Datasets:

eswardivi

/

books

like 0

dure is performed, the UE prioritizes uplink transmission over any V2X sidelink transmission irre- spective of the sidelink MAC PDU's PPPP [6]. Resource pool for transmission of a pedestrian UE (P-UE) may be overlapped with resources dedicated for V2X sidelink communication. For each transmission pool, the resource selection mechanism (i.e., random selection or partial sensing-based selection) is configured. If the P-UE is configured to choose either random selection or partial sensing-based selection for a transmission pool, then it is up to the UE to select a specific resource selection mechanism. If the eNB does not specify a random selection pool, the P-UEs that only support random selection cannot perform sidelink transmission. In excep- tional resource pool, the P-UE uses random selection. The P-UE can send sidelink UE 838 Chapter 7 information message to indicate that it requests resource pools for pedestrian-related V2X sidelink transmission. It is not mandatory for P-UE to support zone-based resource selec- tion. The P-UE reports whether it supports zone-based resource selection as part of UE capability signaling. The P-UEs do not perform CBR measurements; however, they may adjust the transmission parameters based on the default transmission parameter configura- tion, which can be provided to the P-UE via RRC signaling [6]. The LTE V2X messages can be delivered in unicast mode via non-guaranteed bit rate (non-GBR) or GBR bearers. To meet the QoS message delivery requirements for V2X ser- vices, a non-GBR QoS class identifier (QCI) value and a GBR QCI value are used. For broadcast V2X messages, single-cell point-to-multipoint or multimedia broadcast single frequency network transmission can be used. The reception of downlink broadcast V2X messages on different carriers/PLMNs can be supported by having multiple receiver chains in the UE. A GBR QCI value is used for the delivery of V2X messages over eMBMS bearers [6]. 7.6 LTE/NR V2X Security 3GPP LTE has one of the most advanced security mechanisms among al

l wireless technologies and NR V2X will continue to maintain very high standards for V2X security. 3GPP has been developing capabilities that will benefit V2X, starting with enhancements in LTE to support V2X use cases. 5G will evidently support more advanced use cases; therefore, two require- ments are particularly important: (1) the need for direct, ad hoc, broadcast, secure communica- tion without any a priori configuration of security by the network and (2) management of identities for user privacy from the network or other third parties. There are two types of LTE transport-level security mechanisms: (1) LTE security protecting the UE signaling and commu- nications with the LTE network, and (2) LTE D2D or ProSe communications security. LTE security uses a symmetric keying scheme for data protection between the UE and the network. For user-plane data, only confidentiality (encryption) is applied; there is no integrity protection on application-layer messages exchanged with the network. For ProSe/D2D communications, each group of devices conduct one-to-many, confidentiality protected data transfers. This is done using a group symmetric key, which is provisioned by the network to all member devices. There is no integrity protection on the user data and, because the key is shared by all, there is no way to positively identify which of the group members has sent the data. V2X sidelink communication uses the following identities [6]: Source layer-2 ID: Identifies the sender of the data in the sidelink and V2X sidelink communications. The source layer-2 ID is 24 bits long and is used together with desti- nation layer-2 ID and LCID for identification of the RLC UM entity and PDCP entity in the receiver. Vehicle-to-Everything (V2X) Communications 839 Destination layer-2 ID: Identifies the target of the data in the sidelink and V2X sidelink communications. For sidelink communication, the destination layer-2 ID is 24 bits long which is split in the MAC sublayer into two bit strings. The first bit string is the least si

g- nificant bit part (8 bits) of destination layer-2 ID and is forwarded to physical layer as group destination ID. This identifies the target of the intended data in SCI and is used for filtering of packets at the physical layer. The second bit string is the MSB part (16 bits) of the destination layer-2 ID which is carried within the MAC header. This is used for fil- tering of packets at the MAC sublayer. In the case of V2X sidelink communication, desti- nation layer-2 ID is not split and is carried within the MAC header. No access stratum signaling is required for group formation and to configure source layer-2 ID, destination layer-2 ID and group destination ID in the UE. These identities are either provided by higher layer or derived from identities provided by higher layer. In case of groupcast and broadcast, the ProSe UE ID provided by higher layer is used directly as the source layer-2 ID and the ProSe layer-2 group ID provided by higher layer is used directly as the destination layer-2 ID in the MAC sublayer. In the case of one-to-one communica- tions, the ProSe UE ID provided by higher layer is used directly as the source layer-2 ID or the destination layer-2 ID in the MAC sublayer. In the case of V2X sidelink communica- tion, higher layer provides source layer-2 ID and destination layer-2 ID. 5G network security is still under development by 3GPP for device-to-network communica- tions. The C-V2X security was not updated with Rel-15. However, the enhancements devel- oped for network access and associated security aspects will apply to the network-enabled mode of communication. As for the direct mode of operation, the security design for the Rel-14 C-V2X is expected to remain unchanged, namely specifying the reuse of the application-level security already defined by IEEE for DSRC systems. In particular, LTE Rel-14 does not support vehicle identity privacy when it sends V2X traf- fic via the network. Architectural changes would be required to support user/vehicle ano- nymity from the operator. It is expect

ed that 5G core network will enable network mode V2X operation privacy. 5G security is going to define device-to-network authentication methods and transport; specify provisioning and storage of 3GPP credentials for devices; and specify network functions and protocols necessary for secure device operation within the operator network. The V2X architecture relies on the security relationship between device and its home operator network. In addition, secondary authentication schemes can support industrial and virtual private networks that wish to deploy their own authentication methods and credentials, thus it is possible to deploy public key infrastructure (PKI) A public key infrastructure (PKI) is a set of rules, policies, and procedures needed to create, manage, distrib- ute, use, store, and revoke digital certificates and manage public key encryption. In cryptography, a PKI is an arrangement that binds public keys with respective identities of entities. The binding is established through a process of registration and issuance of certificates at and by a certificate authority. 840 Chapter 7 security, where devices use a digital certificate to authenticate themselves to the network and vice versa. 5G security design is being developed to achieve the following goals [23]: Enhanced subscriber/device privacy: This is an improvement over LTE in that the new subscriber permanent identifier (SUPI) is never allowed to be transmitted over the air. Instead, the sensitive part of the SUPI is sent over the radio link protected against spoofing and tracking, which was not the case with the 3G/4G international mobile sub- scriber identity. Enhanced security support at the network level, along with capability for flexible authentication and authorization schemes: A new network function is specified, known as security anchor function (SEAF), which maintains the security anchor deep in a net- work (in a physically secure location). The SEAF also provides flexibility in deploying other network entities such as AMF and SMF. In Rel-

15, the SEAF is co-located with the AMF. Support various types of devices that have different security capabilities and require- ments: Secondary authentication enables support for non-3GPP access links and ses- sions authorized by third-party servers, such as for industrial IoT, V2X, and automation. This functionality is related to the access control for network slices. For example, vari- ous IoT devices may require different credential provisioning and authentication meth- ods. The details for this type of industrial scenarios are being investigated. Support user data and signaling encryption and integrity protection: These features are essential for a secure system. A new feature is data path integrity protection, which may be especially important for certain new services such as industrial IoT. There is also the potential for the user-plane security to be terminated in the network instead of the base station. Separate device credential management and access authentication from data session setup and management: This split results in a separate security context between device and the AMF which is used for mobility management. A different security context is established for session management, between device and the SMF, used to authorize access of the device to specific services; for example, network slices or specific data networks such as enterprise networks. This new type of access control employs a sepa- rate and flexible authentication and authorization procedure; for example, extensible authentication protocol (EAP). Support secure slicing: There could be several services instantiated as a network slice, each with different security requirements. The access to a slice is granted based on the primary authentication and subscription information, but this authorization is carried out Extensible authentication protocol (EAP), defined in IETF RFC 5247, is an authentication framework fre- quently used in wireless networks and point-to-point connections which provides an authentication frame- work for transport

and usage of keying material and parameters generated by EAP methods. Each protocol that uses EAP defines a method to encapsulate EAP messages within that protocol's messages. Vehicle-to-Everything (V2X) Communications 841 by the respective SMF. Therefore, the access security is contained within that network slice and does not rely on the AMF, which may serve multiple slices that may have dif- ferent security requirements. Moreover, an attack mounted on one slice does not result in an increased risk for an attack on a different slice of the same network. In a nutshell, what 5G security is trying to achieve is increased user privacy, robustness to cyber-attacks on the network, and better device hardware security. These goals can be achieved with stronger authentication/authorization schemes between device and network, both radio access and core network functions, secure credential provisioning and storage on device, and new network functions that support device-to-network communications security. Cellular-based V2X systems treat latency as the most important performance metric because the level of protection decreases as the delay in receiving safety information increases. While the delay in non-mission-critical applications may be tolerated to some extent, delayed information in V2X communications could result in serious automobile accidents and injuries. The volume of traffic in V2X communication is much smaller than other appli- cations in cellular systems. Sensing information or safety notifications transmitted via a V2X link can be carried in small packets, thus high-speed data transmissions are less impor- tant in V2X systems. The device-to-core network (AMF) signaling is integrity- and confidentiality-protected. The device link to the radio access network is also protected for both signaling and data traffic. The V2X system can leverage the 5G system security for vehicle device authorization, authentication, and access to the network. 7.7 Implementation and Deployment Considerations The DSRC would require

the deployment of tens of thousands of RSUs embedded or attached to roadway infrastructure to enable an effective network along the country roads. This would be a challenge in rural areas considering the vast distances involved. State high- way administrations and other highway authorities would be responsible for deploying, managing, and operating the RSUs and the associated infrastructure network and inter- connections such as fiber or copper backhaul. After considering how to map each V2V ser- vice to different 3GPP technologies, the conclusion is that Rel-14 is only used for basic safety, while NR-based V2X is used for advanced services to avoid duplication or replace- ment of LTE-based Rel-14 functionalities. Therefore, it is expected that in the beginning of the V2X service deployment, there will be LTE-based Rel-14 V2X devices which later evolve into dual-mode UEs that support both LTE and NR V2X services. Several V2X use cases require vehicles to communicate with an infrastructure. The DSRC security relies on a public key infrastructure that distributes and manages digital certificates for vehicles. This means that vehicles need to have access to this infrastructure, which in Chapter 7 the case of DSRC is provided via the RSUs. The RSU may also be used by vehicles to com- municate with the V2X application server. Owing to various factors, the deployment of RSUs might be limited. It is therefore unrealistic to expect the provision of ubiquitous COV- erage of roadways via DSRC equipment in near future. One challenge with the deployment of V2X technologies is that there is uncertain business incentive for network providers. In the United States, where V2V deployments may be mandated, there is still lack of clarity regarding the plan to implement the infrastructure that would utilize the balance of the channels, and which entities should manage the security network functions. Since govern- ment funding has been a driving force for many V2X pilot programs, it is unclear if com- mercial business models can be a

pplied to accelerate infrastructure deployment or if deployments will be managed by the governments. In some regions, the government agen- cies are mainly promoting pilot projects in order to benefit the economy of cities and regions by improving traffic efficiency, reducing emissions, and minimizing the risk of crashes. With C-V2X, mobile operator involvement can make additional, commercially motivated services available. The beneficiaries of these can be both subscribers and road operators. References 3GPP Specifications11 [1] 3GPP TS 22.185. Service requirements for V2X services; Stage 1 (Release 15); June 2018. [2] 3GPP TS 23.285. Architecture enhancements for V2X services (Release 15); June 2018. 3GPP TS 36.211. Evolved universal terrestrial radio access (E-UTRA). Physical channels and modulation (Release 15); June 2018. [4] 3GPP TS 36.212. Evolved universal terrestrial radio access (E-UTRA). Multiplexing and channel coding (Release 15); June 2018. [5] 3GPP TS 36.213. Evolved universal terrestrial radio access (E-UTRA). Physical layer procedures (Release 15); June 2018. [6] 3GPP TS 36.300. Evolved universal terrestrial radio access (E-UTRA) and evolved universal terrestrial radio access network (E-UTRAN); Overall description; Stage 2 (Release 15); March 2019. 3GPP TS 36.321. Evolved universal terrestrial radio access (E-UTRA). Medium access control (MAC) protocol specification (Release 15); June 2018. [8] 3GPP TS 36.331. Evolved universal terrestrial radio access (E-UTRA). Radio Resource Control (RRC); Protocol Specification (Release 15); June 2018. [9] 3GPP TR 36.885. Study on LTE-based V2X services (Release 14); July 2016. [10] 3GPP TR 37.885. Study on evaluation methodology of new vehicle-to-everything V2X use cases for LTE and NR (Release 15); June 2018. [11] 3GPP TR 38.885. Study on NR vehicle-to-everything (V2X) (Release 16); March 2019. Articles, Books, White Papers, and Application Notes [12] Recommendation ITU-R M.2084-0. Radio interface standards of vehicle-to-vehicle and vehicle-to- infrastructure

communications for Intelligent Transport System applications; September 2015. 3GPP specifications can be accessed at the following URL: http://www.3gpp.org/ftp/Specs/archive/ Vehicle-to-Everything (V2X) Communications [13] 5G Americas White Paper. V2X cellular solutions; October 2016. [14] 5G Americas White Paper. Cellular V2X communications towards 5G; March 2018. [15] K. Lee, et al., Latency of cellular-based V2X: Perspectives on TTI-proportional latency and TTI-independent latency, IEEE Access 5 (2017). [16] M. Boban, et al., Use cases, requirements, and design considerations for 5G V2X, IEEE Vehicular Technol Mag, December 2017. [17] R. Molina-Masegosa, J. Gozalvez, LTE-V for sidelink 5G V2X vehicular communications, IEEE Vehicular Technol Mag, December 2017. [18] Rohde & Schwarz White Paper. LTE-advanced (Release 12) technology introduction; September 2014. [19] Rohde & Schwarz White Paper. Device to device communication in LTE; September 2015. [20] Husain S, et al. An overview of standardization efforts for enabling vehicular-to-everything services. IEEE Conference on Standards for Communications and Networking (CSCN); September 2017. S. Chen, et al., Vehicle-to-everything (V2X) services supported by LTE-based systems and 5G, IEEE Commun Standards Mag, June 2017. [22] H. Seo, et al., LTE evolution for vehicle-to-everything services, IEEE Commun Mag, June 2016. [23] Kousaridas A, et al. Recent advances in 3GPP networks for vehicular communications. IEEE Conference on Standards for Communications and Networking (CSCN); September 2017. [24] ETSI TS 102 687 V1.2.1. Intelligent transport systems (ITS); Decentralized congestion control mechan- isms for intelligent transport systems operating in the 5 GHz range; Access Layer Part; April 2018. [25] Khoryaev A. Evolution of cellular-V2X (C-V2X) technology use cases, technical challenges, and radio- layer solutions for connected cars. IEEE ComSoc Webinar; March 2018. 5GAA Automotive Association. An assessment of LTE-V2X (PC5) and 802.11p direct communications techno

logies for improved road safety in the EU; December 2017. [27] Qualcomm Technologies Inc. Accelerating C-V2X commercialization; 2017. [28] Gaurang Naik et al., IEEE 802.11bd & 5G NR V2X: Evolution of Radio Access Technologies for V2X Communications, Cornell University Online Library, March 2019. CHAPTER 8 Operation in Unlicensed and Shared Spectrum The RF spectrum is divided into licensed and license-exempt/unlicensed bands. The tradi- tional cellular communications systems used to exclusively operate in licensed spectrum. However, due to the insufficiency and cost of licensed spectrum below 6 GHz, the telecom- munication industry has shifted attention to the unlicensed bands to deploy supplementary uplink/downlink carriers or standalone systems. The use of unlicensed spectrum has been increasingly considered by cellular operators as a complementary mechanism to offload traffic from licensed-band RF carriers to increase the overall system throughput. Various approaches to cellular operation in unlicensed spectrum have been investigated and trialed in the past few years. Two practical cellular technologies for communication in unlicensed spectrum, the LTE- U and the LTE-based license-assisted access (LAA), have been investigated by LTE-U Forum and 3GPP. The use of Wi-Fi as a complementary carrier to offload cellular networks was stud- ied earlier and specified by 3GPP under LTE-Wi-Fi link aggregation (LWA) work item [12]. Operation in unlicensed spectrum is subject to various limitations and restrictions which are regional and band specific. The typical limits are in terms of total transmit power, power spectral density (PSD), carrier bandwidth, and duty cycle that each device can use. In addi- tion, sharing protocols may also be specified in some bands to protect other systems in the band or to allow efficient sharing. As an example, dynamic frequency selection (DFS) aims to protect radar systems and listen-before-talk (LBT) allows efficient spectrum sharing by minimizing inter-user interference in unlicensed spe

ctrum. 3GPP LTE began to explore 5 GHz unlicensed band for offloading cellular traffic in Rel-12. The non-standard LTE-U and standard LAA both use LTE technology as the baseline in unlicensed spectrum anchored on a licensed carrier to increase user throughput and system capacity. It must be noted that LTE-U can only be deployed in selected regions due to lack of LBT support whereas LTE-based LAA uses a similar LBT procedure as IEEE 802.11 systems. In addi- tion to LTE-U and LAA, the standalone operation of LTE in the unlicensed spectrum has been considered by some operators. 1 MulteFire Alliance: https://www.multefire.org 5G NR. DOI:https://doi.org/10.1016/B978-0-08-102267-2.00008-7 © 2019 Elsevier Inc. All rights reserved. 846 Chapter 8 The 60 GHz band has yet to attain significant large-scale industry traction although the specification for IEEE 802.11ad2 system was finalized in 2012. The 60 GHz frequency bands have been used for wireless backhaul and indoor and outdoor point-to-point commu- nication. The unlicensed 5 GHz spectrum provides about 500 MHz of usable bandwidth. A large number of Wi-Fi deployments based on the IEEE 802.11 are in 5 GHz band and are mainly used for broadband indoor applications. Cellular operators actively use Wi-Fi to off- load traffic in dense hotspots using unlicensed spectrum; however, due to a high level of inter-cell interference, Wi-Fi-based offloading scenarios are not attractive for outdoor use cases. Multi-antenna transmission and beamforming with large antenna arrays are some of the key features of the new radio systems, which can benefit the NR-unlicensed operation in the sense that transmission power can be reduced due to highly directional antennas. In addi- tion, directional transmission leads to lower inter-user/inter-cell interference, which in turn improves the coexistence conditions. For typical network deployment where the elevation angle of the antenna is manually tilted to one fixed direction, the maximum antenna gain is only achieved in the tilted elevation dire

ction while lower antenna gain is seen in other directions. When both elevation and horizontal beamforming are utilized (FD-MIMO), the maximum beamforming gain can be obtained in each direction via beam tracking proce- dures. Beamforming can improve the link budget and can increase signal-to-interference- plus-noise ratio (SINR) performance. Moreover, the interference between different systems in the unlicensed band can be minimized using highly directional antennas due to the signif- icantly lower collision probability compared to the omnidirectional transmission cases. As a result, higher spatial reuse would be possible, increasing the system throughput as well as improving spectrum efficiency. Some licensed frequency bands are designated for commercial use while others are desig- nated for public safety. The assignment of a frequency band to a user allows the user to uti- lize that frequency and bandwidth for stated purposes using predefined emission parameters. Licensed spectrum allows exclusive use of certain frequencies in specific geo- graphic locations, meaning that when someone is granted the right by a regulatory body to communicate at certain frequencies and in certain locations, everyone else is prohibited from using that frequency in that location. In contrast, in the spectrum that is designated as license-exempt or unlicensed, users can operate without a license but must comply with the regulatory constraints. For instance, the regulations limit the transmit power and the effec- tive isotropic radiated power (EIRP) to minimize interference to other cochannel systems. The unlicensed frequency bands were originally allocated for industrial, scientific, and med- ical (ISM bands) applications. IEEE Standards Association: https://standards.ieee.org/findstds/standard/802.11ad-2012.html Operation in Unlicensed and Shared Spectrum 847 To coexist with Wi-Fi in the unlicensed spectrum, some enhancements in the LTE and NR systems were required including a mechanism for channel sensing based on LBT, dis- contin

uous transmission on a carrier with limited maximum transmission duration, DFS for radar avoidance in certain bands, and multi-carrier transmission across multiple unli- censed channels. The DTX and LBT have had major impacts on various aspects of LTE functionalities including downlink/uplink physical channel design, channel state informa- tion (CSI) estimation and reporting, HARQ operation, and radio resource management. In 5 GHz band, about 500 MHz of unlicensed spectrum is available for LAA use. This unli- censed band can be divided into multiple channels of 20 MHz bandwidth. The selection of LAA carrier(s) with minimal interference is the first step for an LAA node to coexist with Wi-Fi systems in the unlicensed spectrum. However, when a large number of nodes are present, interference avoidance cannot be guaranteed through channel selection, thereby, sharing carriers between different technologies is required. The carrier selection can be further performed periodically by adding or removing unlicensed carriers as required. These carriers are then configured and activated as secondary cells (SCells) for use by LAA-enabled UEs. The 3GPP NR has initiated a study item in Rel-16 to investigate the required amendments in the Rel-15 NR to extend the operation to the unlicensed bands in sub-6 GHz and above 6 GHz. Meanwhile, the LTE track has continued to enhance the LAA and non- homogeneous LTE/Wi-Fi carrier aggregation solutions, which global operators have already started to deploy. In this chapter, we will review the general aspects and use cases of the operation in unlicensed spectrum including identifying the unlicensed frequency bands, overview of IEEE 802.11 operation, which is important for understanding the coexistence studies, as well as the necessary NR and LTE enhancements to address the unlicensed band operation from various aspects, including network architecture, radio access protocol struc- ture, and L1/L2 processing, as well as deployment and implementation considerations. 8.1 General Aspects and Use

Cases 8.1.1 Unlicensed and Shared Spectrum Unlicensed/license-exempt bands are a spectrum that has been defined for use collectively by an undetermined number of independent users without registration or individual permis- sion. For unlicensed bands, the regulatory bodies establish rules for how applications, tech- nologies, and industries must use the spectrum that allows applications and users to coexist with limited interference to each other. The rules are defined openly with no limitation on technologies and applications other than requirements to avoid/reduce destructive interfer- ence. With unlicensed spectrum, there is no process for establishing the right of use, and therefore the band may be utilized by any device that is compliant with usage rules such as 848 Chapter 8 maximum power levels, bandwidth limitations, and duty cycles. The use of unlicensed spec- trum is an important complement for all 5G systems and deployments, particularly in small cell deployments. Table 8.1 provides a summary of such rules for 5 GHz band utilization across the world [9,13,15,21]. In recent years, short-range communication technologies such as Bluetooth, 3 ZigBee, 4 Wi-Fi have utilized unlicensed bands for short-range communication services. The fre- quency mapping of unlicensed bands is country dependent, which determines the type of technology used within designated parts of the spectrum. Despite the original intent, radio communications in the ISM bands are possible as long as the communication systems are designed to tolerate the inter-user interference as well as the cochannel interference poten- tially from other communication systems operating in the same band. Since the advent of mobile devices, more and more short-range, low-power, low-cost wireless communication systems, such as cordless phones, Wi-Fi, Bluetooth, and ZigBee, have utilized some of the unlicensed bands in 902-928 MHz, 2.40-2.4835 GHz, and 5.725-5.875 GHz. The grow- ing number of wireless applications in the ISM bands has motivated the wireless i

ndustry to increase the amount of spectrum available for unlicensed use. In the United States, the Federal Communications Commission (FCC) made 300 MHz of spectrum available in 5.15-5.25 GHz (UNII-1), 5.25-5.35 GHz (UNII-2A), including 5.725-5.825 GHz (UNII- 3), for use by a new category of unlicensed equipment. The latter band is partially over- lapped with the ISM band (5.725-5.875 GHz), hence, is sometimes referred to as Unlicensed National Information Infrastructure (U-NII)/ISM. The FCC further made avail- able additional 255 MHz spectrum in 5.47-5.725 MHz (UNII-2C) band. This aligns the U- NII frequency band in the United States with other parts of the world, thereby allowing the same product to be used in most parts of the world. The frequency ranges 5.250-5.350 and 5.470-5.725 GHz in U-NII-2 are used by radar systems worldwide; thus the use of these bands requires DFS techniques to avoid adverse effects on radar operation [20]. In certain geographical regions, such as the European Union and Japan, support of LBT rule is manda- tory to reduce the interference to other users operating in the same band. The LBT medium access rule prevents a transmitter from continuous transmission and dominating the commu- nication channel, rather it requires the transmitter to check for other radio activities in the channel prior to transmission. In the United States, the use of 5 GHz unlicensed spectrum is subject to FCC regulations. At present, unlicensed wireless systems can access bands 5.15-5.25 GHz (UNII-1), 5.25-5.35 GHz (UNII-2A), 5.47-5.725 GHz (UNII-2C), and 5.725-5.85 GHz (UNII-3). In addition, bands 5.35-5.47 GHz (UNII-2B) and 5.85-5.925 GHz (UNII-4) are also being Bluetooth Special Interest Group: https://www.bluetooth.com/ ZigBee Alliance: https://www.zigbee.org/ Wi-Fi Alliance: https://www.wi-fi.org/ Table 8.1: 5 GHz unlicensed band designations worldwide [18]. 5 GHz UNII-1 UNII-2A UNII-2B Region 100 MHz 100 MHz 120 MHz UNII-2C 255 MHz UNII-3 125 MHz UNII-4 75 MHz (Total Bandwidth) 5.15-5.25 GHz 5.25-5.35 GHz

5.35-5.47 GHz 5.47-5.59 GHz 5.59-5.65 GHz 5.65-5.725 GHz 5.725-5.825 GHz 5.82-5.85 GHz 5.85-5.925 GHz China Indoor Indoor DFS/ Indoor/outdoor DFS/TPC (325 MHz) Europe Indoor LBT Indoor/outdoor Indoor/outdoor (455 MHz) DFS/TPC L DFS/TPC LBT Japan Indoor LBT Indoor/outdoor Indoor/outdoor (455 MHz) DFS/TPC LBT DFS/TPC LBT Korea Indoor Indoor/outdoor Indoor/outdoor Indoor/outdoor (480 MHz) DFS/TPC DFS/TPC The United Indoor/outdoor Indoor/outdoor Indoor/outdoor Indoor/outdoor States DFS/TPC DFS/TPC (580 MHz) 850 Chapter 8 considered for unlicensed use. The FCC has some regulations regarding transmission band- width, maximum transmit power, out-of-band emission, power spectrum density, transmit power control (TPC), and DFS for each unlicensed band. For example, the maximum trans- mit power is 24 dBm in the UNII-1 and UNII-2A bands, and 30 dBm in the UNII-2C and UNII-3 bands. In addition to the maximum transmit power, TPC may further limit the out- put power of a transmitter to minimize interference to users of other wireless technologies. In fact, TPC is required for both the UNII-2A and UNII-2C bands. The DFS is used for unlicensed devices to detect radar signals and to change their operating channels whenever radar activity is detected. The DFS should be adopted in the UNII-2A and UNII-2C bands to protect radar signals. Unlike LTE-based LAA/eLAA that only provide support for the 5 GHz unlicensed band, the NR is required to cover a wide range of unlicensed and shared licensed frequency bands, where regulatory and inter-RAT coexistence requirements may differ for each band. This includes 3.5 GHz band that has been designated as a shared band in the United States as well as the 5 and 60 GHz bands that are unlicensed bands. Moreover, the system design needs to consider the vastly different wireless channel characteristics for the lower carrier frequency bands such as sub-6 GHz as well as higher carrier frequency bands such as 60 GHz [26]. 5G networks are expected to operate over a wide range of licensed and unlicensed fr

equen- cies in low, medium, and high spectrum bands, but some of those frequencies are yet to be specifically defined by 3GPP and ITU-R. It can be assumed that most of the bands currently being used for 4G networks will be ultimately re-allocated to 5G technologies. Meanwhile, worldwide activities have already begun to explore a number of bands between 24.25 and 86 GHz that are being studied for the 2019 World Radiocommunication Conference (WRC), as well as on bands not included in the WRC agenda item. In the United States, the FCC is planning to make 64-71 GHz band available for unlicensed use based on the same rules applicable to the unlicensed 57-64 GHz band. In addition, FCC is studying several other bands in 24 GHz and above. It further plans to make some bands available for unli- censed applications using the same rules applicable to the unlicensed 57-64 GHz band. However, it has been asked to reconsider allocating the entire 64-71 GHz band to unli- censed operations. 3GPP also has a study item on 5G in unlicensed bands below and above 6 GHz. Unlicensed bands above 52.6 GHz covering the FCC 64-71 GHz frequency range will be considered to the extent that waveform design principles remain unchanged relative to that in below 52.6 GHz bands. The FCC has allowed access to spectrum for next- generation wireless broadband in the 28 GHz (27.5-28.35 GHz), 37 GHz (37-38.6 GHz), and 39 GHz (38.6-40 GHz) bands, as well as an unlicensed band at 64-71 GHz. The new rules make available almost 11 GHz of spectrum consisting of 3.85 GHz of the licensed spectrum and 7 GHz of the unlicensed spectrum [31]. As shown in Fig. 8.1, 60 GHz unli- censed spectrum comprises up to four non-overlapping channels, where each channel has a Operation in Unlicensed and Shared Spectrum Channel Channel 2 Channel 3 Channel 4 (57.24-59.40 GHz) (59.40-61.56 GHz) 61.56-63.72 GHz) (63.72-65.88 GHz) US (57.05-64 GHz) Europe (57-66 GHz) Korea (57-64 GHz) Japan (59-66 GHz) China (59-64 GHz) Figure 8.1 Worldwide 60 GHz unlicensed spectrum allocation [26

]. bandwidth of 2.16 GHz and enables four channels in Europe; three channels in the United States, Canada, South Korea, and Japan, and two channels in China. The requirements and regulations concerning the EIRP for 60 GHz vary by geographic areas and are different from the ones in 5 GHz. For example, the requirement for peak EIRP is 43 dBm in the United States and 40 dBm in Europe where the maximum PSD is also required to be 13 dBm/MHz. The main challenge of operation in 60 GHz band is excessive path loss and blockage relative to operation in 5 GHz band. However, the large propagation loss at 60 GHz can be mitigated by increasing the antenna array gain obtained by large array sizes. Fig. 8.1 summarizes the worldwide availability and channelization of 60 GHz spectrum. The US FCC established Citizen Broadband Radio Service (CBRS) for shared commercial use of the 3.5 GHz (3550-3700 MHz) band with the incumbent military radars and fixed satellite stations in 2015. Dynamic spectrum sharing rules have been defined to make additional spectrum available for flexible wireless broadband use while ensuring interfer- ence protection and uninterrupted use by the incumbent users. Under the plan, a three-tier sharing framework coordinates spectrum access among the incumbent military radars and satellite ground stations and new commercial users. The three tiers are as follows: incum- bent, priority access license (PAL), and general authorized access (GAA) users, as shown in Fig. 8.2. The incumbent military radar systems, satellite ground stations, and wireless ISPs are always protected from possible interference from the lower-tier PAL and GAA users. Tier-2 PAL users have the next highest priority access and are protected from GAA users. 852 Chapter 8 3550MHz 3600MHz 3650MHz 3700MHz US military radars (on-ground, ship-borne) Tier-1 (Incumbents) Fixed satellite stations Primarily in the coastal areas and some inlands Wireless ISPs to transition to Tier-2/3 Wireless ISPs Tier-2 (PAL) Priority access license (PAL) Up to 7x10 MHz ch

annels Three-year term by census tract Tier-3 (GAA) General authorized access (GAA) At least 80 MHz of commercial use Figure 8.2 CBRS three-tier shared spectrum licensing structure [19]. PAL licenses within the 3550-3650 MHz portion of the band are assigned based on spec- trum auctions. Each PAL license covers a 10 MHz channel for a single census tract (i.e., a small, relatively permanent statistical subdivision of a county) for a 3-year term. For any given census tract, up to seven total PALs may be assigned. With over 50,000 census tracts in the United States, each PAL spectrum license is expected to cost much less and encour- age participation from a variety of participants. It should be noted that a PAL frequency range may change over time based on incumbent activity. The lowest-tier GAA users are permitted to use any portion of the 3.5 GHz band not assigned to higher-tier users. With an open-access rule, GAA provides free access to the spectrum similar to unlicensed spectrum. Because PAL licenses are limited to a maximum of 70 MHz in any given census tract, a minimum of 80 MHz bandwidth is available for GAA use when not in use by the incumbent users. While GAA operation does not require a costly license, GAA operators must coordi- nate their use of the spectrum through the dynamic spectrum sharing system [19]. As shown in Fig. 8.3, a key element of the CBRS spectrum sharing architecture is the spec- trum access system (SAS). A SAS maintains a database of all CBRS base stations, also known as CBRS devices (CBSDs), including their tier status, geographical location, and other pertinent information to coordinate channel assignments and manage potential inter- ferences. To mitigate possible interference to tier-1 military radar systems, environmental sensors known as environmental sensing capability (ESC) will be deployed in strategic loca- tions near naval stations, mostly along coastal regions, to detect incumbent activities. When incumbent use is detected, the ESC alerts the SAS, which then directs CBSDs util

izing impacted CBRS channels in that area to move to other channels. The cloud-based SAS enforces the three-tier spectrum sharing mechanism based on FCC rules via centralized, dynamic coordination of spectrum channel assignments across all CBRS base stations in a region. The CBRS rule-making defines two classes of base stations: class-A and class-B. A Operation in Unlicensed and Shared Spectrum 853 Environmental sensing Spectrum access system capability ((IIII) (())) (())) ((a)) ((< HI) CBSD Class A (indoor small cell) CBSD Class B (outdoor station) Figure 8.3 CBRS network architecture [19]. class-A base station can be an indoor or low-power outdoor small cell with a maximum conducted power of 24 dBm (per 10 MHz) and a maximum EIRP of 30 dBm. This type of small cell is similar to commercial enterprise-class small cells with 250 mW transmit power and a typical 2 dBi omnidirectional antenna or up to 6 dBi directional antenna. The class-B base station is meant for outdoor use with a maximum EIRP of 47 dBm. The highly direc- tional antenna makes the outdoor CBRS base stations more suitable for fixed wireless use cases. While indoor and outdoor base stations can be assigned to either GAA or PAL, more indoor GAA deployments are expected until ESC certification and PAL auctions are con- cluded [19]. 8.1.2 Use Cases and Deployment Models The introduction of carrier aggregation in LTE-advanced required the distinction between a primary cell (PCell) and SCell. The PCell is the main cell with which a UE communicates and maintains its connection with the network. One or more SCells can be allocated to and activated for the UEs supporting carrier aggregation for bandwidth extension. Because the unlicensed carrier is shared by multiple systems, it can never match the licensed carrier in terms of mobility, reliability, and quality of service (QoS). Hence, in LAA, the unlicensed carrier is considered only as a supplemental downlink or uplink SCell assisted by a licensed PCell via carrier aggregation. LAA deployment scenarios enc

F1: Licensed carrier(s) Ideal/non-ideal Ideal/nonideal backhaul F2: Licensed backhaul carrier(s) Small-cell Small-cell Cluster Cluster F3: Unlicensed carrier(s) Ideal backhaul Ideal backhaul (co-located) F3: Unlicensed (co-located) carrier(s) Small-cell Small-cell Cluster Cluster Scenario 3 Scenario 4 Figure 8.4 LAA deployment scenarios [12]. Operation in Unlicensed and Shared Spectrum 855 context of the carrier aggregation through ideal backhaul with a licensed cell. In scenarios where carrier aggregation is used within the small cell with carriers in both licensed and unlicensed bands, the backhaul between macrocell and small cell can be ideal or non-ideal. The deployment scenarios depicted in Fig. 8.4 can be described as follows. In scenario 1, carrier aggregation between licensed macrocell (F1) and unlicensed small cell (F3) is used, whereas in scenario 2, carrier aggregation between licensed small cell (F2) and unlicensed small cell (F3) without macrocell coverage is intended. In scenario 3, licensed macrocell and small cell (F1) with carrier aggregation between licensed small cell (F1) and unlicensed small cell (F3) is utilized, and in scenario 4, licensed macrocell (F1), licensed small cell (F2), and unlicensed small cell (F3) and further carrier aggregation between licensed small cell (F2) and unlicensed small cell (F3) is used. In the latter scenario, if there is ideal back- haul between macrocell and small cell, there can be carrier aggregation between macrocell (F1), licensed small cell (F2), and unlicensed small cell (F3). If dual connectivity is enabled, there can be dual connectivity between macrocell and small cell. In the study to support deployment in the unlicensed spectrum for the above scenarios, carrier aggregation functionalities are used as a baseline to aggregate PCell/PSCell6 on a licensed carrier and SCell on an unlicensed car- rier. When non-ideal backhaul is applied between a macrocell and a small cell cluster in scenar- ios 3 and 4, a small cell on an unlicensed carrier must be aggreg

ated with a small cell on a licensed carrier in the small cell cluster through ideal backhaul. The NR-unlicensed (NR-U) study item has considered different unlicensed bands or shared bands such as 2.4, 3.5, 5, 6, 37, and 60 GHz band. Some bands are available globally whereas others are only regionally available. The NR-U study item does not target sub- gigahertz unlicensed bands due to the lack of sufficient spectrum to support efficient NR-U operation. Five deployment scenarios have been identified for NR-U as follows [12]: Scenario A: Carrier aggregation between licensed-band NR (PCell) and NR-U (SCell) NR-U SCell may have both downlink and uplink or only downlink coverage Scenario B: Dual connectivity between licensed-band LTE (PCell) and NR-U (PSCell) Scenario C: Standalone NR-U Scenario D: An NR cell with a downlink in an unlicensed band and uplink in a licensed band In LTE Rel-10 and Rel-11 carrier aggregation, PUCCH was configured only on the PCell. Therefore, uplink control information, that is, HARQ ACK/NACK, CSI of SCells were transmitted on PUCCH of the PCell, if not multiplexed with SCell's own PUSCH. In LTE dual connectivity, it was determined not suitable to carry UCIs of SeNB in PUCCH of PCell in MeNB due to backhaul latency. Therefore, PUCCH is configured on a special SCell of SeNB called Primary SCell or PSCell. The PSCell is never deactivated and RACH procedure needs to be initiated upon its initial configuration. 856 Chapter 8 Scenario E: Dual connectivity between licensed-band NR (PCell) and NR-U (PSCell) The references to sub-7 GHz bands are intended to include unlicensed bands in 6 GHz region that are under consideration by some regulatory bodies. NR-U is trying to identify additional functionalities that may need a new physical layer design (except channel access procedures) for operation in the unlicensed spectrum that are applicable to a particular fre- quency range (e.g., sub-7 GHz, 7-52.6 GHz, above 52.6 GHz). Optimizations for some fre- quency bands may be necessary due to different re

quirements for each band such as PSD limitation or occupied channel bandwidth (OCB), defined as the bandwidth containing 99% of the signal power, which falls between 80% and 100% of the nominal channel bandwidth. Channel bandwidths below 5 MHz are not included in this study. The study further targets the design of channel access procedures for frequency bands based on coexistence and regu- latory considerations applicable to the band. The study includes identification of procedures for technology-neutral channel access for frequency bands that may become available sub- ject to new regulations. If the absence of a Wi-Fi system cannot be guaranteed in sub- 7 GHz band where NR-U is operating, the baseline assumption is that NR-U operating bandwidth is an integer multiple of 20 MHz. The following physical layer procedures may be impacted as a result of the study for NR-U: HARQ operation; configured grant support in NR-U; and channel access mechanisms. For the latter, LTE-based LAA LBT mechanism is used as the baseline. The study may further encompass enhancement of the baseline LBT mechanism; techniques to cope with directional antennas/transmissions: receiver-assisted LBT; on-demand receiver-assisted LBT, for example, receiver-assisted LBT enabled only when needed; and techniques to enhance spatial reuse, preamble detection, and enhance- ments to baseline LBT mechanism above 7 GHz. The citizens broadband radio service can potentially lower the barrier to entry for non-traditional wireless carriers. The limited propagation characteristics of the 3.5 GHz spectrum facilitate indoor and floor-by-floor deployment options that could compete with the existing Wi-Fi networks. Due to the significantly lower cost of PALs compared to licensed spectrum costs, private operators now have access to 150 MHz of spectrum on every floor. This may allow enterprise applications, industrial Internet of things, and densely populated venues. Trials are already happening in the industrial IoT and smart home areas. PALs fit the need for loca

l connectivity in remote or temporary locations for industrial com- plexes such as mines, power plants, oil platforms, factories, and warehouses. Private and localized LTE deployments combine the QoS of LTE and the low cost of unlicensed spec- trum. Aside from industrial IoT players and general network operators looking for more avail- able bandwidth, the SAS model for CBRS encourages more players to participate in the eventual deployment of 5G technology. Low license fees and neutral hosts allow non-traditional cellular carriers to build private networks independent of exclusively licensed frequencies or heavily congested unlicensed spectrum in 5 GHz band. Operation in Unlicensed and Shared Spectrum 857 8.1.3 Principles of IEEE 802.11 Operation IEEE 802.11 family of standards is a group of wireless local area network radio access tech- nologies developed by IEEE 802 LAN/MAN Standards Committee. These standards define non-synchronous contention-based multiple access schemes based on carrier sense multiple access with collision avoidance (CSMA/CA). Unlike LTE, Wi-Fi takes a decentralized approach to initiate transmissions from different devices. The rule is to listen before you talk, and if your transmission collides with another transmission, wait a random period before you try again. Therefore, when a Wi-Fi device wants to begin a transmission, it senses the medium and performs a clear channel assessment (CCA). If the channel is detected to be free for a time duration referred to as distributed interframe space, or DIFS, the transmission proceeds. Otherwise, the Wi-Fi device draws a random number, between 0 and 16 (or between 0 and 32 for IEEE 802.11b/g), starts a counter and backs-off from trans- mission while the channel is busy. When the counter reaches zero, the device that was attempting to transmit starts transmission over the channel. However, if other devices were also sensing the carrier at the same time and tried to transmit, a collision may occur. When a transmission fails (which is detected by the ab

sence of an ACK from the receiver), a ran- dom backoff number is drawn and the process repeats. With every backoff, the random counter value is doubled, that is, increasing as 16, 32, 64, etc. This random-access process, referred to as CSMA/CA, is illustrated in Fig. 8.5. Most wireless systems operating in unlicensed spectrum employ CSMA as the basis for channel access. In this case, the most commonly used medium access mechanism is distrib- uted coordination function (DCF), which is based on CSMA/CA. The DCF concept is illus- trated in Fig. 8.5. A transmitting node senses the channel, and if the channel is idle for a certain time duration, that is, the DCF interframe space (DIFS), the node starts to transmit; otherwise, it continues to monitor the channel until it becomes idle for a time duration equal to DIFS. At this time, the node generates a random backoff timer, uniformly distributed within a contention window (CW). The random backoff helps avoid potential collisions, which may happen when two or more nodes are simultaneously waiting for the channel to be cleared. The backoff timer is decremented as long as the channel is idle but remains fro- zen when a transmission is detected and is reactivated after the DIFS period of time as soon as the channel is free. A node refrains from transmission until its backoff timer expires. Note that the DCF mechanism tries to ensure that only one transmission is present in a channel at any time, and each node has a fair share of the channel. The channel use for each node at a particular time is not guaranteed. Thus there is no deterministic schedule transmission, reflecting the random and contentious nature of communication in the unli- censed spectrum. Reliable services and efficient resource usage are typically hard to achieve in unlicensed operation. This property is very different from that of operation in a licensed spectrum. 858 Chapter 8 Device A Random Medium busy Medium free backoff Device B (attempts to send) Carrier sensing Medium busy Medium free Medium busy De

vice C (attends to send) Carrier sensing Collision Random backoff = 14 Device A 13 12 11 10 9 8 7 6 5 Data A Random backoff = 10 Random backoff = 7 Device B 9876543210 Data B Random backoff = 16 Device C -DIFS Data C Signal Preamble symbol Data symbols Figure 8.5 Illustration of the asynchronous Wi-Fi channel access concept [30]. A generic IEEE 802.11 frame structure consists of a set of preambles (new and legacy pre- ambles), a signal field, and multiple data symbols. The preamble is a special waveform designed for signal identification, AGC adjustment, timing and frequency synchronization, and channel estimation. The preamble is particularly suited for activity detection in a chan- nel because waveform detection is 10-20 dB more sensitive than the energy-detection- based CCA. Furthermore, because the IEEE 802.11 frame structure does not have a fixed timing, the preamble detection is crucial for a receiver to synchronize to the incoming frame. The signal symbol following the preamble contains information that includes the modulation and coding scheme and the total number of octets for the following data sym- bols that carry a MAC PDU. Depending on the payload, the frame can be very short (e.g., ACK frame) or long in the case of data frame for user traffic. Each IEEE 802.11 MAC PDU contains a transmission duration field for informing the neighboring nodes of the Operation in Unlicensed and Shared Spectrum 859 medium occupancy time of the current burst. This is the amount of time that all nodes must wait, if they receive it. A local timer, called a network allocation vector (NAV), of a neigh- boring node is updated after the node reads the duration value from the ongoing transmis- sion. This node avoids medium access until the NAV counter expires. Using this virtual medium sensing mechanism, Wi-Fi utilizes a special clear-to-send (CTS)-to-self message to deal with the newer versions of Wi-Fi frames coexisting with a legacy node. CTS-to-self is a standard Wi-Fi CTS message except that it is addressed to the transmi

tting node itself, as the name implies. Nevertheless, it is meant for the neighboring nodes, if it is detected. A new generation Wi-Fi node transmits a CTS-to-self frame immediately before transmitting. The duration field of the CTS-to-self packet contains the time of the following traffic frame, thereby providing more effective protection of the subsequent frame than relying on physi- cal medium sensing. IEEE 802.11 supports two network architecture types namely infrastructure and ad-hoc modes. The basic service set (BSS) is the basic building block of an IEEE 802.11 network. Direct association of stations in ad-hoc network forms an independent BSS or IBSS. The interconnection of a number of BSS through a distributed system creates an extended ser- vice set. IEEE 802.11 specifications define multiple physical layers and a common MAC layer for wireless local area networks (as shown in Fig. 8.6). Another important aspect of this family is that they use unlicensed spectrum in 2.4, 5, and 60 GHz for operation. The IEEE 802.11 family comprises many technologies that have evolved from direct sequence spread spectrum (DSSS) and complementary code keying (CCK) in the first gen- eration to OFDM waveforms combined with advanced coding and modulation techniques and spatial division multiplexing (SDM) multi-antenna schemes in the latest generations that, depending on channel conditions, can provide data rates in the excess of a few gigabits per second within short distances. Table 8.2 summarizes the key physical layer characteris- tics of IEEE 802.11 air-interface technologies. There are other IEEE 802.11 family mem- bers that each provides an extension to the baseline standard by adding new features such as handover and roaming, mesh networking, QoS, security, regulatory, and measurement for various regions of the world [30]. In IEEE 802.11, the stations and the access point are not synchronized except when they exchange data or control information in the downlink or uplink. An LBT method combined with CSMA/CA is used to g

ain access to the medium and to ensure collision avoidance with other contenders wishing to access the shared medium. The stations either passively scan the beacons transmitted by nearby access points or actively scan neighboring access points by transmitting a probe signal. In IEEE 802.11, there are two options for medium access. The first is a centralized control scheme that is referred to as point coordination function (PCF), and the second is contention-based approach known as DCF. The PCF mode supports time-sensitive traffic Application layer Layers 3-7 data Presentation layer Data link layer MSDU is encapsulated in the MPDU Session layer (layer 2) Transport layer MAC header and trailer are added or removed MPDU = PSDU, the PSDU is encapsulated in the PPDU Network layer Physical layer Preamble and PHY header IEEE 802.2 Logical Link Control (LLC) Data link-layer (layer 1) are added/removed IEEE 802.11 MAC Header 0100010001110111000111 IEEE 802.11 PLCP Header Data is transmitted as bits PhysicaHayer RF/PHY Medium (DSSS, OFDM, etc.) Figure 8.6 OSI and IEEE 802.1 11 protocol relationships [ 30]. Table 8.2: Evolution of IEEE 802.11 air-interface technologies [30]. Frequency Bandwidth Transmission Highest Order Spatial Peak Data Rate Standard (GHz) (MHz) Scheme Modulation Coding Rate Streams (Mbps) 802.11 Convolutional coding with coding rates 802.11b DSSS/CCK 1/2, 2/3, 3/4, and 5/6 802.11a 64QAM 802.11g OFDM, 64QAM DSSS/CCK 802.11n 2.4 and 5 20, 40 OFDM/SDM 64QAM 6.5-600 802.11ac 20, 40, 80, OFDM/SDM/ 256QAM 433 (80 MHz and MU-MIMO one spatial stream) 6933 (160 MHz and 8 spatial streams) 802.11ad 256QAM Up to 6912 802.11ax 2.4 and 5 20, 40, 80, OFDMA/ 1024QAM Convolutional coding and LDPC with 600.4 (80 MHz and SDM/MU- coding rates 1/2, 2/3, 3/4, and 5/6 one spatial stream) 9607.8 (160 MHz and 8 spatial streams) 862 Chapter 8 flows where the access points periodically send beacon frames to communicate network management and identification, which is specific to that Wi-Fi node. Between the sending of these frames,

PCF splits the timeframe into a contention-free period and a contention period. If PCF is enabled on the remote station, it can transmit data during the contention- free polling periods. However, the main reason why this approach has not been widely adopted is because the transmission times are not predictable. The other approach, DCF, relies on the CSMA/CA scheme to send/receive data. Within this scheme, the MAC layer sends instructions for the receiver to look for other stations' transmissions. If it sees none, then it sends its packet after a given interval and waits for an acknowledgment. If one is not received, then it knows its packet was not successfully delivered. The station then waits for a given time interval and checks the channel before retrying to send its data packet. This can be achieved because every packet that is transmitted includes a value indicating the length of time that transmitting station expects to occupy the channel. This is noted by any station that receives the signal, and only when this time has expired other stations consider transmitting. Once the channel appears to be idle, the prospective transmitting station must wait for a period equal to the DCF interframe spacing. If the channel has been active, it must first wait for a time consisting of the DIFS plus a random number of backoff slot times. This is to ensure that, if two stations are waiting to transmit, then they do not trans- mit together, and then repeatedly transmit together. A time known as contention window is used for this purpose. This is a random number of backoff slots. If a transmitter intending to transmit senses that the channel becomes active, it must wait until the channel becomes free. If the channel is still busy, the transmitter continues waiting a random period for the channel to become free, but this time allowing a longer CW (see Fig. 8.7). While the system works well in preventing stations to transmit together, the result of using this access system is that, if the network usage level is high, then th

e time that it takes for data to be success- fully transferred increases. This results in the system appearing to become slower for the users. In view of this, Wi-Fi may not provide a suitable QoS in its current form for systems where real-time data transfer is required. To introduce QoS, a new MAC layer was developed as part IEEE 802.11e. The user traffic is assigned a priority level prior to transmission. There are eight user priority levels. The transmitter prioritizes the data by assigning one of the four access categories. The QoS- enabled MAC layer has combined features from the DCF and PCF schemes into hybrid coordination function (HCF). In this approach, the modified elements of the DCF are termed the enhanced distributed channel access (EDCA), whereas the elements of the PCF are termed the HCF controlled channel access. A new class of interframe space called an arbitration interframe space (AIFS) has been introduced for EDCA. This is chosen such that the higher the priority the message, the shorter the AIFS and associated with this there is also a shorter CW. The transmitter then gains access to the channel in the normal manner, but in view of the shorter AIFS and shorter CW, the higher the chance of it gaining access CWmin(10) CWmin(7) CWmin(0) AIFS[i] Immediate access when medium is free > DIFS/AIFS[i] AIFS[i] Voice random backoff range Video random backoff range Best-effort random backoff range Contention window Contention window Slot time DIFS/AIFS Busy medium Backoff slots Busy medium Next frame Next frame Backoff window Defer access Decrement backoff as long as medium is idle Slot time Select slot and decrement Defer access backoff as long as medium is idle Figure 8.7 Medium access methods in IEEE 802.11 30]. 864 Chapter 8 to the channel. Although statistically a higher priority message will usually gain the chan- nel, this will not always be the case. IEEE 802.11 frame contains a MAC header, a variable length frame body (0-2304 bytes) and 32-bit frame check sequence. The MAC header contains inform

ation related to the type of frame, source and destination addresses, frame control, and sequence control. IEEE 802.11 standard defines three major frame types: (1) management frames which do not carry service data units, (2) control frames to assist delivery of data frames, and (3) some data frames with service data units, some without service data units, and a null frame to inform the access point of client's power save status. IEEE 802.11 physical layer is divided into two sublayers: physical layer convergence proto- col (PLCP) sublayer and physical medium dependent (PMD) sublayer (see Fig. 8.6). The PLCP sublayer receives a frame for transmission from the MAC sublayer and creates the PLCP protocol data unit (PPDU). The PMD sublayer then modulates and transmits the data as bits. The MAC protocol data unit that is delivered to the physical layer is referred to as the PLCP service data unit (PSDU). As part of the processing, the PLCP sublayer adds a preamble and header to the PSDU. When the PLCP layer receives the PSDU from the MAC layer, the appropriate PLCP preamble and header are added to the PSDU to create the PPDU. When transmitting data, the transmitting station provides the receiving station spe- cial synchronization sequences at the beginning of each transmission. The access point periodically broadcasts a special signal called beacon (once every 102.4 ms). When the Wi-Fi communication module is turned on, the device first detects and decodes the beacon signal and establishes physical synchronization with the sender. After establishing syn- chronization, the access point and the device initiate the authentication procedure followed by the association procedure. There are two types of scanning: passive and active scanning. As shown Fig. 8.8, during passive scanning, the device scans and detects the beacon signal from the access point and establishes synchronization based on the beacon signal. In the active scan mode, the device broadcasts a probe request to all access points or a specific one. If there is

any access point that detects the probe request, it sends a probe response to the device. Once the client and the access point go through authentication and association procedure, the client can send or receive data. Unlike cellular standards, IEEE 802.11 does not support dedicated control/traffic channels and it does not have the MAC scheduling functionality in previous generations of Wi-Fi. However, IEEE 802.11a adds an OFDMA scheduling capa- bility and more granular resource allocation scheme. The stations are allowed to transmit at any time as long as the medium is not occupied by transmissions from other stations. To determine whether the medium is free and to overcome the hidden-node problem, the sta- tion transmits a short request to send (RTS) frame containing source address, destination address, and the duration of upcoming data transmission. Other stations located around the transmitting station may receive the RTS burst and check if the RTS is meant for them. If the RTS is meant for a station and the medium is free, the receiving device would transmit Operation in Unlicensed and Shared Spectrum Device Passive scan AP Device Active scan Beacon signal Probe request Decode beacon Detect probe Initial sync Probe response Authentication request Request to send (RTS) Challenge phrase Clear to send (CTS) Encrypted challenge phrase Encrypted challenge phrase Decode beacon Contention Association request Association response Figure 8.8 Scanning and association procedures [30]. a CTS frame containing the duration of the transaction. At this time, other devices in the vicinity of the communicating stations know that the medium will be occupied for certain time duration and set their NAV counters, which is a MAC-level contention control timer accordingly SO that it would not try sensing the medium and transmitting during that period (see Fig. 8.9). When a packet arrives at the MAC layer from higher layers, a sequence number is assigned to it, and if the packet length is larger than a single MAC frame, it is segmen

ted into multi- ple fragments. In this case, a fragment number is assigned to each segment. When a packet is segmented into multiple MAC frames, those fragmented frames are assigned the same sequence number and different values for the fragment number. IEEE 802.11 can transmit a maximum of 2304 bytes of higher layer data. 8.1.4 LTE and NR Solutions for Operation in Unlicensed Spectrum The traditional methods of interworking between cellular networks and Wi-Fi have proved to be cumbersome, with practical limitations in handling the mobility of the IP flows 866 Chapter 8 Backoff following delayed transmission Source Destination Contention window Others NAV (RTS) NAV (CTS) Defer Access Source Frag 0 Frag 1 Frag 2 Destination Others NAV (RTS) NAV (Frag 0) NAV (Frag 1) NAV (CTS) NAV (ACK0) NAV (ACK 1) Defer Access Figure 8.9 Medium access and contention control in IEEE 802.11 [30]. between cellular and Wi-Fi networks. All these schemes have suffered from complexity to define which Wi-Fi networks can be selected for traffic offload, and in which condition a Wi-Fi network should be selected SO that it provides the best performance. For these needs, 3GPP has specified access network discovery and selection function (ANDSF), where an ANDSF server in the operator network controls the connection manager at the device. The ANDSF may have location-dependent network selection policies, defining how and in which priority order to select Wi-Fi networks. It is noted that the UE local operating envi- ronment and user preferences are important. The ANDSF has recently included comparative signal and load thresholds to instruct the UE when it should use the cellular base station and when to use the selected Wi-Fi access points for any given flow. Due to the lack of commercial interest in ANDSF, 3GPP has decided to integrate Wi-Fi into cellular networks by RAN-level features, which can provide scalability as a Wi-Fi access node acts in a simi- lar manner as an eNB does, rather in the packet data network. A Wi-Fi access point may be in

tegrated into an eNB. This type of interworking enables timely accounting of radio aspects in the offloading decisions and transparent integration of the Wi-Fi access with a single node connection to the cellular core network (in this case, the eNB), which not only reduces control signaling but simplifies network management operations. In this section, we focus on 3GPP RAN-level features, for example, LTE-Wi-Fi aggregation (LWA) and LTE- Wi-Fi radio-level integration with IP security tunnel (LWIP) introduced by 3GPP in Rel-13 and extended in Rel-14 [16,22]. In both features, the aggregated LTE link (licensed Operation in Unlicensed and Shared Spectrum LTE/NR Network LTE/NR (KH)) (HH)) ((a)) LTE/NR Licensed LTE/ Network Network NR RCell NB/eNB Wi-FiAP NB/eNB Licensed LTE/ Unlicensed NR SCell Licensed LTE/ Wi-Fi (SCell) Licensed LTE/ LTE/NR SCell Licensed LTE/ NR PCell NR PCell Licensed LTE/ NR SCell NR PCell LTE/NR Carrier Aggregation using Licensed Bands (Link Licensed LTE/NR + Unlicensed Wi-Fi Carrier Aggregation Licensed LTE/NR + Unlicensed LTE/NR Carrier Aggregation Level at MAC Sub-layer) (Link Levelat PDCP Sub-layer) (Link Level at MAC Sub-layer) Figure 8.10 Comparison of various solutions for cellular operation in unlicensed spectrum [18]. operation) provides robust mobility and coverage, whereas the aggregated Wi-Fi link (unli- censed operation) allows routing of high data rate traffic, providing higher user throughput and more efficient use of both types of radio access. Despite their commonalities, LWA and LWIP have significant differences, mainly in the protocol layer at which the aggregation occurs and in the adopted user-plane security mechanisms. Fig. 8.10 summarizes various unlicensed access mechanisms that have been studied in 3GPP. In the following sections, we discuss in more details various schemes that have been used to enable LTE and NR access and operation in unlicensed spectrum. 8.1.4.1 License-Assisted Access Licensed assisted access was introduced in LTE Rel-13. It uses carrier aggregation

in the downlink to combine an LTE carrier in the unlicensed spectrum with LTE carrier(s) in the licensed spectrum (see Fig. 8.12). This aggregation of spectrum provides higher data rates, improved indoor connectivity, and network capacity. This is done by maintaining a persis- tent anchor in the licensed spectrum that carries the control and signaling information and combining it with one or more carriers from the unlicensed spectrum. The LAA has been designed as a single global solution that can adapt to unique regional regulatory require- ments while coexisting with Wi-Fi and other unlicensed-band radios. For regulatory compli- ance and coexistence with other devices operating in the unlicensed band, devices supporting LAA must utilize LBT, which is mandated in Europe and Japan. The LBT used in LAA is similar to the method used by Wi-Fi nodes. A new frame structure (type 3) was defined exclusively for LAA cells, supporting discontinuous time-limited transmissions to ensure that the LTE access nodes do not dominate the channel. It also enables the use of incomplete subframes below 1 ms for more flexible adaptation to transmission opportunities after a successful LBT. The enhanced LAA (eLAA) in Rel-14 extended the downlink-only LAA to the uplink. For LAA operation in some geographical regions (e.g., Japan), 868 Chapter 8 + (TMCOTIT|-1) Tjs (us) Tjs = 34 us ;=4ms Transmit Transmit Is the channel idle during this period? Is the detected energy below a threshold? Figure 8.11 Discontinuous transmission in LAA [22]. discontinuous transmission is part of the specification. An eNB can transmit again on the LAA SCell for a maximum duration of T=ms immediately after sensing the channel to be idle for a sensing interval of Tjs=34 us as shown in Fig. 8.11. The Rel-13 LAA functionality relies on LBT, frame structure type 3 along with discontinu- ous transmission with limited use of channel, use of incomplete subframes, downlink only, no PBCH (PBCH is not transmitted within frame structure type 3), use of Band 46 (5.150-5.92

5 GHz), and discovery reference signals (DRS) to enable radio resource manage- ment. In discontinuous transmission, the first subframe in a burst can last 1 ms (starts at symbol 0) or 0.5 ms (starts at symbol 7), which is signaled to the UE via RRC message subframeStartPosition in the dedicated physical configuration. The last subframe in a burst can last 1 ms or the duration of a special subframe with normal cyclic prefix (3, 6, 9, 10, 11, 12) OFDM symbols, which is signaled to the UE via DCI format 1C scrambled with CC-RNTI [8]. In LAA, the configured set of serving cells for a UE always include at least one SCell oper- ating in the unlicensed spectrum. The LAA SCells act as regular SCells where the eNB and the UE use LBT before performing transmission in the SCell. The combined time of trans- missions by an eNB may not exceed 50 ms in any contiguous 1 second period on an LAA SCell. The LBT type (i.e., type 1 or type 2 uplink channel access) which the UE applies is signaled via uplink grant for physical uplink shared channel (PUSCH) transmission on LAA SCells. For type 1 uplink channel access on autonomous uplink (AUL)7, the eNB signals the channel access priority class for each logical channel and the UE must select the lowest channel access priority class of the logical channel(s) with MAC SDU multiplexed into the MAC PDU. The MAC control elements, except padding buffer status report (BSR), use the highest channel access priority class. For type 2 uplink channel access on AUL, the UE In autonomous uplink, uplink transmissions are allowed without requiring a prior scheduling request or an explicit scheduling grant from the eNB. Operation in Unlicensed and Shared Spectrum ((Q)) Unlicensed SCells User IP packets 12345 Unlicensed band 54321 Licensed PCell licensed Figure 8.12 Illustration of LAA operation concept [18]. may select logical channels corresponding to any channel access priority class for uplink transmission in the subframes signaled by eNB in common downlink control signaling. For uplink LAA operatio

n, the eNB does not schedule the UE more than the minimum number of subframes necessary to transmit the traffic corresponding to the selected channel access prior- ity class or lower than the following: (1) The channel access priority class signaled in uplink grant based on the latest buffer status report and received uplink traffic from the UE if type 1 uplink channel access procedure is signaled to the UE. (2) The channel access priority class used by the eNB based on the downlink traffic, the latest BSR and received uplink traffic from the UE, if type 2 uplink channel access procedure is signaled to the UE [5]. Four channel access priority classes are defined as shown in Table 8.3, which can be used when performing uplink and downlink transmissions on unlicensed carriers. In LTE-based LAA, each channel access priority class, used by user traffic, is associated with the stan- dardized LTE QoS class identifiers (QCIs); i.e., channel access priorities 1,2,3,4 are associ- ated with QCI values 1,3,5,65,66,69,70),(2,7),(4,6,8,9), respectively [5]. A non- standardized QCI (operator defined QCI) should use a suitable channel access priority class associated with the standardized QCIs, which best matches the traffic class. If a downlink transmission on physical downlink shared channel (PDSCH) is intended for which channel access has been obtained using channel access priority class pe {1, 2, 3, 4}, the LTE system must ensure that the transmission duration does not exceed the minimum duration needed to transmit all available buffered traffic; and further the transmission duration does not exceed the maximum channel occupancy time (MCOT)8 8 as defined in Table 8.3 for each channel Maximum MCOT is the maximum continuous transmission time after channel sensing, while Wi-Fi may transmit for much shorter time duration. The maximum continuous transmission time in license-assisted access is limited to 8 or 10 ms regardless of frame size and data rate. 870 Chapter 8 Table 8.3: Channel access priority classes in the downlink/upl

ink [3,4]. Channel Access Direction Priority Class p min (CWp) max (CWp) T'MCOT (ms) Allowed CWP Sizes Downlink {3,7} {7,15} 8 or 10 {15,31,63} 8 or 10 {15,31,63,127,255,511,1023} Uplink {3,7} {7,15} 6 or 10 {15,31,63,127,255,511,1023} 6 or 10 {15,31,63,127,255,511,1023} access priority class. The downlink burst in the above refers to the continuous transmission by LTE after a successful LBT. For uplink PUSCH transmission, there is no additional restriction on the UE regarding the type of traffic that can be carried in the scheduled sub- frames [10]. LTE Rel-15 further enhanced the uplink transmission in an LAA SCell in the unlicensed spectrum by specifying support for multiple uplink starting and ending point in a subframe, and support for AUL transmission, including channel access mechanisms, core and RF requirements for base stations and UEs, and RRM requirements. The key functionalities introduced in LTE Rel-15 include additional starting and ending points for PUSCH trans- missions on an LAA SCell (starting PUSCH transmission at the slot boundary; ending PUSCH transmission after the third symbol or at the slot boundary; and UE-based selection of the starting point for PUSCH transmission at the subframe or slot boundary depending on successful channel access); AUL access where a UE can be RRC-configured with a set of subframes and HARQ processes that it may use for autonomous PUSCH transmissions [8,11]. The AUL operation is activated and released with DCI format 0A (transmission mode [TM1]) or DCI format 4A (TM2). The UE skips an AUL allocation if there is no data in the uplink buffers. The PRB allocation and MCS as well as DMRS cyclic shift and orthogonal cover code are indicated to the UE via AUL activation DCI. The UE informs the eNB along with each AUL transmission of the selected HARQ-process ID, new data indicator, redundancy version, UE ID, PUSCH starting and ending points, as well as whether the UE-acquired channel occupancy time can be shared with the eNB. The eNB may provide the UE with HARQ feedback

for AUL-enabled HARQ processes, transmit power command, and transmit PMI [5,6]. LTE Rel-13 introduced UE RSSI measurements with configurable measurement granularity and time instances of the reports which can be used for the assessment of hidden nodes by an eNB which is located near specific UEs. For example, if UE measurement shows a high RSSI value when the serving cell is inactive due to LBT, it can imply the presence of Operation in Unlicensed and Shared Spectrum 871 hidden nodes and can be considered for channel (re) selection. The DFS is a regulatory requirement for certain frequency bands in some regions, for example, to detect interference from radar systems and to avoid cochannel operation with these systems by selecting a dif- ferent carrier on a relatively slow time scale. The corresponding time scales for DFS are in the order of seconds and can, therefore, be at an even slower time scale than carrier selec- tion. This functionality is an implementation issue and does not impact the specifications. As we mentioned earlier, LBT procedure is defined as an algorithm by which a UE per- forms one or more CCAs prior to transmitting on the channel. It is the LAA equivalent of the DCF and EDCA MAC protocols in Wi-Fi (see Fig. 8.13). 3GPP has specified four categories of LBT as follows: Cat-1: no LBT Cat-2: LBT without random backoff Cat-3: LBT with random backoff and fixed CW Cat-4: LBT with random backoff and variable CW A simple approach to ensure coexistence with Wi-Fi is to make the LAA LBT procedure for both data and discovery reference signal 9 as similar as possible to IEEE 802.11 DCF and EDCA protocols. This was the underlying design principle in Rel-13 LAA LBT mecha- nism for detecting the presence of Wi-Fi which can be summarized as follows: An LAA-enabled node must sense the carrier to be idle for a random number of 9 us CCA slots prior to data transmission. If the energy in a CCA slot is sensed to be above an energy detection (ED) threshold, then the process is suspended and the counter is stopped

. The backoff process is resumed and the counter is decremented once the carrier has been idle for the duration of the deferred period. The energy detection threshold is both channel type and output power related, that is, -72 dBm for 23 dBm PUSCH and 62 dBm for DRS. An important component of LBT design is the choice of ED threshold, which determines the level of sensitivity to declare the existence of ongoing transmissions. 3GPP has stud- ied mechanisms to adapt the ED threshold. An eNB accessing a carrier on which LAA SCell(s) transmission(s) are performed, sets the ED threshold ED threshold to less than or equal to the maximum ED threshold max(EDthreshold) = min(Tmax + 10 dB, Nr), if the absence of any other technology sharing the carrier can be guaranteed on a long-term A downlink transmission by the eNB may include a discovery signal and without physical downlink shared channel on a carrier on which license-assisted access SCell(s) transmission(s) are performed, immediately after sensing the idle channel for a sensing interval Tdrs = 25 us and if the duration of the transmission is less than 1 ms. The parameter Tdrs consists of a duration Tf=16 us immediately followed by one slot duration Tslot = and Tf includes an idle slot duration Tslot at start of Tf. The channel is considered idle for Tdrs, if it is sensed to be idle during the slot durations of Tdrs. Chapter 8 Is the channel idle during this period and the detected energy eNB/gNB needs below a threshold? to transmit Transmission Tslot min(mp) max(mp) min(Tdefer) max(Tdefer) Backoff Channel busy LAA downlink burst eCCA slot time PDCCH Pcell subframe boundary PDSCH PDSCH PDSCH PDSCH Subframe = 1ms Pcell subframe boundary Starting downlink partial subframe Ending downlink partial subframe Downlink transmission Uplink transmission Subframe Reservation boundary boundary signal MCOT = 8 ms Defer duration Channel Transmission Initial partial subframe Ending partial subframe 0.5 ms (12,11,10, 9, 6, 3 OFDM symbols) Reservation Slot duration Subframe signal boun

dary Figure 8.13 Frame structure type 3 and LBT mechanism in LAA 22]. Operation in Unlicensed and Shared Spectrum 873 basis, where Nr is maximum ED threshold defined by regulatory requirements in dBm when such requirements are defined; otherwise [4]: -72 log10(BW MHz/20 MHz) dBm, ED threshold max (PH + 10log10(BW MHz/20MHz) PTX) where PH is a reference power equal to 23 dBm, TA = 10 dB for transmission(s) includ- ing PDSCH or TA = 5 dB for transmissions including discovery reference signal trans- mission(s) and excluding PDSCH, PTX is the configured maximum transmit power for the carrier in dBm, and it is given by Tmax = dBm/MHz + 10 log10(BW), where BW is the channel bandwidth in MHz. In a nutshell, the energy detection threshold can be increased if the bandwidth increases and/or the transmit power decreases. The detected energy level needs to be below this threshold for a certain period of time with a slot duration Tslot us and defer time Tdefer Tf + Mp, where Tf = 16(us)mp is based on the channel access priority class and is, therefore, traffic-type dependent and has a duration of at least one slot. The defer time is required to specifically protect Wi-Fi ACK/NACK transmission between the access points and the clients. As a result, the channel needs to be idle for an initial CCA period of 34 us and a maximum wait time of 88 us before an LAA-capable eNB can start its transmission. If the channel is sensed to be clear, the transmitter can only transmit for a limited amount of time defined as the maximum channel occupancy time T'MCOT: If the channel is sensed to be occupied dur- ing that time or after a successful transmission, the enhanced CCA period is started by generating a random number that is within the contention window. Because there are different traffic types (VoIP, video, background traffic, etc.), different channel priority access classes have been defined with different CW sizes and T'MCOT (see Table 8.3). If the most recent downlink transmission burst showed 80% or more decoding errors, as reported

via HARQ feedback (NACKs) from UEs, then the CW is doubled for the next LBT (see Fig. 8.14). Once downlink transmission is complete, a new random backoff is chosen and used with the next transmission. A single, short CCA period of 25 us can be used to transmit control information without accompanying data such as DRS. This is aligned with the CCA duration used for Wi-Fi beacon frames. Rel-13 LTE defined an LAA equivalent to the four Wi-Fi priority classes in the form of four sets of minimum and maximum CW sizes, maximum channel occupancy times, and deferred period CCA slots. An LAA/Wi-Fi coexistence study suggests that the coexistence performance is more sensi- tive to factors that affect the channel occupancy (e.g., control signals) rather than to the 874 Chapter 8 Downlink SCell Update CW based on HARQ feedback and draw counter End of burst 0<N<CW update CW Defer Channel Channel Freeze period Post TX Backoff Short Uplink PCell report report Figure 8.14 An example of LAA downlink transmission with LBT CW updates based on HARQ ACK/NACK feedback. The UE provides HARQ feedback and CSI reports on the licensed carrier [17,23]. choice of parameters in the LBT CCA and backoff algorithms. Consequently, the coexis- tence is highly affected by the behavior of the upper-layer protocols; bursty traffic pattern; HARQ-based CW slot update; and either CTS-to-self or support for lower Wi-Fi energy detection thresholds, which seems to be a fundamental feature to be supported by LAA to allow coexistence with Wi-Fi and protect the LAA performance in the presence of hidden nodes [12]. To minimize interference to Wi-Fi transmissions, an LTE system can mute its operation in almost blank subframes (ABS). These subframes are called almost blank because LTE can still transmit some broadcast signals, control signals, and synchronization signals over these subframes. However, these signals only use a small fraction of system resources in the time/ frequency domain and reduced power. Therefore, cochannel interference is much less during t

he ABS transmission periods. The LTE-based LAA defines methods for UEs in the LAA SCell coverage to provide radio resource management measurements and reporting. The Rel-12 discovery reference signals, originally designed for small cells, are being used with certain modifications in LTE-based LAA SCells. As shown in Fig. 8.15, the LTE-based LAA DRS can be transmitted within a periodically occurring time window called the DRS measurement timing configuration occasion that has a duration of 6 ms and a configurable period of 40, 80, or 160 ms. The radio frame and subframe in which DRS can be transmitted depends on the RRC parameters (dmtcOffset, dmtcPeriodicity) that are signaled to the UE. The LTE-based LAA DRS can be transmitted following a single idle observation interval of at least 25 us. To compensate for Operation in Unlicensed and Shared Spectrum Silent subframes (no CRS, PSS, SSS, PDSCH) Discovery signal sent with a 40,80, or 160 ms period in SF5 (CRS, PSS, SSS, PDCCH,/PDSCH for SIB1) Data transmitted to UE Unlicensed SCell Licensed PCell Frame in Frame n+4 Regular subframes with CRS (even when no PDSCH) DMTC periodicity (40, 80,160 ms) SFNmod T = DMTCoffset/10 T = DMTCperiodicity/10 DMTC occasion (6 ms) Subframe = DMTCoffsetmod10 DMTC offset LAA SCell 1 LAA SCell 2 Primary synchronization signal (PSS) LAA DRS occupies the first 12 OFDM symbols within a subframe Secondary synchronization signal (SSS) Cell specific reference signal (CRS) Figure 8.15 Structure and timing of discovery signals in LAA SCell [14,20]. potential DRS transmission blocking due to LBT and increase the probability of successful of DRS transmission, the network can attempt DRS transmission in any subframe within the DMTC occasion [26]. LAA operation on multiple unlicensed carriers is a key requirement for maximizing throughput. IEEE 802.11ac/ax supports transmission bandwidths of up to 160 MHz, which would span eight contiguous 20 MHz unlicensed channels in the 5 GHz band. The LAA 876 Chapter 8 multi-carrier transmission on multiple unl

icensed SCells adheres to the principle of fair channel utilization while identifying improved transmission opportunities across available spectrum. Rel-13 LAA supports two methods for identifying and utilizing secondary chan- nels: (1) prior to transmission, a single random backoff is completed on any carrier along with CCA on other channels; (2) multiple SCells must individually perform random backoff before transmitting simultaneously. The single and multi-carrier LBT schemes are illustrated and compared in Fig. 8.16 for a scenario in which three LAA SCells are assigned a common random backoff counter. The performance of multi-carrier LBT over 80 MHz has been eval- uated and shown in Fig. 8.17. The overall system performance in terms of mean user throughput as a function of served traffic per access point per operator show that, from the coexistence point of view and the impact on the non-replaced Wi-Fi network, both classes of multi-channel LAA LBT schemes are viable and can increase the performance of a multi-carrier Wi-Fi network compared to the case when it is coexisting with another Wi-Fi network [17,23]. Backoff complete, quick End of burst check on other SCells Post-TX backoff SCell 1 SCell 2 Interferer SCell 3 Interferer End of burst Backoff complete, wait until wait Ilimit Post-TX backoff Wait limit SCell 1 SCell 2 Interferer SCell 3 Interferer Figure 8.16 Comparison of LAA multi-carrier LBT access schemes with single and multiple random backoff channels (17,23,31,32). Operation in Unlicensed and Shared Spectrum 877 Downlink throughtput Uplink throughtput Operator Wi-Fi uplink Operator B Wi-Fi uplink (LAA Alt 1) Operator B Wi-Fi uplink (LAA Alt 2) Operator Wi-Fi downlink Operator Wi-Fi downlink Operator LAA Alt Operator Wi-Fi downlink Operator LAA Alt 2 Operator B Wi-Fi downlink Total served traffic per operator per AP (Mbps) Total served traffic per operator per AP (Mbps) Figure 8.17 Mean user throughput versus served traffic per access point per operator for the indoor multi- carrier deployment scen

ario with FTP traffic using up to 80 MHz transmission bandwidth 17,23]. As mentioned earlier, LBT is a mechanism used by devices operating in the unlicensed bands to determine the presence of other signals in the channel prior to transmission and to avoid collisions with other transmissions. This protocol allows several users and different technologies to use the same channel without pre-coordination. In LAA, the LBT procedure is initialized when the SCell has data to transmit. Prior to that, the cell is completely turned off to all transmissions, including cell reference signals. Then, the LAA cell will initiate CCA to determine if the channel is idle. If there are no signals detected in the channel, then the transmission can proceed. This procedure is illustrated in Fig. 8.18. If the channel is not idle, the device performs a slotted random backoff procedure, in which a random number of slots are withdrawn from the contention window. The contention window is increased expo- nentially with the occurrence of collisions and is reset to the minimum value once the trans- mission succeeds. Given the random nature of the backoff procedure, different devices will have different backoff intervals, improving channel adaptation [14,20]. 8.1.4.2 LTE-Wi-Fi Aggregation An alternative approach to deploying LTE in the unlicensed spectrum, which was more acceptable to the Wi-Fi industry, was through the aggregation of LTE and Wi-Fi carriers. This solution was meant to enhance LTE performance by partly offloading traffic to an available Wi-Fi link. In this scheme, an LTE payload is split at the PDCP level and some traffic is tunneled over the Wi-Fi link while the remaining traffic is sent over the native LTE connection. The LWA approach uses Wi-Fi access points to augment LTE RAN by encapsulating LTE data in IEEE 802.11 MAC frames, such that it appears as a Wi-Fi frame 878 Chapter 8 Frame-based equipment (FBE) LBT Fixed frame period Fixed frame period Fixed frame period CCA failure CCA success CCA failure CCA success CCA period

is the significant difference compared to LTE-unlicensed and LAA schemes. In the case of LWA solution, the aggregation takes place at the radio link level. The Wi-Fi access points can use the LTE core network functions (e.g., authentication, security, etc.). It will be dis- cussed in Section 8.2 that the LWA architecture consists of LWA eNB, LWA-aware Wi-Fi access points, and LWA UEs. The LWA eNB and Wi-Fi AP can be co-located or non-co- located. If non-co-located, data is delivered through IP tunnel between two systems. The LWA eNB performs scheduling of PDCP packets and transmits some of the IP packets over the LTE air-interface and others over the Wi-Fi after encapsulating them in Wi-Fi MAC frames. All packets, received over either LTE or Wi-Fi, are then aggregated at the PDCP sublayer of the LWA UE. The Wi-Fi APs are connected to LWA eNBs and report informa- tion on channel conditions to the LWA eNB. The LWA eNB then determines whether to use the Wi-Fi link or not. The LWA eNB can improve LTE performance by managing radio resources in real-time according to the RF and load conditions of both LTE and Wi-Fi. A Wi-Fi AP can work as a native Wi-Fi AP while not serving the LWA purpose. This Operation in Unlicensed and Shared Spectrum 879 ((Q)) IEEE 802.11 MAC IEEE 802.11 PHY IP packets Wi-Fi carrier User IP packets ethernet Wi-Fi 12345 Access point 54321 Licensed PCell Figure 8.19 Illustration of LWA principle of operation [18]. eliminates potential coexistence issues in the unlicensed bands. However, as LTE data must be split at eNB and then aggregated at UE, the involved nodes such as eNB, Wi-Fi AP, and UE must be LWA-enabled. The LWA architecture, protocol, and operations will have to be defined as well. Fig. 8.19 shows the high-level architecture and operation of LWA scheme [18]. 8.1.4.3 LTE Unlicensed and MulteFire In 2014, a group of companies formed the LTE-U Forum to develop technical specifications based on LTE Rel-12 in the 5 GHz band. LTE-U Forum initially focused on non-LBT mar- kets without modifying

the LTE physical layer and MAC functionalities. The deployment scenarios included supplemental LTE downlink operating in an unlicensed spectrum com- prising multiple contiguous 20 MHz bands. The LTE network with an uplink and downlink anchored on the licensed spectrum (up to 20 MHz) for delivering essential control signaling and synchronization related to radio resource management and connection/mobility provides reliability, mobility, and coverage, whereas the unlicensed spectrum was exclusively used for increasing the downlink data rates. LTE-U was designed to comply with the US regulations and targets other regions of the world with no LBT requirements. However, it does not necessarily meet the needs of other regions which do require LBT feature such as Europe and Japan. While LTE-U and LAA are both LTE-based solutions for operation in the unlicensed band, they differ in the way they meet the regulatory requirements. LTE-U is well suited for regions that do not require LBT and address coexistence issues with Wi-Fi using an algorithm called carrier sensing adaptive transmission (CSAT) in the LTE-U eNB, which is based on the use of duty cycle that adapts the duration of the ON period by sensing the average power in the channel as shown in Fig. 8.20. LTE-U also limits the duration of the ON period to 20 ms SO that low 880 Chapter 8 LTE in OFF period senses channel CCA success utilization by Wi-Fi Channel access by Wi-Fi when LTE is OFF LAA frame failure LTE on TE off LTE on LE off CCA failure CCA success Duty cycle Wi-Fi Wi-Fi Random backoff Carrier sensing adaptive transmission (CSAT) Licensed assisted access (LAA) Figure 8.20 Comparison of CSAT and LBT mechanisms [17,23]. latency applications can coexist. On the other hand, LAA is a global solution and includes support for LBT and a more flexible frame structure to better adapt to channel availability. LAA is generally preferred by operators as it is the only method that provides full cellular capabilities, whereas LWA or LWIP are less frequently used because t

hey require either a Wi-Fi network or partnership with a Wi-Fi operator. In an ultra-dense deployment of Wi-Fi and LTE-U small cells, there is a possibility that no clean channel can be found. In such cases, LTE-U can share the channel with the neighbor- ing Wi-Fi or another LTE-U system by using the CSAT algorithm. Typical cochannel coex- istence techniques in unlicensed bands such as LBT and CSMA, used by Wi-Fi systems, are based on the contention-based access. In these techniques, transmitters are expected to sense the medium to ensure that there is no activity prior to transmission. The goal of these algo- rithms is to provide coexistence across different technologies in a time-division multiplex- ing (TDM) manner. LTE-U in unlicensed spectrum uses CSAT scheme that relies on the same concept of TDM coexistence and medium sensing. In CSAT, the small cell senses the medium for longer (than LBT and CSMA) duration (up to 200 ms), and according to the observed medium activities, the algorithm proportionally switches off LTE transmission. In particular, CSAT defines a time cycle where the small cell transmits in a fraction of the cycle and remains off in the remaining duration. The duty cycle of transmission relative to OFF interval is determined by the sensed medium activity of other technologies. The CSAT is conceptually similar to CSMA except that it has longer latency, an impact that is miti- gated by avoiding channels where Wi-Fi APs use for discovery signals and QoS-enforced traffic. The LTE-U, which is deployed on the SCell, is periodically activated and deacti- vated using LTE MAC control elements. The procedures and timeline are chosen to ensure compatibility with legacy LTE UE behavior. During the LTE-U OFF period, the channel is released to neighboring Wi-Fi APs, which can resume normal transmissions. The small cell will measure Wi-Fi medium utilization during the LTE-U OFF period and adaptively adjust Operation in Unlicensed and Shared Spectrum ON/OFF duty cycle. The adaptive cycle can effectively accom

modate the activation/de-acti- vation procedures while controlling the data transmission delay. Because the anchor carrier in license band is always available, the supplemental downlink carrier in the unlicensed band can be used opportunistically. When the downlink traffic of the small cell exceeds a certain threshold and there are active users within the unlicensed band coverage area, the supplemen- tary downlink (SDL) carrier can be turned on for offloading the traffic. When the traffic load can be managed by the primary carrier or there is no user within the unlicensed band cover- age area, the SDL carrier is turned off. Opportunistic SDL mitigates the interference from LTE continuous cell-specific reference signal transmission in an unlicensed channel. MulteFire was proposed by the MulteFire Alliance in late 2015 as a standalone version of LTE for small cells, which operates in the unlicensed spectrum and can provide service to users with or without a universal subscriber identity module card (i.e., an operator subscrip- tion). It combines the advantages of LTE technology and the simplicity of Wi-Fi deploy- ments and can be deployed either by traditional mobile operators or neutral hosts. It specifies two different architectures: (1) a PLMN access mode, which allows mobile net- work operators to extend their coverage into the unlicensed band, especially in the case where a licensed spectrum is not available at certain locations; and (2) a neutral host net- work access mode, which is similar to Wi-Fi, a self-contained network deployment that pro- vides access to the Internet. Because of the nature of transmission in the unlicensed band and the need to adhere to the LBT requirements, MulteFire has introduced some modifica- tions to the radio interface of LTE [16]. 8.1.4.4 NR-Unlicensed The NR-U study item in Rel-16 is identifying and evaluating solutions for the NR when operating in the unlicensed spectrum. 3GPP will further investigate the coexistence with NR-based systems in the licensed and unlicensed bands

as well as with LTE-based LAA and other incumbent RATs to ensure compliance with regulatory requirements. 3GPP Rel- 15 NR introduced a flexible frame structure with various subcarrier spacing options and slot formats. It further introduced mini-slot configuration, whereby PDSCH resource allocation can start from almost any symbol in a slot. The NR-U is expected to support both baseline NR Type A (slot-based) and Type B (non-slot-based) resource allocation schemes. Even with restricted mini-slot length of 2, 4, and 7 symbols for downlink, the PDSCH transmis- sion (without gap) can still start at almost any symbol position, if we allow multiple mini- slots to be allocated within a slot. In the uplink, however, the mini-slot length and starting symbol are not restricted, thus uplink transmission in NR-U (without gap) can start at any symbol position. For sub-7 GHz unlicensed bands, 15 kHz SCS with normal cyclic prefix that was supported in LTE-based LAA and the same SCS must be supported in NR-U to achieve similar cover- age when the component carrier bandwidth is 20 MHz. On the other hand, if a larger SCS is 882 Chapter 8 used, the OFDM symbol duration will be shorter, which increases the channel access oppor- tunities, reduces latency, and enhances resource utilization. A larger SCS can facilitate sin- gle wideband-carrier operation. In carrier aggregation scenarios (similar to LTE-based LAA), the device performs LBT per component carrier and then transmits on any available component carrier. Operating a single carrier allows reducing the control overhead and avoiding the use of inter-carrier guard bands. For the same system bandwidth, a larger SCS requires a smaller FFT size, hence reducing the implementation complexity. The studies suggest that an NR system with SCS of 60 kHz operating as a wideband carrier obtains about 60% average downlink user throughput gain and 80% average uplink user throughput gain over LTE-based LAA system for the same system bandwidth [14,20]. The performance gain comes from the finer g

ranularity of channel occupation in the time domain. In addition, the processing delay for the short symbol demodulation is reduced, SO retransmission for short symbols in the same MCOT becomes possible and HARQ combining gain can be achieved with low latency. Furthermore, the overhead of guard band with 60 kHz SCS wideband carrier is also smaller than that of LTE-based LAA. However, in carrier aggrega- tion scenarios, the performance gain of 60 kHz SCS relative to LTE-based LAA diminishes because of the spectrum efficiency loss from guard bands in 60 kHz SCS and 20 MHz band- width. The occupied channel bandwidth of an SS/PBCH block with subcarrier spacing 15 and 30 kHz is 3.6 and 7.2 MHz, respectively. An SS/PBCH block with subcarrier spacing 15 kHz in 5 GHz unlicensed band cannot meet the OCB requirements because system/nomi- nal bandwidth in most cases would be greater than 5 MHz. The NR-U should support simi- lar SCS in 6-7 GHz unlicensed bands. To support 60 kHz SCS for all NR-U signaling in operation below 6 GHz, some modifications in Rel-15 NR are required because NR does not support SS/PBCH block transmission using 60 kHz. In addition, the candidate SS/PBCH positions for a half frame are not defined for 60 kHz SCS. Note that the use of 60 kHz SCS for PRACH transmission in NR is also restricted to frequencies above 6 GHz. For outdoor deployments, even if the output power is limited, considerable delay spread can still be observed. For 60 kHz SCS with normal CP, the cyclic prefix is 1.17 us corresponding to a range of 175.5 m, which is very typical in outdoor environments. The cyclic prefix of 60 kHz SCS with extended CP is 4.17 us corresponding to a range of 625.5 m, which would be sufficient for most outdoor small cell deployments. In NR, the SCS for each channel or signal is configured through the BWP information ele- ment (IE) configuration, and the subcarrier spacing in the BWP IE is used for all channels and reference signals unless explicitly configured. The reference SCS of the slot structure config

uration is included in SIB1. Furthermore, the SCS for SIB1, Msg.2/Msg.4 for initial access, and broadcast SI-messages is included in the MIB. For NR-U, these SCS configura- tion schemes can be reused directly. In licensed bands, NR supports dynamic and semi- statically configured slot configuration. The semi-static DL/UL assignment is done via cell- specific and UE-specific RRC configuration. With the cell-specific configuration, the UE is Operation in Unlicensed and Shared Spectrum Fixed frame period Fixed frame period LAA frame LAA frame Idle period Idle period Fixed frame period Fixed frame period Content Content Content Content Slot 0 Slot M-1 Slot 0 Slot M-1 M contention M contention slots slots Figure 8.21 Illustration of timing relation for FBE [24,25]. allocated flexible resources that can be assigned to downlink or uplink using UE-specific configuration or dynamic signaling. The UE-specific RRC signaling or dynamic signaling includes per-slot-basis indication that can override the unknown allocation in the cell- specific configuration. Resources in the slot configuration that are without downlink/uplink indication are considered flexible resources. Frame-based equipment (FBE) 10 is a channel access mechanism wherein the transmit/ receive frame structure has a periodic timing with a periodicity known as the fixed frame period (FFP), which is between 1 and 10 ms. The transmitting device performs a single-shot LBT before starting transmission in a channel at the beginning of an FFP, and the channel occupancy time (COT) associated with a successful LBT (see the example shown in Fig. 8.21). In addition to the load-based equipment (LBE) operation mode, NR-U can also support the FBE operation mode for use cases in which other LBE-based networks (e.g., Wi-Fi, LTE-based LAA) are excluded such as in a factory or private NR-U network. In comparison to the LBE mode, the FBE operation mode can achieve higher spectrum utiliza- tion in such scenarios considering much simpler LBT process. The necessity of signaling to i

ndicate FFP and channel occupancy time to the UE is being studied. FBE is a scheme where the transmit/receive frame structure is not directly demand-driven but has fixed tim- ing as opposed to LBE where the transmit/receive frame structure is not fixed in time but demand-driven. 884 Chapter 8 Dynamic and semi-statically configured downlink/uplink configuration can be used in NR- U. For example, some semi-static DL/UL assignments can be used to indicate downlink slots to the UE to receive the SS/PBCH blocks, and other slots can be assigned to the UE to transmit PRACH or granted uplink transmission. If gNB performs LBT successfully on the semi-statically configured downlink slot, it can dynamically override the flexible resource (s) through DCI format 2_0. Furthermore, the use of a flexible frame structure in NR-U can take advantage of MCOT. Multiple DL-to-UL switching points in the MCOT can be permit- ted to reduce the scheduling and feedback delay. However, more DL-to-UL switching points create more gaps to perform LBT upon each change of transmission direction which may result in resource wastage or channel loss during this time. Therefore, there is a tradeoff between fast switching and overhead for potential use cases. Because the MCOT duration is different for various channel access priorities and regional regulations, different maximum number of DL-to-UL switching points for each MCOT duration should be defined. In LTE- based LAA, the eNB can only start transmission at slot boundaries (symbol 0 or 7) with 15 kHz subcarrier spacing. The reservation signal is transmitted before the downlink trans- mission start position, that is, the slot boundary, to reserve the carrier and prevent other devices from occupying the frequency band, which is considered an overhead in LTE-based LAA system. Therefore, the overhead of reservation signal should be minimized. In NR-U, different slot types including slot and mini-slot (non-slot-based scheduling) can be utilized to minimize the reservation signal length. The gNB or UE s

hould be able to send DL/UL transmissions as soon as possible whenever LBT/CCA procedure is completed, resulting in different slot duration. As illustrated in Fig. 8.22, the gNB can choose the type of the first slot, that is, mini-slot based on LBT/CCA success location and slot boundary, and then select the type of the following slots with a larger size. The type of the last slot(s) is chosen according to the remaining time within the MCOT. In the example shown in Fig. 8.22, the remainder of the MCOT is not enough for a normal slot, thus two mini-slots are allocated [17,28]. In the downlink, because a CORESET can be configured to start at any symbol, gNB can start downlink transmission from the nearest symbol after a successful LBT such that the reservation signal overhead can be minimized. However, this will cause increased blind detection complexity for the UE. Therefore, NR-U would rely on an efficient mechanism for the UE to detect the beginning of a downlink transmission. In LTE-based LAA, the ini- tial signal was discussed to facilitate the detection of downlink burst. The NR supports non- slot-based PUSCH scheduling in the uplink and the start symbol can be dynamically indi- cated by DCI. Therefore, the UE can transmit PUSCH starting from the indicated symbol. However, this position may not align with the nearest symbol where LBT succeeds. Consequently, more reservation signals may be needed for the uplink. Because NR already supports multiple start positions and durations, it can provide sufficient transmission flexi- bility for NR-U [10,33]. Operation in Unlicensed and Shared Spectrum Minimization of initial signal Minislot minislots CCA success Slot with 60 kHz SCS Slot with 15 kHz SCS 0.25 ms 7 OFDM symbol minislot with 60 kHz SCS 0.125 ms Slot with 15 kHz SC Slot with 15 kHz SCS 2 OFDM symbol minislot with 15 kHz SCS Figure 8.22 Multiple slot durations in an MCOT with different SCS [17,23]. In NR licensed operation, there are three patterns for multiplexing the SS/PBCH block and the RMSI transmissions

as shown in Fig. 8.23. The number of PRBs for SS/PBCH blocks is 20 whereas the minimum number of PRBs for RMSI is 24. Using a 20 MHz bandwidth, the OCB requirement can be met if the numerology of SS/PBCH block and RMSI is set to 60 kHz. In that case, Pattern 1 can be considered for NR-U in the sub- 7 GHz bands. The TDM method in Pattern 1 allows the RMSI to provide consecutive transmissions between SS/PBCH blocks over the time span of DRS. In contrast, if 30 kHz SCS is selected, the bandwidth of SS/PBCH blocks would only be about 40% of a 20 MHz. In this case, a frequency division multiplex pattern such as Pattern 2 or 3 is preferred in terms of OCB requirement. When RMSI is multiplexed with SS/PBCH blocks using Pattern 2 or 3, the RMSI may not completely fill the gaps between the SS/ PBCH block transmissions. 886 Chapter 8 Pattern 1 Pattern 2 Pattern 3 Figure 8.23 Multiplexing patterns of SS/PBCH and RMSI transmission [33]. In NR-U, because a channel occupancy time can start at any time, the UE would need to perform frequent PDCCH monitoring, even when there is no downlink transmission intended for it. To solve this issue, a wake-up signal may be designed for NR-U to wake up the UE only when necessary. The transmission of a wake-up signal is also subject to LBT. To reduce the LBT overhead, the wake-up signal can be embedded in an existing signal such as DRS that is transmitted periodically, resulting in significant UE power saving while maintaining low signaling overhead. Similar to LTE, NR supports SRS-based frequency-selective scheduling and periodic/aperi- odic SRS transmission. NR also supports SRS-based downlink beamforming and semi- persistent SRS transmission, which are not supported in LTE. In NR-U, the transmission time of the SRS resource as well as the configuration of SRS resource and resource map- ping adapt to LBT and channel availability. Due to the uncertainty of channel availability, periodic SRS would not be feasible. The aperiodic SRS which is triggered by DCI has more flexibility for NR-U op

eration. The SRS transmission should also meet the OCB require- ments; thus NR-U can only support wideband and BWP/subband SRS transmission. In LTE-based LAA revisions, downlink and uplink transmissions start and end at certain points. More specifically, using frame structure type 3, a downlink or uplink transmission can start at slot boundaries. Because LBT may end at any time which is not necessarily aligned with the symbol boundaries, the interval between the end of LBT and the start of a PDCCH/PUSCH transmission may be wasted. The partial ending subframe reduces over- head by defining more ending points, that is, OFDM symbol positions 2, 5, 8, 9, 10, 11, and 13 for the downlink and 6, 12, and 13 for the uplink can be used as partial subframe ending Operation in Unlicensed and Shared Spectrum Starting position 1: symbol 1 Starting position 2: 012345678910111213 symbol 3 Starting position 3: 012345678910 symbol 5 Starting position 4: 012345678910111 symbol 7 Starting position 5: symbol 8 Starting position 6: 0123456789 symbol 10 Starting position 7: 012345678910 symbol 12 2 Symbol minislot 4 Symbol minislot 7 Symbol minislot Figure 8.24 Flexible starting point with nonuniform minislot patterns [33]. points. The NR-U provides more frequent start/end points than LTE-based LAA allowing more flexible transmission as shown in Fig. 8.24. As we mentioned earlier, LBT is a spectrum sharing mechanism that can operate across dif- ferent RATs. However, it suffers from the hidden node and the exposed node problems. These issues arise when the coverage of the nodes in a network are different. The problem worsens when the sensing and transmit coverages are different. This may occur when an omnidirectional antenna pattern is used for carrier sensing whereas a directional antenna (or array) pattern is used for transmission, resulting in a higher node exposure likelihood. In the example depicted in Fig. 8.25, device A senses the carrier using omnidirectional antenna pattern and overhears the transmission from device C to D. Dev

ice A subsequently refrains from transmitting to its targeted receiving device B, which is unnecessary because the trans- mission from A to B using highly directional beams does not interfere with the transmission from C to D. If the direction of the communication is known, directional carrier sensing may be useful; however, this would create a different problem as shown in Fig. 8.25. The effects of directionality on carrier sensing and transmission are studied in NR-U to ensure optimal system operation and performance [34]. 888 Chapter 8 ((9)) Figure 8.25 (A) Omnidirectional carrier sensing. (B) Two simultaneous directional links [34]. The NR supports CSI-RS with 1, 2, and 4 OFDM symbols and up to 32 antenna ports. It further supports periodic, aperiodic, and semi-persistent CSI-RS transmission patterns. The CSI-RS can be used for RRM measurement in the connected mode for mobility, the RSRP measurement for beam management, and CSI measurement for scheduling. The CSI-RS can be transmitted in conjunction with the SS/PBCH block and/or CORESET/RMSI using time or frequency multiplexing following an LBT process. If frequency-domain multiplexing is used, a CSI-RS resource is configured on the resources that are not occupied by SS/PBCH block and CORESET. Due to the uncertainty of channel availability for transmission of CSI-RS and the corresponding measurement report, the periodic CSI measurement and report is not feasible. Furthermore, the interference condition in the unlicensed spectrum may fluctuate severely and is unpredictable due to the larger delays between the measure- ment and the report as well as beamforming-based transmission. In LTE-based eLAA, block interlaced frequency division multiple access (B-IFDMA) was introduced in the uplink transmission to ensure compliance with OCB and maximum PSD requirements while maintaining a transmit power level that can support the desired cell COV- erage. In NR-U, given similar regulatory requirements, the B-IFDMA serves as the baseline design for the uplink transmission.

NR supports a large number of channel bandwidth com- binations and subcarrier spacings, which makes the realization of a unified B-IFDMA design very challenging. A typical B-IFDMA design can be characterized by three para- meters: the number of subcarriers per block Nsc, number of blocks per interlaces L, and number of interlaces per symbol N as illustrated in Fig. 8.26. In LTE-based eLAA, NRB =100 and with RB-based interlaced design, Nsc = 12,L = 10, and N = 10. The set of subcarriers Sn allocated for a specific interlace n can be represented as Sn = + m|0 m <Nsc, ISI<|NRB/N Note that it is not always possible to Operation in Unlicensed and Shared Spectrum NscxN Unused RBs Interlace 0 NSCXN Interlace 0 Figure 8.26 B-IFDMA design parameters and various design options [35]. divide NRB by N. As an example, in NR, for a channel bandwidth of 20 MHz and subcarrier spacing of 15 kHz, NRB = 106, and if the LTE-based eLAA parameters are maintained, then six RBs will not be used by any interlace. To avoid resource wastage, one can assign the remaining blocks to some of the interlaces. For example, if we assign R remaining blocks to the first R interlaces, then the set of subcarriers S' allocated for a specific interlace n can be represented as S'n = {Nsc m00m<Nsc, OSI<(NRB - n - 1)/N}. In this case, the number of blocks per interlace is a function of n as shown in Fig. 8.26. To meet the OCB requirement, one can design interlaces such that all interlaces occupy channel band- widths larger than what the minimum OCB requires. In that case, the interlace parameters Nsc,- L, and N need to be carefully chosen. In particular, assuming nominal channel band- width B and subcarrier spacing Af, the minimum normalized OCB among all interlaces is given by Bmin Nsc 12NRB/(NscN) - 1]N + 1) /B. As an example, in LTE-based eLAA, = 15 kHz and B=20 MHz, thus Bmin 0.819, which indicates that eLAA inter- lace scheme can only satisfy 80% of the minimum OCB requirements. If we consider an NR numerology with Af = 60 kHz, B = 20 MHz, and NRB = 2

4 and if we adopt RB-based interlace Nsc = 12 to satisfy 80% of the minimum OCB requirements, the maximum number of permissible interlaces would be N = 2. For some NR bandwidth and SCS combinations, due to regulatory requirements and design constraints, the number of interlaces available 890 Chapter 8 Interlace 0 for UE 0; PTX=P Interlace 1 for UE 1; PTX P 24 PRBs Interlace 0 Interlace 1 Interlace 0 for UE 0; PTx = 2P Interlace 2 for UE 1; PTX = 2P Interlace 2 Interlace 3 24 PRBs 20 PRBs Figure 8.27 RE-group-level interlace design [35]. (A) Uniform RB level interlace. (B) Uniform RE group level interlace. (C) Nonuniform RE-group level interlace. per symbol can be very limited; in such cases, RE-group-level interlace schemes may be utilized (see Fig. 8.27). In Rel-15 NR, five PUCCH formats are defined, where PUCCH formats 0 and 2 are short PUCCH formats which occupy up to two OFDM symbols; PUCCH formats 1, 3, and 4 are long PUCCH formats which occupy between 4 and 14 OFDM symbols. For PUCCH for- mats 0 and 1, the number of UCI bits is 1 or 2, whereas, for PUCCH formats 2, 3, and 4, the UCI bits can be very large. For PUCCH formats 2 and 3, the maximum number of occu- pied PRBs is 16, whereas PUCCH formats 0, 1, and 4 use only one PRB. Considering the use of LBT and the regulatory requirements on OCB, NR-U makes special considerations about UCI payload size and transmission efficiency as well as UE multiplexing capacity. To overcome the uncertainty of LBT and to improve the spectral efficiency in the unli- censed band, the network may schedule uplink transmission of multiple UEs within the same channel occupancy time. However, due to the OCB requirement, the number of inter- laces per symbol may be limited. As described earlier, a larger SCS value leads to a more restrictive interlace design. Operation in Unlicensed and Shared Spectrum 8.2 Network Architecture and Protocol Aspects In the previous section, we described the prominent methods that cellular industry has con- sidered for access to the unlicensed spectru

m, among which LWA is the only scheme that has network architecture implications that will be discussed in this section. The LWA has been supported in 3GPP specs since Rel-13, wherein a UE in RRC_CONNECTED state is configured by the eNB to utilize radio resources of LTE and Wi-Fi. Two scenarios are sup- ported depending on the backhaul connection between LTE and Wi-Fi namely non-co- located LWA scenario for a non-ideal backhaul and co-located LWA scenario for an ideal/ internal backhaul. The overall architecture for the non-co-located LWA scenario is illus- trated in Fig. 8.28, where the Wi-Fi Termination (WT) terminates the Xw interface for Wi- In LWA, the radio protocol architecture that a particular bearer uses depends on the LWA backhaul scenario and the way the bearer is set up. The LWA supports two bearer types: (1) vEPC (MME/S-GW) ((p)) Figure 8.28 Non-co-located LWA architecture [5]. 892 Chapter 8 Split LWA Switched LWA Split LWA Switched LWA bearer bearer bearer bearer bearer bearer LWAAP LWAAP Figure 8.29 LWA radio protocol architecture for the colocated and non-co-located scenarios [5,7]. split LWA bearer and (2) switched LWA bearer. These two bearer types are depicted in Fig. 8.29 for the co-located and the non-co-located scenarios. In the non-co-located LWA sce- nario, the eNB is connected to one or more WTs via Xw interface. In the co-located LWA scenario, the interface between LTE and Wi-Fi is implementation specific. For LWA, the only required interfaces to the core network are S1-U and S1-MME which are terminated at the eNB. The Wi-Fi has no core network interface. In the downlink and for the PDUs that are sent over the Wi-Fi link, the LTE-Wi-Fi aggrega- tion adaptation protocol (LWAAP) entity generates an LWAAP PDU (i.e., a PDU with data radio bearer (DRB) ID generated by LWAAP entity for transmission over Wi-Fi) containing a DRB identity and the WT. An LWA UE determines that the received PDU belongs to an LWA bearer and uses the DRB identity to determine to which LWA bearer the PDU belongs to.

In the uplink and for the PDUs sent over Wi-Fi, the LWAAP entity in the UE generates LWAAP PDU containing a DRB identity. The LWA supports split bearer operation where the PDCP sublayer handles sequential delivery of upper-layer PDUs based on the reordering procedure introduced for dual con- nectivity. An LWA UE may be configured by the eNB to send PDCP status report or LWA status report in cases where feedback from WT is not available. Note that only RLC AM and RLC UM can be configured for an LWA bearer. The LTE RAN does not simulta- neously configure LWA with dual connectivity, LWIP, or RAN-controlled LTE-Wi-Fi interworking for the same UE. If LWA and RAN-assisted Wi-Fi interworking are simulta- neously configured for the same UE in RRC_CONNECTED, the UE only utilizes LWA. For LWA bearer, if the data available for transmission is greater than or equal to a thresh- old configured by LTE RAN, the UE decides which PDCP PDUs are sent over Wi-Fi or LTE links. If the data available is below the threshold, the UE transmits PDCP PDUs on LTE or Wi-Fi as configured by LTE RAN. For each LWA DRB, LTE RAN may configure the IEEE 802.11 access category value to be used for the PDCP PDUs that are sent over Wi-Fi in the uplink. Operation in Unlicensed and Shared Spectrum 893 In the non-co-located LWA scenario, Xw user-plane interface (Xw-U) is defined between eNB and WT. The Xw- interface supports flow control based on feedback from WT. The flow con- trol function is applied in the downlink when an E-UTRAN radio access bearer (E-RAB) is mapped to an LWA bearer, that is, the flow control information is provided by the WT to the eNB to control the downlink user data flow to the WT for the LWA bearer. The operation, administration and maintenance (OAM) configures the eNB with the information of whether the Xw downlink delivery status provided by a connected WT concerns LWAAP PDUs successfully delivered to the UE or successfully transferred toward the UE. The Xw-U interface is used to deliver LWAAP PDUs between eNB and WT. In an LW

A architecture, the S1-U terminates at the eNB and, if Xw-U user data bearers are associated with E-RABs for which the LWA bearer option is configured, the user-plane data is transferred from the eNB to the WT using the Xw- - U interface. Fig. 8.30 shows the user-plane connectivity of eNB and WT in LWA, where the S1-U terminates at the eNB and the eNB and the WT are interconnected via Xw-U. In the non-co-located LWA scenario, Xw control-plane interface (Xw-C) is defined between eNB and WT. The application layer signaling protocol is referred to as Xw appli- cation protocol (Xw-AP). The Xw-AP protocol supports the transfer of Wi-Fi metrics from WT to eNB; LWA for a UE in ECM-CONNECTED state; establishment, modifica- tion, and release of a UE context at the WT; control of user-plane tunnels between eNB and WT for a specific UE for LWA bearers; general Xw management and error handling func- tions including error indication, setting up Xw, resetting Xw, and updating the WT configu- ration data. The eNB-WT control-plane signaling for LWA is performed through Xw-C - interface. There is only one S1-MME connection per LWA UE between the eNB and the MME. The coordination between eNB and WT is performed via Xw interface signaling. vEPC (S-GW) vEPC (MME) S1-MME ((q)) ((9)) User-plane Control-plane Figure 8.30 User-plane and control-plane connectivity of eNB and WT for LWA [10]. 894 Chapter 8 Fig. 8.30 shows control-plane connectivity of eNB and WT, where the S1-MME is termi- nated at the eNB and eNB and the WT are interconnected via Xw-C. The Wi-Fi mobility set is a set of one or more Wi-Fi access points identified by their BSSID/ESSID/SSIDs (these identifiers are used to describe different sections of a Wi-Fi network), in which Wi-Fi mobility mechanisms apply while the UE is configured with LWA bearer(s), that is, the UE may perform mobility between Wi-Fi APs belonging to the mobility set without informing the eNB. The eNB provides the UE with a Wi-Fi mobility set. When the UE is configured with a Wi-Fi mobility set, it wi

ll attempt to connect to a Wi-Fi AP whose identifiers match the ones in the configured mobility set. The UE mobility relative to Wi-Fi APs not belonging to the UE mobility set is controlled by the eNB, for example, updating the Wi-Fi mobility set based on measurement reports provided by the UE. A UE is connected to only one mobility set at a time. All Wi-Fi APs belonging to a mobility set share a common WT which terminates Xw-C and Xw-U. The termination points for Xw-C and Xw-U may differ. The Wi-Fi identifiers belonging to a mobility set may be a subset of all Wi-Fi identifiers associated with the WT [5]. The UE supporting LWA may be configured by the LTE RAN to perform Wi-Fi measure- ments. Wi-Fi measurement object can be configured using Wi-Fi identifiers (BSSID, ESSID, or SSID), Wi-Fi carrier information and Wi-Fi band (2.4, 5, and 60 GHz). The Wi- Fi measurement reporting is triggered using RSSI measurement. A Wi-Fi measurement report contains RSSI and Wi-Fi identifier and may contain Wi-Fi carrier information, Wi-Fi band, channel utilization, station count, admission capacity, backhaul rate, and an indication whether the UE is connected to the Wi-Fi [5]. The LTE and Wi-Fi integration architecture requires Wi-Fi-specific core network nodes and interfaces (as shown by dotted lines in Fig. 8.31). However, the LWA scheme is different vEPC/EPC PDCP aggregation and Circuit switch PDCP reordering (IMS) PDCP splitting Gy/Gz S1-MME LWA UE LTE Uu modem Wi-Fi IEEE 802.11 Wi-FiAl modem Gateway Figure 8.31 LWA network architecture (non-co-located case) [5,18]. Operation in Unlicensed and Shared Spectrum LWAUE Application TCP/UDP/IP LTE modem LTE Uu Wi-Fi modem LWAAP Wi-FiAP LWAAP GTP-L GTP-U IEEE 802.11 L1/L2 01/L2 LWAeNB Figure 8.32 LWA protocol architecture (non-co-located case, user-plane) [5,18]. because LTE and Wi-Fi are aggregated at the radio link level. The LWA protocol architec- ture, data aggregation at the PDCP sublayer, signaling and interfaces between eNB and Wi- Fi AP, etc., have been specified by 3GPP and

illustrated in Figs. 8.31 and 8.32. The LWA architecture introduced a new interface Xw which was defined for communica- tion between eNB and Wi-Fi AP, which is similar to X2, where user data is delivered through GTP tunnel, while control messages are delivered as Xw-AP - messages over SCTP. Upon arriving at eNB, downlink user traffic is split in PDCP sublayer and is for- warded over LTE and Wi-Fi. Some PDCP packets are delivered via data radio bearer over the LTE radio link and other packets are delivered to Wi-Fi AP by eNB, which adds a DRB ID to the packets to indicate which DRB they belong to and delivers the LWA PDUs to Wi-Fi AP through the IP tunnel established over Xw [5]. Wi-Fi AP then sets the packet type to PDCP and forwards the LWA PDU to LWA UE over IEEE 802.11 interface. Upon receiving IEEE 802.11 frames, the LWA UE forwards the frames to LTE PDCP sublayer, if the packet type is set to PDCP. The PDCP sublayer then collects PDCP packets received from LTE and Wi-Fi that belong to the same LWA bearer by checking their DRB IDs and aggregates them through reordering. The LWA adaptation protocol supports LWA operation as shown in Fig. 8.32 [5]. 8.3 Physical Layer Aspects To support a flexible LAA operation, a new type 3 frame structure was introduced in LTE Rel-13, in which UE considers each subframe as empty unless downlink transmission is detected in that subframe. For LAA, the LBT procedure can be completed at any time and downlink transmission may not start/end at the subframe boundary. Frame structure type 3 896 Chapter 8 is applicable to LAA unlicensed carrier exclusively with normal cyclic prefix. Each radio frame is 10 ms long and consists of 10 subframes of length 1 ms. Any of these 10 subframes can be used for uplink/downlink transmission or can be empty. LAA transmission can start and end at any subframe and can consist of one or more consecutive subframes in the burst. Partial subframes are introduced by 3GPP, as part of frame structure type 3, to support LBT and for efficient use of unlicensed

spectrum by LAA scheme. Partial subframe in LAA uplink burst is transmitted either from the start of the zeroth symbol or from the start of the first symbol in a subframe. The transmission can also start between the zeroth and the first symbol [1]. Therefore, the zeroth symbol in the first subframe of an LAA uplink burst can be partially filled, fully filled, or empty. The LAA uplink transmission can end either at the 12th or 13th OFDM symbol in a subframe [2]. Therefore, the last subframe in an LAA uplink transmission can be filled with 14 OFDM symbols or can be partially filled with 13 OFDM symbols. The LAA downlink transmission can start from the zeroth OFDM symbol (subframe boundary) or from the seventh OFDM symbol (second slot starting position) of a subframe [3,4]. The LAA downlink transmission can either end at the subframe boundary or at any existing downlink pilot time slot (DwPTS) symbol (in frame structure type 2). Therefore, the last subframe can be completely occupied with 14 OFDM symbols or consist of any of DwPTS symbols, that is, 3, 6, 9, 10, 11, or 12 OFDM symbols [1]. Example con- figurations of frame structure type 3 are shown in Fig. 8.33 [1]. LAA uplink burst LAA downlink burst Partial ending subframe Partial starting subframe Partial ending subframe Slot 0 Slot 1 11 12 13 11 12 13 12 Symbols Partial starting subframe Partially filled 1st 13 Symbols symbol Subframe boundary 11 Symbols Completely empty 1st 10 Symbols symbol LAA burst 9 Symbols 11 12 13 6 Symbols 3 Symbols Figure 8.33 Example configurations of frame structure type 3 [1]. Operation in Unlicensed and Shared Spectrum Cell selection and synchronization rely on the reception of the synchronization signals, that is, primary and secondary synchronization signal (PSS/SSS) as well as cell-specific reference sig- nals (CRS). In particular, PSS/SSS can be used for physical cell identity detection, and the CRS can be used to further improve the performance of cell ID detection. These synchroniza- tion signals are further used to acquire c

oarse and fine time/frequency synchronization, although large time/frequency offset between two successive downlink bursts is unlikely due to multiple DRS transmission opportunities within a DRS measurement timing configuration (DMTC) occasion which has a duration of 6 ms. Therefore, the synchronization based on DRS in LAA systems can achieve reliable performance. The presence of downlink subframe must be detected by the LAA UE because the eNB does not always transmit in LAA scenar- ios. The exact detection method depends on UE implementation. LTE-based LAA supports transmission modes (TMs) with CRS-based CSI feedback, includ- ing TM1, TM2, TM3, TM4, and TM8, and those with CSI-RS-based CSI feedback, includ- ing TM9 and TM10 [3]. The CSI-RS/CSI-IM for CSI measurement is present in the configured periodic CSI-RS/CSI-IM subframes within downlink transmission bursts. Similar to the legacy LTE systems, both periodic and aperiodic CSI reports are supported in LAA scenarios. Unlike the legacy LTE system where the CRS/CSI-RS transmission power, or energy per resource element, is fixed, CRS/CSI-RS transmission power on LAA SCell is only fixed within a downlink transmission burst and can vary across downlink transmission bursts. Thus, the UE should not average CRS/CSI-RS measurements across different trans- mission bursts. The UE could either rely on CRS detection or common control signaling to differentiate the downlink bursts. LTE supports two different scheduling approaches, namely cross-carrier scheduling and self-scheduling. With cross-carrier scheduling, the con- trol information including scheduling indication on PDCCH and the actual data transmission on PDSCH take place on different carriers, whereas they are transmitted on the same carrier in the case of self-scheduling. Due to the uncertainty of channel access opportunities on unlicensed carriers, the synchronous uplink HARQ protocol with fixed timing relation between retransmissions was difficult to use in LAA. Thus, asynchronous HARQ protocol is used for LAA d

ownlink and uplink. For LAA uplink, in particular, UEs would need to rely on the uplink grant from eNB for (re)transmissions. Enhanced licensed-assisted access was introduced in 3GPP Rel-14. It defines how a UE can access 5 GHz band to transmit data in the uplink direction. Because all uplink transmissions in LTE are scheduled and controlled by the serving eNB, the channel contention between devices and the LBT scheme, which was defined in LAA for downlink operation, needed to be adapted to the uplink direction. Furthermore, the regulatory requirements for access to 5 GHz band vary in different regions. For example, ETSI requires that the OCB, defined by 3GPP as the bandwidth containing 99% of the signal power, to be between 80% and 100% of the declared nominal channel bandwidth [29]. As an initial approach, multi-cluster PUSCH transmission, which was specified in 3GPP Rel-10, was considered to fulfill this 898 Chapter 8 ETSI requirement [29]. Multi-cluster PUSCH allows two clusters of resource blocks to be scheduled far enough from each other to fulfill, for example, the 80% bandwidth require- ment. However, further studies in 3GPP suggested that the latter was not an efficient solu- tion. Furthermore, the PSD limits defined for the 5G GHz band was another limitation of the multi-cluster PUSCH approach as shown in Fig. 8.34. The LTE-based eLAA supports multi-subframe scheduling, where an uplink grant can schedule PUSCH transmissions in a number of consecutive subframes ranging from two to four subframes. It further supports DCI formats 0A and 4A for single subframe scheduling and DCI formats 0B and 4B for multi-subframe scheduling for single-layer and multi-layer PUSCH transmissions, respec- tively (see Fig. 8.35). As an example, the 10 dBm/MHz defined by the ETSI over 5150-5350 MHz frequency range is shown in Fig. 8.34, where ETSI allows a PSD of 17 dBm/MHz (over 5470-5725 MHz) if TPC can be applied [29]. The US FCC defines 11 dBm/MHz PSD for 5150-5350 MHz frequency range. As a result, 3GPP adopted the principl

e of B-IFDMA Occupied Bandwidth > 80% of the nominal channel bandwidth First RB Last RB Occupied bandwidth Channel bandwidth Power spectral density (PSD) per MHz limits maximum output power (10 dBm/MHz) 23 dBm total power with transmit power control (TPC) Guard band Occupied bandwidth (LTE channel) 6 PRB allocation 6 PRB allocation Figure 8.34 ETSI requirements for transmission 5 GHz band [22,27,29]. Operation in Unlicensed and Shared Spectrum Single subframe scheduling Multi subframe scheduling Figure 8.35 Multi-subframe scheduling versus single subframe scheduling [23]. for eLAA. The available number of resource blocks is organized in interlaces that are equally spaced in the frequency domain. A UE can now transmit on one or multiple inter- laces. Similar to LTE uplink transmissions, the total number of allocated resource blocks must be multiples of 2, 3, and 5 to minimize the complexity of the DFT precoding. To incorporate this specific access mode into the standard, 3GPP Rel-14 specified a new uplink resource allocation type 3 that is only applicable for LAA. As a result, two new scheduling grants (DCI format 0A/0B and 4A/4B) were introduced. A DCI transports downlink, uplink or sidelink scheduling information, requests for aperiodic CQI reports, LAA common infor- mation, or uplink power control commands for one cell and one RNTI. The RNTI is implic- itly encoded in the CRC. DCI formats 0A and 4A schedule single subframes for single antenna (OA) and multi-antenna (4A) transmission, whereas DCI formats OB and 4B allow scheduling of up to four consecutive subframes for SISO and MIMO, respectively. These new uplink grants provide the UE with a resource indication value RIV that is either 5 or 6 bits (5 bits for 10 MHz channel and 6 bits for 20 MHz channel). The RIV represents a start value RB start and the actual number of allocated resource blocks L. For a 20 MHz LTE channel, corresponding to 100 resource blocks, there are 10 interlaces with 10 RBs/inter- lace. Interlace 0 contains resource blocks (0, 10, 20, 3

0, 40, 50, 60, 70, 80, 90) as shown in Fig. 8.36. If the UE transmits on 4 out of the 10 available RBs per interlace L = 4, then the RIV is calculated as RIV = N(L - 1) + RB start. With N = 10, RB start can have a value between 0 and 6 and RIV can be between 30 and 36. The RIV is signaled via DCI to the UE. In addition to this and other information, the DCI format also includes the method on how to access the channel, that is, how to perform LBT in the uplink. As shown in Table 8.3, there are two channel access types defined for eLAA, whose usage is signaled with the uplink scheduling grant (DCI formats 0A, OB, 4A, 4B). The type 1 channel access procedure is identical to the procedure for LAA. The difference is that there are separate Chapter 8 Interlace 0 Interlace 9 Interlace 0: (0,10,20,...,90) Interlace 1: (1,11,21,...,91) Interlace 9: (9,19,29,...,99) Figure 8.36 eLAA interlaces for 20 MHz LTE channel 22]. channel access priority classes defined for the uplink. Type 2 is a procedure similar to trans- mitting DRS in the downlink. After sensing that the channel is idle for 25 us, the device can start its PUSCH transmission [22]. It must be noted that DCI format 0A is used for the scheduling of PUSCH in an LAA SCell and DCI format 0B is used for the scheduling of PUSCH in each of the multiple subframes in an LAA SCell. Furthermore, DCI format 1C is used for very compact scheduling of one PDSCH codeword, notifying multicast control channel change, notifying single-cell multicast control channel change and direct indication, reconfiguring TDD, and LAA common information, whereas DCI format 4 is used for the scheduling of PUSCH in an LAA SCell with multi- antenna port transmission mode. DCI format 4B is used for the scheduling of PUSCH with multi-antenna port transmission mode in each of the multiple subframes in an LAA SCell. The LTE-based LAA supports two alternative approaches to multi-carrier LBT. In the first approach, the eNB is required to designate a carrier requiring LBT with random backoff and the eNB ca

n sense other configured carriers with single interval LBT, only if the eNB com- pletes the LBT with random backoff on the designated carrier. In the second approach, the eNB performs LBT with random backoff on more than one unlicensed carriers and can transmit on the carriers that have completed the LBT with potential self-deferral to align transmissions over multiple carriers. An LTE eNB can access multiple carriers on which LAA SCell(s) transmis- sion(s) are performed using either Type A or Type B procedures as follows [4]: Type A multi-carrier access procedure: The eNB performs channel access on each car- rier CiEC, where C is a set of carriers on which the eNB intends to transmit and Operation in Unlicensed and Shared Spectrum 901 Ncarrier - 1 is the carrier index. The eNB transmission may include PDSCH/ PDCCH/ePDCCH on an LAA SCell(s) carrier(s), after sensing the channel to be idle during the slot durations of defer period Tdefer and after the counter N is initialized adjusted by sensing the channel for additional slot duration(s). The counter N is deter- mined for each carrier Cj and is denoted as Nci. In Type A1, the counter N is indepen- dently determined for each carrier. If the absence of another technology sharing the carrier cannot be guaranteed over a long period of time, and when the eNB stops trans- mission on any carrier CjECACi # Cj, it can continue to decrement Nci when idle slots are detected either after waiting for a duration of 4Ts or after reinitializing Nci. In Type A2, the counter Ncj is determined for the carrier CjEC, where Cj is the carrier that has the largest contention window CW value. For each carrier Ci, Nci = Ncj When the eNB stops transmission on any carrier for which Nci is determined, it reinitializes Nci for all carriers. Type B multi-carrier access procedure: A carrier CjEC is selected by the eNB by ran- domly choosing Cj from C before each transmission on multiple carriers CiECVi=0,1, Ncarrier - 1, or the eNB selects Cj less frequently, where C is a set of carriers on whi

ch the eNB may transmit. To transmit on a carrier Ci # CiVciEC, the eNB performs carrier sensing for a minimum sensing interval Tmc 25 us prior to trans- mitting on carrier Cj, and it may transmit on carrier Ci immediately after sensing the car- rier Ci to be idle for at least the sensing interval Tmc. The carrier Ci is considered to be idle for Tmc if the channel is sensed to be idle during all time intervals in which such idle sensing is performed on carrier Cj in a given time interval Tmc. The eNB will not continuously transmit on a carrier Ci # CiTCiEC for a period exceeding T'MCOT' where the value of the latter parameter is determined using the channel access class used for carrier Cj. In Type B1, a single CWp value is maintained for the set of carriers C. For determining CW for performing channel access on carrier Cj, if at least 80% of HARQ feedback corresponding to PDSCH transmission(s) in reference subframe k of all car- riers CiEC are determined as NACK, then CW is increased for each priority class pe {1, 2, 4} to the next higher permissible value. In Type B2, a CW value is main- tained independently for each carrier C. For determining Ninit for the carrier Cj, CW value of carrier Cj1 EC is used, where Cj1 is the carrier with largest CW among all car- riers in set C. Discontinuous transmission on an unlicensed carrier with limited maximum transmis- sion duration has an impact on some LTE/NR essential functionalities that support AGC setting, time and frequency synchronization of the UEs, and channel reservation. Channel reservation refers to the transmission of signals by an LAA node after gaining channel access via a successful LBT operation, SO that other nodes that receive the transmitted signal with energy above a certain threshold sense the channel to be occupied. 902 Chapter 8 8.4 Layer 2/3 Aspects In this section, we discuss the L2/L3 aspects and impacts of unlicensed band operation in LTE and NR-based systems. The uncertainty of the cell availability for HARQ retransmis- sions on unlicensed carr

iers would create some limitations for normal HARQ operation. The uncertainty may arise due to the LBT operation needed to acquire the channel or because the maximum transmission duration for LAA is exceeded. Fig. 8.37 illustrates an example in which the maximum transmission duration of an LAA cell has been reached before the third retransmission of a HARQ process can be performed. There are two ways to address this issue. One approach is to keep the HARQ retransmissions on the same LAA cell or to ensure that the HARQ process is completed within the maximum transmission duration of an LAA cell as shown in Fig. 8.37. A new HARQ process is started when the LAA cell reacquires the channel after an LBT operation. Alternatively, a retransmission can be delayed until the LAA cell acquires the channel again. The RLC retransmission may be invoked if HARQ transmission is not successfully completed. Another alternative would be to move HARQ retransmissions to another cell (note that each carrier in the carrier aggrega- tion framework is referred to as a cell). The retransmission may also be performed via another cell which would be either the PCell or another SCell. This would change the base- line LTE HARQ protocol as it may be linked to two or more cells. In synchronous HARQ, the UE identifies the HARQ process that is associated with the cur- rent transmission time interval (TTI). Depending on whether spatial multiplexing is used, each TTI has one or two associated HARQ processes, and for the identified HARQ 3rd Transmission Alternative 2 Primary cell 2nd Transmission Alternative 1 3rd Transmission 1st Transmission LAA cell 1 2nd Transmission 1st Transmission New data Figure 8.37 Illustration of HARQ issue in LAA cell [9,12]. Operation in Unlicensed and Shared Spectrum 903 processes, the UE will perform a (re)transmission. The association between TTIs and HARQ processes relies on uplink HARQ being synchronous and is derived from the timing relation. If there has been an initial transmission in a certain subframe for a ce

rtain HARQ process, then Tround-trip later, the same HARQ process is considered for retransmission. In LTE, the UE receives uplink HARQ feedback from the eNB on PHICH. PHICH received in subframe k relates to the transmission in the subframe k - 4. If the UE receives NACK on PHICH (but does not receive a grant), a non-adaptive retransmission is automatically trig- gered. LTE-based LAA uses asynchronous HARQ in the uplink. To enable asynchronous HARQ, the eNB needs to know the HARQ process that the UE is using when performing a transmission/retransmission to correctly choose the soft-buffer where the received transmis- sion should be combined. Hence, the eNB needs to indicate to which HARQ process a grant is related and the redundancy version that should be used SO that the UE uses the correct HARQ process with the correct redundancy version when performing a transmission or retransmission. Therefore, with the uplink asynchronous HARQ protocol, all transmissions or retransmissions are scheduled via (e)PDCCH. The process index is indicated in the HARQ process index field in the uplink grant. For synchronous uplink HARQ, the number of HARQ processes is not explicitly specified. Instead, the supported number is derived from the HARQ timing. For asynchronous uplink HARQ, a maximum number of HARQ processes may need to be specified. The exact number of uplink HARQ processes in the eNB may be left to implementation [12]. In synchronous uplink HARQ, the UE can expect to receive uplink grants at a specified time as the HARQ process follows a fixed pattern. It should be noted that the UE monitors (e)PDCCH once in every HARQ RTT even if the UE has received ACK in PHICH. With asynchronous uplink HARQ, the UE does not know when to expect grants as the eNB may send them at any time. In addition, if HARQ buffer is not flushed, the UE would never stop monitoring (e)PDCCH according to the current mechanism. Therefore, with asynchronous uplink HARQ, the DRX behavior needs to be modified. In carrier aggregation, the same DRX operatio

n applies to all configured and activated serving cells (i.e., identical active time for PDCCH monitoring). In other words, a common DRX is applied to all the serving cells. The difference in the case of LAA is that due to LBT there is no guarantee that the channel is obtained for scheduling the UE at the exact moment desired by the eNB. In addi- tion, even if CCA succeeds, the eNB transmitter can only occupy the channel for a limited duration due to limited maximum transmission duration requirement. This means that the DRX timers (on-duration, inactivity timer) should be long enough or DRX cycles should be short enough to allow time for obtaining access to the channel. In LAA, there are scenarios where multiple operators may operate in the same frequency channel. If these operators do not coordinate the allocation of physical cell identifier (PCI) values across their cells, it may lead to PCI confusion or PCI collision cases. PCI confusion refers to the case where a UE discovers and configures the LAA cell of another operator 904 Chapter 8 with the same PCI value as its own operator. PCI collision concerns a UE that is in the COV- erage area of two (or more) cells which have the same PCI value where the cells belong to different operators. If the same PCI value is used for cells on the same carrier frequency in the same area, PCI confusion or collision may occur. In LTE and NR, the number of PCIs is limited to 504 and 1008, respectively. If operators do not coordinate PCI assignment in their cells, the probability of PCI confusion or collision depends on the number of LAA cells of other operators the UE can find in the PCell coverage area, increasing the risk of collision when there is a dense deployment of LAA cells in a large PCell coverage area. PCI collision will happen with a lower probability than PCI confusion because PCI collision happens when the coverage areas of two cells with the same PCI partially overlap. Considering the LTE requirement for carrier aggregation where the UE is only required to handl

e a maximum timing difference of approximately 30 us between the PCell and an SCell, the UE may not be able to receive anything from the other operator's LAA cell, if the downlink signal of that cell arrives later than + 30 us of the UE's PCell. The probabil- ity that the other operator's LAA cell overlaps with the UE's PCell is 6% assuming random timing of cells. Furthermore, for self-scheduling, the UE will not decode downlink assign- ments of the other operator's cell unless the downlink transmissions are scrambled with a C-RNTI value matching the C-RNTI assigned to the UE. Considering that there are 65536 C-RNTIs (16-bit values), if the other operator serves 20 UEs in a cell, the probability that the same C-RNTI is used is roughly 0.03%. Therefore, the probability that the UE can decode a downlink message from another operator's LAA cell is 0.0018% [14,20]. In the event of a PCI confusion when self-scheduling from the LAA cell, the probability that the UE could decode downlink is negligible (0.0018% in the scenario described earlier). However, in case of cross-carrier scheduling from the PCell, the UE would never be able to decode the downlink data for which the UE has received a downlink assignment as the data is sent in another cell. In the uplink, the UE will not be able to acquire uplink synchroniza- tion and would therefore not be able to transmit uplink traffic in the other operator's cell. However, it may send unnecessary random-access preambles to another operator's LAA cell (e.g., if it receives a PDCCH order from the PCell). It is expected that the eNB can detect PCI confusion by observing that the UE is reporting good quality for a cell, but no data communication is succeeded for this UE. The eNB can then resolve the confusion by changing the PCI for the problematic cell(s). In the event of a PCI collision, there is a non-negligible likelihood that both LAA cells become unusable to the UEs that fall in their common coverage area. This would result in the UE not being able to receive the downlink an

d/or transmit in uplink while in those cells. It is also expected that in some cases, that is, not in hidden-node cases where the two cells with the same PCI are hidden from each other but are heard by the UEs, the network can detect PCI collision by listening to carriers and avoid the collision by changing the PCI values of their cells. PCI confusion and PCI collision can be avoided completely if Operation in Unlicensed and Shared Spectrum 905 operators coordinate the PCI values for their LAA cells. Otherwise, the probability of occur- rence would be scenario dependent. In LWA, the UE mobility is transparent to the eNB while in a Wi-Fi mobility set, that is, a group of Wi-Fi access points that are controlled by one WT logical entity. As long as the UE moves between APs of the same mobility set, it does not need to inform the eNB about its movement. When the UE leaves the mobility set, the eNB may change the access point based on the Wi-Fi measurements, thus it informs the UE of its decision. Coexistence with other technologies is very important for LAA, and therefore accessing and using already congested frequencies/channels that are used by Wi-Fi access points and cli- ents should be avoided. Because efficient radio resource management is critical for the overall performance of LAA scheme, LTE defines signal quality criteria such as reference signal received power [RSRP (dBm)] and reference signal received quality [RSRQ (dB)] metrics to effectively quantify the [shared] channel conditions. The RSSI serves as the key performance indicator for interference on a given carrier. To measure RSSI, the DRS needs to be present. However, because DRS is subject to LBT, any RSSI measurement report of an LAA-capable UE needs to include a time stamp indicating when the measurements were conducted. Therefore, higher layers configure an RSSI measurement time configuration (RMTC) with a measurement period [40, 80, 160, 320, or 640 ms], a subframe offset [0,...,639] and a measurement duration [1, 14, 28, 42, or 70 OFDM symbols].

(28 OFDM symbols) (28 OFDM symbols) rmtc-SubframeOffset-r13 rmtc-SubframeOffset-r13 rmtc-Period-r13 (40 ms) < rmtc-Period-r13 (40 ms) rmtc-Period-r13 (80 ms) rmtc-Period-r13 (80 ms) Measurement Measurement period report to the eNB (CO, average RSSI) Channel Occupancy (CO)=% RSSI samples > RSSI_Threshold 70 OFDM symbols 70 OFDM symbols 70 OFDM symbols measurement measurement measurement duration duration duration RSSI measurement RSSI measurement Periodicit Periodicit (40 ms) (40 ms) Figure 8.38 RMTC configuration for channel occupancy measurements based on R attempt. In NR-U, power ramping is not applied when preamble is not transmitted due to LBT failure. This will require an indication from the physical layer to the MAC sublayer. In addition, ra-Response Window is not started when the preamble is not transmitted due to LBT failure. It is assumed that ra-ContentionResolutionTimer may need to be extended with larger values to overcome the LBT impact. For 2-step RACH, the msgA is a signal to detect the UE and a payload while the second message is for contention resolution for CBRA with a possi- ble payload. The msgA will include the equivalent information, which is transmitted in msg3 for 4-step RACH [12]. 8.5 Implementation and Deployment Considerations The main challenge in the implementation and use of LTE/NR in the unlicensed bands, partic- ularly in 5 GHz band, is the coexistence with already deployed Wi-Fi networks, where LTE/ Operation in Unlicensed and Shared Spectrum 907 NR operation would adversely impact the performance of Wi-Fi systems, while the perfor- mance of LTE/NR would remain unchanged due to reliance of Wi-Fi systems on CSMA/CA mechanism. The issue is caused because of different channel usage and access procedures of these technologies. LTE/NR is designed to operate in the licensed bands based on the assump- tion that one operator has exclusive control of a given spectrum. They will continuously trans- mit with minimum time gap even in the absence of user traffic. LTE/NR also has an almost cont

inuously transmitting protocol, as well as a periodically transmitting protocol to transmit a variety of control and reference signals. Wi-Fi, on the contrary, is designed to coexist with other technologies through random backoff and channel sensing. As a result, Wi-Fi users would have a slight chance to sense a clear channel and to transmit. To ensure fair spectrum sharing and [practically] minimum inter-system interference among different wireless technologies operating in the unlicensed spectrum, 3GPP has specified a number of Wi-Fi coexistence mechanisms. These mechanisms operate in time, frequency, or power domains. In the frequency and time-domain schemes, the goal is to separate trans- missions of LTE and Wi-Fi in frequency and time, respectively, while in the power domain, the goal is to adjust the output power of LTE nodes to a tradeoff between LTE throughput and opportunistic Wi-Fi transmission. Prior to any specification work on LAA, 3GPP con- ducted studies to investigate the feasibility of LTE operating in unlicensed bands [9]. The focus of those studies was fair sharing and coexistence with Wi-Fi systems where the crite- rion used to ensure coexistence was that an LAA network does not impact existing Wi-Fi neighbors more than another Wi-Fi network. We discussed earlier that there are two design options for LTE-based LAA LBT, that is, asynchronous and synchronous LBT. The main difference between them lies in the fact that the asynchronous LBT is based on the current DCF protocol. In this case, the LBT scheme may use IEEE 802.11 RTS/CTS signals to ensure that the channel is idle just at that moment. However, synchronous LBT is considered as a special version of asynchronous LBT, wherein, data subframes are synchronized with the licensed LTE carrier. This LBT approach required minimal changes to the LTE specifications and could use inter-cell inter- ference coordination (ICIC) mechanism already defined in earlier releases of LTE to man- age the interference among LTE base stations. The ICIC mechanism i

s illustrated in Fig. 8.39. In this figure, different shades represent different frequencies in the outer sectors where the same frequency is used in the inner sectors. Deterministic channel sharing relies on LTE centralized scheduling to periodically turn off its transmission SO that Wi-Fi users can access the shared channel. Among time-domain coexistence mechanisms, we have discussed CSAT and blank-subframe allocation. A blank- subframe is an LTE subframe where transmission is muted SO that Wi-Fi users can access the channel. Similar to CSAT, a blank-subframe allows time-domain sharing between LTE unlicensed and Wi-Fi networks. In each radio frame, the eNB can configure a certain 908 Chapter 8 Sector 1 Sector 6 Sector 2 Inner sector 1 Sector 5 Sector 3 Sector 4 Figure 8.39 Illustration of LTE ICIC scheme where cell-center and cell-edge users have frequency reuse factor of 1 and 3, respectively. Radio frame (10 ms) LTE on subframes Blank subframes Figure 8.40 Example of blank-subframe allocation in an LTE radio frame [25]. number of blanked subframes based on the measurement of Wi-Fi's traffic load (see the example in Fig. 8.40). Fairness can be achieved by adjusting the number of blank subframes in each radio frame. Blank subframe offers more flexibility than CSAT as the ratio between the non-blank and blank subframes can be dynamically adjusted at the frame level, which is shorter than a CSAT cycle. Moreover, the positions of these blank subframes in each frame do not need to be consecutive. A blank-subframe is similar to the ABS used for enhanced ICIC (eICIC) mechanism in LTE-based heterogeneous networks. An ABS is a subframe Operation in Unlicensed and Shared Spectrum Opportunistic air time Opportunistic air time of Wi-Fi before LTE of Wi-Fi after LTE lowers Tx power lowers Tx power LTE Tx power (before) LTE Tx power (after) Wi-Fi energy detection threshold Time (ms) Figure 8.41 An example of Wi-Fi transmission opportunities before and after LTE power reduction [25]. during which only control and reference s

ignals are transmitted with reduced transmit power. In contrast to ABS, a blank subframe does not include transmission of control and reference signals. The coexistence between LTE and Wi-Fi networks by adjusting the output power of LTE nodes is an alternative method to minimize the interference. The Wi-Fi nodes typically employ energy detection to determine the activities of other users. More specifically, if the aggregate received energy is above a threshold, a Wi-Fi node would consider the channel to be busy and would postpone its transmission. For LTE/Wi-Fi coexistence, one may increase the transmission opportunity of Wi-Fi nodes by reducing the output power of the LTE nodes. As illustrated in Fig. 8.41, when the LTE transmit power is reduced, the transmission window for a Wi-Fi node becomes larger and the Wi-Fi node can more opportunistically transmit. On the other hand, the reduction in LTE transmit power will also result in lower LTE throughput due to the decrease of the SINR as a result of increased Wi-Fi transmissions. Considering the potentially large number of deployed Wi-Fi APs and/or LTE-based LAA nodes, the backhaul and inter-node connectivity is another key challenge in such heteroge- neous networks. An ideal backhaul (a dedicated point-to-point connection) is considered a link which provides (one-way) transport latency less than 2.5 ms and a throughput of up to 10 Gbps. Other types of backhaul links are considered non-ideal. The unlicensed and licensed carriers in ideal backhaul deployments can be co-located or inter-connected. 910 Chapter 8 Inter-node synchronization is another deployment consideration in heterogeneous networks. Both synchronous and asynchronous scenarios have been considered between LTE-based LAA and/or Wi-Fi small cells as well as between the small cells and the macrocell(s). Modern UEs often implement multiple RATs, where very close proximity of the RF compo- nents would cause in-device coexistence (IDC) interference. 3GPP LTE Rel-11 introduced several solutions for handling t

his interference and those solutions can be used to protect Wi-Fi networks during LAA operation. The basic principle is that the UE indicates IDC interference to the serving eNB which later resolves the issue by configuring the UE with an appropriate DRX cycle, performing a handover of the UE to another cell, or completely releasing one or more SCells [5]. However, it must be noted that the use of LBT in LAA would complicate the UE scheduling because there is no guarantee that a channel can be obtained for the UE at the exact time instant determined by the eNB. The LBT mechanism also limits the duration when the channel can be occupied; therefore the DRX timers should be adjusted to be long enough or the DRX cycles should be short enough to allow time for obtaining the channel access. In LTE, when the UE detects an IDC condition, it would initially try to solve the problem internally. If this fails, the UE can inform the eNB that it is experiencing an IDC condition. Note that the detection of IDC condition in a UE is implementation specific. The UE identi- fies the frequencies that are suffering from IDC interference. If the UE determines that the IDC problem can be solved in a TDM-manner (i.e., by multiplexing the use of the interfering transceivers across time) the UE can indicate a TTI bitmap or DRX cycles that are affected by IDC interference to the eNB. When the eNB receives the indication, it can solve the pro- blems by performing a handover of the UE to other frequencies, removing the problematic cell or configuring the UE with a new DRX configuration which would solve the problem. The existing IDC solutions can be used to support Wi-Fi background scanning during LAA operation. The existing IDC solutions can also be used to indicate interference problems for cases where the UE intends to use Wi-Fi on the same or adjacent carrier to the unlicensed car- rier. If the eNB does not support IDC, the only way for the UE to enable Wi-Fi transmission would be to perform detach and attach procedures, and changing it

s capabilities to indicate that LAA is not supported. However, from a system operation viewpoint, having the UE per- form detach and attach procedures is considered undesirable. Hence, eNB should enable IDC indications and respond to the IDC requests from the multi-radio UEs. The QoS of some radio bearers might suffer when LAA is used due to support of LBT as there can be various interference sources in the unlicensed spectrum such as other RATs and LAA nodes of other operators. To improve the QoS, the characteristics of an LAA cell should be considered when mapping traffic from radio bearers to carrier(s). For example, it is better not to send critical control information, delay-sensitive data or guaranteed bit rate bearers through LAA cells, if the LBT operation is required. Operation in Unlicensed and Shared Spectrum References 3GPP Specifications [1] 3GPP TS 36.211, Evolved universal terrestrial radio access (E-UTRA), Physical Channels and Modulation (Release 15), June 2018. 3GPP TS 36.212, Evolved universal terrestrial radio access (E-UTRA), Multiplexing and channel coding (Release 15), June 2018. [3] 3GPP TS 36.213, Evolved universal terrestrial radio access (E-UTRA), Physical layer procedures (Release 15), June 2018. 3GPP TS 37.213, Physical layer procedures for shared spectrum channel access (Release 15), June 2018. [5] 3GPP TS 36.300, Evolved universal terrestrial radio access (E-UTRA) and evolved universal terrestrial radio access network (E-UTRAN); Overall Description; Stage 2 (Release 15), June 2018. [6] 3GPP TS 36.321, Evolved universal terrestrial radio access (E-UTRA), Medium Access Control (MAC) Protocol Specification (Release 15), June 2018. [7] 3GPP TS 36.323, Evolved universal terrestrial radio access (E-UTRA), Packet Data Convergence Protocol (PDCP) specification (Release 15), June 2018. [8] 3GPP TS 36.331, Evolved universal terrestrial radio access (E-UTRA), Radio Resource Control (RRC); Protocol Specification (Release 15), June 2018. [9] 3GPP TR 36.889, Study on licensed-assisted access to u

nlicensed spectrum (Release 13), June 2015. [10] 3GPP TS 38.300, NR; NR and NG-RAN overall description, Stage 2 (Release 15), June 2018. [11] 3GPP TS 38.323, NR; Packet data convergence protocol (PDCP) specification (Release 15), June 2018. [12] 3GPP TR 38.889, Study on NR-based access to unlicensed spectrum (Release 15), December 2018. Articles, Books, White Papers, and Application Notes [13] B. Ren, et al., Cellular communications on license-exempt spectrum, IEEE Communications Magazine, May 2016. [14] Keysight Technologies Application Note, 4G LTE-A in Unlicensed Band - Use Cases and Test Implications, December 2017. [15] J. Zhang, et al., LTE on license-exempt spectrum, IEEE Commun. Surveys Tutor. 20 (1) (First Quarter 2018). M. Labib, et al., Extending LTE into the unlicensed spectrum: technical analysis of the proposed variants, IEEE Communications Standards Magazine, December 2017. A. Mukherjee, et al., Licensed-assisted access LTE: co-existence with IEEE 802.11 and the evolution toward 5G, IEEE Communications Magazine, June 2016. [18] Netmanias Report, Analysis of LTE - Wi-Fi Aggregation Solutions, NMC Consulting Group, March 2016. [19] K. Mun, CBRS: new shared spectrum enables flexible indoor and outdoor mobile solutions and new busi- ness models, Mobile Experts CBRS White Paper, March 2017. A. Savoia, LTE in the unlicensed spectrum, Keysight Technologies, March 2016. H.-J. Kwon, et al., Licensed-assisted access to unlicensed spectrum in LTE Release 13. IEEE Communications Magazine, February 2017. [22] Rohde & Schwarz White Paper, LTE-Advanced Pro Introduction, eMBB Technology Components in 3GPP Release 13/14, May 2018. R. Karaki, et al., Uplink performance of enhanced licensed assisted access (eLAA) in unlicensed spec- trum, in: IEEE Wireless Communications and Networking Conference (WCNC), 2017. [24] B. Chen, et al., Co-existence of LTE-based LAA and Wi-Fi on 5GHz with corresponding deployment sce- narios: a survey, IEEE Commun. Surveys Tutor. 19 (1) (First Quarter 2017). 3GPP specifications can be acc

essed at the following URL: http://www.3gpp.org/ftp/Specs/archive/ 912 Chapter 8 [25] Y. Huang, et al., Recent advances of LTE/Wi-Fi co-existence in unlicensed spectrum, IEEE Netw. (March/April 2018). B.L. Ng, et al., Unified access in licensed and unlicensed bands in LTE-A Pro and 5G, APSIPA Transactions on Signal and Information Processing (SIP), vol. 6, Cambridge University Press, July 2017. T. Levanen, et al., 5G new radio and LTE uplink co-existence. IEEE Wireless Communications and Networking Conference (WCNC), April 2018. [28] A. Mukherjee, et al., System architecture and co-existence evaluation of licensed-assisted access LTE with IEEE 802.11, in: 2015 IEEE International Conference on Communication Workshop (ICCW), June 2015. [29] ETSI EN 301.893 v1.7.1, Broadband radio access networks (BRAN); 5 GHz high performance RLAN, 2012. [30] Part 11: Wireless LAN medium access control (MAC) and physical layer (PHY) specifications, IEEE Std 802.11-2018, May 2018. 4G Americas, LTE aggregation & unlicensed spectrum, November 2015. 5G Americas, Spectrum landscape for mobile services, November 2017. 3GPP R1-1808318, Discussion on frame structure for NR-U, ZTE, August 2018. [34] 3GPP R1-1808058, NR numerology and frame structure for unlicensed bands, Huawei, HiSilicon, August 2018. [35] 3GPP R1-1808274, On channel access procedure in NR-U, MediaTek, August 2018. Index Note: Page numbers followed by "p" and "t" refer to figures and tables, respectively. Active antenna systems (AASs), Antenna 574, 655 Access and mobility management antenna-port virtualization, 575 Active self-interference function (AMF), 40-41, aperture, 737 cancellation techniques, 115-117, 143, 243, 806 arrays, 311, 737 Access network (AN), 1, 112-140 bearing, 307 bearers and identifiers, 124-132 Additional maximum power calibration, 743 reduction (A-MPR), 676, radio bearers and packet data configurations, 733 connectors, 741 unit sessions, 124-129 radio network identifiers, Adjacent channel interference ratio design considerations, 129-132 (ACIR), 683-

685 726-727 entities and interfaces, 112-123 Adjacent channel leakage ratio panels, 307 (ACLR), 668-671, 682, E1 control-plane functions port, 392 and procedures, 121-122 Aperiodic CSI trigger state Adjacent channel selectivity subselection, 206 F1 control-plane/user-plane (ACS), 673-674, Aperiodic SRS, 604, 886 functions and procedures, 683-685, 690 119-121 Application functions (AFs), 109 Advanced driver assistance Application protocol identity (AP NG control-plane/user-plane systems (ADAS), 15-16 ID), 131-132 functions and procedures, Aerial vehicle control, 37-38 122-123 Arbitration interframe space Aggregated maximum bit rate (AIFS), 862-864 Xn control-plane/user-plane (AMBR), 172-173 functions and procedures, Area traffic capacity, 41 117-118 Aggregation level (AL), 764 Area-specific SIB, 240-241 user-plane and control-plane Allocation and retention priority Asynchronous HARQ protocol, (ARP), 146-147, 182 214, 897 protocol stacks, 132-140 Almost blank subframes (ABS), Access network discovery and Asynchronous LBT, 907 874, 907-909 selection function Atmospheric effects, 300-301 AMF Name, 130 Authentication, 185, 751 (ANDSF), 845 Analog beamforming (ABF), Authentication, authorization, and Access point name (APN), 564-573,738 124-125, 803-804 accounting (AAA), 12 Analog front-end (AFE), 714-715 Access security management entity Authentication and key agreement design considerations, 719-723 (AKA), 144-145 (ASME), 185 Access stratum (AS), 10, 60-62, Analog to digital converter (ADC), Authentication credential 199-200 321-322, 704, 706, 783 repository and processing Angle of arrival (AoA), 287 function (ARPF), 188 security, 187 Angle of departure (AoD), Authentication server function Access-network-initiated paging, 295-296 (AUSF), 143 Angular spread (AS), 291, Automated driving-related Acoustic filter technologies, 295-296 services, 796-797 735-736 914 Index Automatic gain control (AGC), 721 Bhattacharyya parameter, 516, CBG transmit indicator (CBGTI), Automatic neighbor relation 518-519 492-493 scheme (ANR s

cheme), Binary phase shift keying (BPSK), 50-51 535, 772 cell-level mobility, 259-260 Automatic repeat request (ARQ), Bipolar junction transistor (BJT), 66 cell-specific reference signals, 68-72, 197-198 Bit error rates (BERs), 76, error control mechanism, 515-516 cell-specific SIB, 240-241 539-542 Blank-subframe, 907-909 densification, 300 Autonomous uplink (AUL), Block interlaced frequency range extension, 52-53 868-869 division multiple access reselection, 165, 248 Average mutual information, (B-IFDMA), 888-890 search and random-access 367-368 Blockage, 296-298 procedure, 777-780 Average power tracking (APT), 724 Blocking, 689-690 selection, 265-266, 897 Average spectral efficiency, 39 Bluetooth, 55-58, 747, 848 virtualization, 59 Averaging window, 178, 184 Bragg diffraction, 291-292 Cell global ID (CGID), 50-51 Break-before-make mode, 109 Cell Radio Network Temporary Broadcast channel (BCH), 138, Identifier (C-RNTI), 130, Backhaul (BH), 48, 84-85, 90 Broadcast control channel Cell-defining SSBs (CD-SSBs), Backoff indicator (BI), 205 (BCCH), 137-138 Balian-Low theorem, 338-339 Broadcast multicast service center Cell-specific reference signals Bandwidth adaptation (BA), (BM-SC), 803-804 (CRS), 699, 757-758 394-395 Buffer status report (BSR), 208 Cellular IoT (CIoT), 750-751 Bandwidth part (BWP), 379-380, Bulk acoustic wave (BAW), Cellular network, 350 393-396 735-736 Cellular V2X (C-V2X), 789, 841 inactivity timer, 169-170 filter, 736-737 Central network controller, 65 Base graph, 530 Business support system (BSS), Central offices (COs), 56-57 Base station transceiver RF 15-16 Central unit (CU), 26-27, 60-62 conducted and radiated BS Butler matrix, 739 Central unit of logical gNB (gNB- transmitter/receiver CU), 26-27, 119 characteristics, 671-691 gNB-CU-CP, 121-122 requirements, 665-671 UE FIAP ID, 120, 132 Baseband modems, 719 Camping on cell, 246 Centralized C-RAN architectural Baseband signal processing, Candidate waveforms, 335-350 concept, 30 62-63 CAPEX, 21 Centralized-RAN architecture (C- Baseband unit (

BBU), 7, 67-68 Carrier aggregation (CA), 53-54, RAN architecture), 28-30, Basic service set (BSS), 859 280-283 scenarios, 881-882 high-capacity cloud-RAN, 15 reporting, 511-512 Carrier frequency offsets (CFOs), Channel busy ratio (CBR), 836 steering, 743 326-328,777 Channel coding, 515-539 sweeping, 510 Carrier sense multiple access with low density parity check coding tracking, 743 collision avoidance principles, 529-531 Beam management (BM), (CSMA/CA), 857 modulation schemes and 411-412, 509-515 Carrier sensing adaptive modulation coding scheme Beam-level mobility, 253, 260 transmission (CSAT), determination, 535-539 Beamforming, 567, 846 879-880 NR low density parity check Beamforming unit (BFU), 739 Carrier Wi-Fi, 56-57 coding, 532-534 Belief propagation (BP), 359-360 Cauchy distribution, 329-330 NR polar coding, 523-528 algorithm, 531 CBG flush indicator (CBGFI), polar coding principles, decoder, 515-516, 518 492-493 516-523 Index 915 Channel hardening, 305, 583 Cloud-based radio access, 17-18 Conducted BS transmitter/receiver Channel impulse response, 310 Cloud-radio access network characteristics, 671-691 Channel matrix, 556 (C-RAN), 7-10, 56-98 Conducted UE transmitter/receiver Channel measurement (CM), 431 architectural aspects, 57-62 characteristics, 694-699 Channel models, 287-313 backhaul transport options, Configured grant Type 1, Channel occupancy (CO), 905 86-98 212-213, 405-406 Channel polarization, 517-518 with distributed RRHs, 54-55 Configured scheduling RNTI (CS- Channel quality indicator (CQI), Coarse/dense wavelength division RNTI), 405 multiplexing (CWDM/ Connected mode procedures, Channel raster, 382 DWDM), 7-10 251-253 Channel realizations, 307-309 Cochannel dual polarization Connection density, 41-42, 54 Channel reservation, 901 (CCDP), 96-97 Connection management states, 243 Channel state information (CSI), Code block concatenation process, CM-IDLE and CM- 411-412, 571, 847 490-491 CONNECTED states, 243 measurement and reporting, Code block group (CBG), 461 Constellation rotation, 328-3

29 496-509 CBG-based retransmissions, Container technology, 13-14 Channel state information 214-215 Content caching, 6, 79 reference signals (CSI-RS), Code construction algorithms, 517 Contention window (CW), 857 425-431, 450, 888 Codebooks, 360-361 Contention-based performance, CSI-RS-based intrafrequency codebook-based precoding, 404-405 measurement, 273 638-640 Contention-based random-access, Channel state information signal- determination algorithms, 550 778-779 to-interference-plus-noise resolution, 502-503 Context-aware communication, 80 ratio (CSI-SINR), 508-509 subset restriction concept, 641 Control channels (CCH), 137, 800 Channel state information-received Coding-based NOMA schemes, PBCH, 432-439 signal strength indicator 368-369 PDCCH, 439-477 (CSI-RSSI), 508 Coherence in uplink physical layer, Channel state information- bandwidth, 293-294 605-631 reference signal received distance, 296 physical random-access power (CSI-RSRP), 431, time, 295 channel, 617-631 507-508 Collapsed gNB deployment, 26 PUCCH, 605-617 Channel state information- Co-located LWA scenario, Control elements (CEs), 203 reference signal received 891-892 Control plane entity (CP entity), 6, quality (CSI-RSRQ), 508 Commercial Mobile Alert System Check node (CND), 529-530 (CMAS), 239 Control resource set (CORESET), Ciphering, 187, 229-233 Common control channel 416-417, 441-443, Circulant matrix, 530 (CCCH), 137-138, 446-448 Circulator, 374-375 204-205 Control-and user-plane separation Citizen Broadband Radio Service Common phase error (CPE), 416, (CUPS), 28-29, 33-35, 141 (CBRS), 851-853 423-424 Control-plane Classic eigen-beamforming Common Public Radio Interface CIoT EPS optimization, scheme, 567 Forum (CPRI Forum), 7, 750-752 Clear channel assessment (CCA), 32-33, 74 CU-CP entity, 26-27 transport, 75 handover procedures, 260-263 Clear-to-send-to-self message Common resource block (CRB), latency, 43-44 (CTS-to-self message), 381-382 protocols, 113-114, 129-130, 858-859 Complementary metal-oxide 808-809 Closed access mode, 52-53 semic

onductor process stacks, 149-152 Cloud infrastructures, 13 (CMOS process), 714, 717 Conventional open-loop Cloud-based application, 794-795 Component carrier (CC), 53-54 beamforming, 585 916 Index Cooperative C-RAN architectural Data plane development kit basics and transmission concept, 30 (DPDK), 13-14 characteristics, 330-335 Cooperative radio transmission, 30 Data radio bearers (DRBs), 115, QPSK waveform, 323-325 Coordinated multipoint (CoMP), 124, 197 subcarrier mapping in, 331-332 28-29, 49 Decoding Diameter, 34-35 transmission and reception algorithms for LDPC codes, 531 Diffuse scattering, 291-292 feature, 54-55 process, 522 Digital beamforming (DBF), Core network (CN), 1, 141-152, Decoupling of RAT and core 564-573, 738-739 network, 141 Digital down-conversion (DDC), 708 architecture, 17-18 Dedicated control channel Digital precoding, 571 assistance information, 249 (DCCH), 137-138, 217 Digital predistortion (DPD), 683, CN-initiated paging, 166 Dedicated EPS bearer, 146-147 CN-level interworking, Dedicated short-range Digital subscriber line (DSL), 158-161 communication (DSRC), 48-49, 91 control-plane protocol stacks, Digital-to-analog converter (DAC), 149-152 Dedicated traffic channel (DTCH), entities and interfaces, 142-146 137-138, 217 Direct frame number (DFN), protocol data unit sessions and Deep learning models, 80 835-836 5GC identifiers, 146-149 Default EPS bearer, 146-147 Direct RF sampling transceivers, user-plane protocol stacks, 152 Default QoS rule, 177-178 704-716 Counter DAI (cDAI), 550 Delay, 293-294 Nyquist zones and sampling of Coupling loss, 47 spread, 291-295 wideband signals, 708-711 Coverage, 47 Demodulation reference signals receiver design, 714-716 Crest factor reduction (CFR), (DM-RS), 415, 417-423, transmitter design, 712-714 347-348 592-597, 772, 814, Direct-conversion RF DAC Cross-carrier scheduling, 53-54, 829-830 technology, 713-714 for PUCCH, 596-597 Direction-of-arrival beamforming, Cross-polarization discrimination for PUSCH, 593-596 (XPD), 291-292 Demodulator, 521 Disagg

regated gNB deployment, Cross-polarization interference Dense Urban-eMBB test 26-27 cancelation (XPIC), 96-97, environment, 39, 44-45, Disaggregation, 26-28 291-292 48, 307-309 Discontinuous reception (DRX), Cross-polarized signal components, Densification of wireless networks, 43, 138, 169, 171, 200, 291-292 55-56 206, 255-258 CSI-RS resource indicator (CRI), Deployment models, 853-856 Discovery reference signals (DRS), Deployment scenarios, Cumulative adjacent channel 158-164 Distributed antenna system (DAS), leakage ratio (CACLR), Extreme Long Distance 55-57 685-686 deployment scenario, 49 Distributed coordination function Cumulative distribution function Destination layer-2 ID, 839 (DCF), 857, 859-862 (CDF), 39-40, 321-322, Device energy efficiencey, 46 Distributed CRC polar code (D- Device-to-device (D2D), 80 CRC polar code), 523 Customer-facing service (CFS), communications, 22 Distributed interframe space 104-105 D2D-based positioning (DIFS), 857 techniques, 55-58 Distributed unit (DU), 26-27, mode, 792 30-32, 36 technology, 791 Distributed unit of logical gNB Data converters, 711 Device-to-network mode, 792 (gNB-DU), 26-27, 119, 121 Data network (DN), 143 DFT-spread OFDM (DFT-S- ID, 132 name, 149 OFDM), 346-347, 353 UE F1AP ID, 132 Index 917 Diverse user interactions, 14-15 CSI measurement and reporting, E1 control-plane functions and dl-UL-TransmissionPeriodicity 496-509 procedures, 121-122 parameter, 389 hybrid ARQ operation and Earthquake and Tsunami Warning Doherty amplifier, 725 protocols, 539-550 System (ETWS), 238 Domain name system (DNS), 36, PDSCH, 487-496 eCPRI, 32-33, 63. See also 105-106 reference signals, 413-432 Common Public Radio Doppler frequency component, synchronization signals, Interface Forum (CPRI 477-487 Forum) Doppler power spectrum, 296 Downlink pilot time slot (DwPTS), characteristics, 77 Doppler shift, 311, 589 895-896 specification, 78-81 Doppler spread, 292-293, 295 Downlink shared channel (DL- transport, 76-77 angular spread, 295-296 SCH), 138, 200 eCPRI radio equipment Doub

le-symbol DM-RS, 420 DRS measurement timing (eRE), 78 Downlink (DL), 51-52, 197 configuration (DMTC), 897 eCPRI radio equipment control assignment, 212 Dual connectivity (DC), 127-128, (eREC), 78 beamforming, 450 199-200 Edge computing, 109 channel-dependent scheduling, duplication, 225-226 Edge proximity services, 80 schemes, 153-158 Effective number of bits (ENOB), MIMO schemes, 551-590 Dual registration approach, 161 704-706 analog, digital, and hybrid Dual-connectivity, 728 Eigen-beamforming, 567 beamforming, 564-573 Dual-polarized antenna, 376-377 Eigenvector transform, 580-581 capacity of MIMO channels, array, 588 Electromagnetic shielding, 552-558 Duplex schemes, 369-379 376-377 FD-MIMO, 573-582 frequency and time division Element management (EM), large-scale MIMO systems, duplex schemes, 369-371 19-20 582-586 full-duplex schemes, 373-379 Element management systems NR multi-antenna half-duplex and flexible-duplex (EMSs), 15-16 transmission schemes, schemes, 371-373 Elementary procedures (EPs), 120 586-590 Duplication activation/deactivation Elementary signal estimator (ESE), SU-MIMO techniques, for MAC CE, 205-206 357-358 558-564 DwPTS. See Downlink pilot time Elevation spread of departure multiple access, 757 slot (DwPTS) (ESD), 578-579 scanning and synchronization Dynamic 3D beamforming, 574 Embedded wafer-level-ball grid + broadcast channel Dynamic blockage, 296-297 array package (eWLB grid acquisition, 43 Dynamic frequency selection array package), 718 scheduler, 210-211 (DFS), 845, 870-871 Emergency response service, transmission, 871-873 Dynamic scheduling, 475-477, 13-14 Downlink assignment index 636-637,820 Emission mask, 683 (DAI), 460 Dynamic spectrum sharing rules, Energy detection threshold (ED Downlink control information threshold), 871-873 (DCI), 207-208, 420, Energy efficiency (EE), 30, 46, 52 763-764, 822 Energy per resource element format, 457-465 (EPRE), 500-501 Downlink physical layer, 412. E-mode pHEMT device, 734 Enhanced distributed channel See also Uplink physical E-URAN random acce

Group (ISG), 99 Feed-back precoding matrix network slices accommodating ETWS. See Earthquake and selection techniques, 571 different use cases, 18 Tsunami Warning System Femtocells, 3-4, 46-47, 55-56, (ETWS) flows, 126-127 Euclidean norm of square matrix, Fifth generation network (5G framework, 172-184 568-569 network), 1, 13-18, 81, model, 177 European Conference of Postal and 728, 850-851 RAN architecture, 28-29 Telecommunications cellular systems, 303-304 reference architectures, Administrations (CEPT), deployment scenarios, 158-164 111-152 656-657 design principles and prominent access network, 112-140 European Telecommunications network topologies, 3-111 core network, 141-152 cloud-radio access network Standards Institute (ETSI), security, 185-191, 839 and virtual-radio access spectrum and regulations, Evolved packet core (EPC), 11, 26 network, 56-98 70-78 control-and user-plane heterogeneous and ultra-dense system and service separation of EPC nodes, networks, 46-56 requirements, 16-18 33-35 MEC, 98-109 3GPP 5G standardization network and service Evolved packet system (EPS), 51, activities, 66-70 750-751 requirements, 10-11 use cases, 21-25 AKA procedure, 185 network sharing, 109-111 categories, 15 bearer, 124, 172 network slicing, 37-46 and deployment scenarios, Evolved Universal Terrestrial NFV, 12-26 18-36 Radio Access (E-UTRA), separation of control and user Fifth percentile user spectral 224-225 planes, 26-36 efficiency, 39-40 Index Figure of merit (FoM), 722-723 resource allocation in, 400-403 Global gNB ID, 130 Filter bank multicarrier (FBMC), PRB bundling, 403 Global ITS spectrum, 801-802 337-340 Zadoff-Chu sequence, 830-831 Global navigation satellite system Filtered-orthogonal frequency Frequency range (FR), 287 (GNSS), 55-58, 89-90, division multiplexing (F- Fresnel zone, 288 797-798 OFDM), 336-337 Frobenius norm, 321-322 Global positioning system (GPS), First Fresnel ellipsoid, 288 Fronthaul bandwidth, 73-74 89-90 Five-point beam test, 743 Fronthaul latency, 73-74 Global synchronization channel

5G core network (5GC network), Fronthaul transport options, 62-86 number (GSCN), 383 1, 26 Fronthaul-I (NGFI-I), 84-85 gNB, 28 identifiers, 146-149 Fronthaul-II (NGFI-II), 84-85 entity, 115 5G QoS identifier (5QI), 175-177, Full-dimension MIMO (FD- scheduler, 210, 213-214 180-181 MIMO), 67-68, 573-582 UE context, 120-121 Fixed frame period (FFP), 883 Full-duplex schemes, 373-379 UE NG application protocol ID, Flexible-duplex scheme, Full-functional base station, 47 371-373 Fully depleted/partially depleted gNB Identifier (gNB ID), 130 Flip chip method, 718 silicon on insulator (RF- Gold sequence, 421, 427 ForCES, 35 SOI/FD-SOI), 718 GPRS tunneling protocol, Forward error correction (FEC), Fully-qualified domain name 117-118, 152 539-542 (FQDN), 130 Grant-based performance, Forwarding and control element Functional splits, 68 404-405 separation, 28 options, 62-86 Grant-free/semi-persistent Fractional-T/E carrier, 91 scheduling, 403-407 Frame structure Group-common PDCCH (GC- NB-IoT, 757 PDCCH), 439 and numerology, 383-391 Gabor system, 338-339 Guaranteed bit rate (GBR), 51-52, type 3, 886-887 Gabor's theory of communication, 172-173 Frame timing synchronization, 821 314-315 Guaranteed flow bit rate (GFBR), Frame-based equipment (FBE), Gallium arsenide (GaAs), 175-176 716-717 Guard subcarriers, 320 Free-space conditions, 288 E-mode pHEMT technology, Guard-band, 754 Free-space path loss, 287-288 734, 736-737 mode, 755 Frequency GaAs-based devices, 734-735 bands, 303-304 mHEMT technology, 735 Gap interval, 765-767 error, 589 frequency-selective fading Gateway core network approach Half-duplex FDD (H-FDD), effects, 294-295 (GWCN approach), 109 371-372 frequency-selective resource General authorized access (GAA), Half-duplex scheme, 371-373 allocation, 586 Handover, 258-274 hopping, 771-772 Generalized frequency division interruption time, 47-48 flag, 459 multiplexing (GFDM), Handover Command message, synchronization, 89-90 343-344 Frequency division duplex (FDD), Generator matrix, 519, 529-530 Handover Complete messag

e, 49-50, 314, 369-371, Generic PA, 348-350 826-827 Generic public subscription Hardware virtualization, 13 Frequency division duplex/time identifier (GPSI), 149 Header compression, 222-223 division duplex (FDD/ Geometry-induced blockage, function, 226-229 TDD), 49 296-297 Heterogeneous networks, 46-56 Frequency domain, 391-392 Gigabit passive optical network, Heterojunction bipolar transistor equalization, 332 91-94 (HBT), 66, 716-717 920 Index HetNet Hybrid NOMA, 364-365 Information and Communication architecture, 91 Hyper frame number (HFN), Technologies (ICT), 83 deployments, 52-53 229-230, 757 Information element (IE), topology, 90 Hypervisor, 13-14 882-883 High-electron-mobility transistor Information-centric networking (HEMT), 734-735 (ICN), 66-67, 141 High-level architectural Infrastructure layer, 43-44 requirements for 5G, Idle mode procedures, 246-248 Integrated LNAs, 732 50-53 IDLE state, 43-44 Integrated passive devices (IPDs), High-level language for SDN IEC 61907 2009 standard, 58-59 76-78 applications, 30-31 IEEE 1588v2 precision timing Integrity High-power UE (HPUE), 730 protocol, 89-90 check, 187 High-spatial-resolution CSI, 503, IEEE 1914 working group, 32-33 protection functions, 229-233 505-506 IEEE 1914.1 Next-Generation Intelligent transport systems (ITS), Higher accuracy positioning, Fronthaul Interface, 82-83 60-62, 791 55-58 IEEE 1914.3 Radio-Over-Ethernet Intended number of paging Hilbert-Schmidt norm, 321-322 Transport, 81 attempts, 269 Home eNB concept (HeNB IEEE 802.11 operation principles, Inter-gNB handover procedure, concept), 52-53 846-847, 857-865 259-260 Home subscriber server (HSS), IEEE P1914.1 standard, 32 Inter-RAT handover process, IEEE P1914.3 standard, 32 59-60 Hotelling transform, 580-581 IKE protocol, 145 Inter-RF bandwidth gap, 690-691 Hotspot 2.0, 56-57 IKEv2 protocol, 145 Intercell coordination function, Hybrid automatic repeat request Imperfect IQ mixer, 377-378 54-55 (HARQ), 763-764, In-band, 754 Intercell interference cancelation 796-797, 847, 902-903 emissions, 6

96 (ICIC), 53, 907 entity, 135 mode, 755 Interdigital transducers (IDTs), functionality in MAC sublayer, In-channel selectivity (ICS), 677, 736-737 213-214 Interference management, 55 HARQ-ACKs, 214 In-device coexistence (IDC), 796 Interference measurement (IM), operation and protocols interference, 910 semi-static/dynamic codebook In-phase and quadrature (I/Q) Interference mitigation techniques, hybrid ARQ-ACK data message type, 78-80 72-73 multiplexing, 548-550 digital radio samples, 62-63 Intermodulation products, UE processing times, HARQ Inactive mode procedures, 348-350 protocol and timing, 248-251 Internal-group identifier, 149 542-548 Inactive RNTI (I-RNTI), 130 International Mobile Equipment principles, 539-542 INACTIVE state, 43 Identity (IMEI), 129-130 protocols, 411 Inactivity-timer, 171 International Mobile Subscriber retransmissions, 44-45 Incremental redundancy HARQ Identity (IMSI), 129-130, time, 43 scheme (IR HARQ 148-149,781-782 and scheduling, 775-776 scheme), 540-542 International Mobile synchronous uplink, 903 Indoor hotspot (InH), 287 Telecommunications timing, 64 Indoor hotspot-eMBB test (IMT), 20-21, 82-83, Hybrid beamforming (HBF), environment, 39, 44-45, 656-657 564-573,739 48, 307-309 International Mobile Hybrid coordination function Industrial, scientific, and medical Telecommunications-2020 (HCF), 862-864 (ISM bands), 846 (IMT-2020), 62-66 Hybrid fiber-coaxial cable (HFC radio communications in, 848 test environment mapping, 49 cable), 91-94 Industrial automation, 53 3D channel model, 307-309 Index International Telecommunication Linear MMSE (LMMSE), Union (ITU), 18 357-358 Karhunen-Loève theorem, ITU-T, 66-67 580-581 receiver, 561 International Telecommunication KASME, 185 Link adaptation in uplink physical Union Radio- Key performance indicators layer, 643-651 Communication Sector (KPIs), 36-62, 85-86 Link-to-system mapping, 44-45 (ITU-R), 18, 62-66 Kronecker model, 312 Listen-before-talk (LBT), 845, LoS path loss model, 301 product correlation model, 877, 887 Study Group 5, 63-64 5

80-581 Load-based equipment (LBE), 883 Internet Engineering Task Force Local oscillator (LO), 712 (IETF), 66-67, 222-223 Local radio resource allocation Internet of Things (IoT), 11, strategies, 55 13-14, 24, 62, 82, 101, Large antenna arrays, 564 Location/positioning service, 167, 747, 840, 856. See also design and implementation, 55-58 Narrowband IoT (NB-IoT) 737-744 Log likelihood ratios (LLRs), general aspects and use cases, array calibration, 740 489-490 749-750 Large-scale MIMO systems, Logarithmic encoding algorithms, 64 implementation and deployment 16-17, 411-412, Logical channel groups (LCGs), considerations, 783-785 577-578, 582-586 layer 2/3 aspects, 775-783 Layer 2/3 aspects Logical channel ID (LCID), 203, network architecture and IoT, 775-783 808-809 protocol aspects, 750-756 cell search and random-access Logical channels, 137-138, physical layer aspects, 757-775 procedure, 777-780 201-202,776-777 Internet protocol (IP), 12, hybrid automatic repeat prioritization procedure, 208 750-751, 804-805 request and scheduling, Long-PUCCH, 608 Internet Engineering Task Force 775-776 Long-term evolution (LTE), 11, (IETF), 67 paging and mobility, 66-67,657-659 791 Inter-node radio resource 781-783 - co-deployment scenarios, 30 aggregation, 153 physical, logical, and control mechanisms, 69-70 Inter-node synchronization, 910 transport channels, deployment scenarios, 26 Interspersed ROHC feedback, 776-777 eNB, 900-901 232-233 PSM, 780-781 LTE-advanced base stations, 93 Inter-symbol interference (ISI), V2X communications, 831-841 LTE-Advanced Pro, 66-67 292-293 Layer indication (LI), 498 LTE-based cellular radio access Inter-vendor interoperability, 81 License-assisted access (LAA), technology, 747 Intra-AMF/UPF handover in NR, 845, 854-855, 867-877 LTE-based LAA, 845, 874-875 260-263 deployment scenarios, 854-855 LTE/NR Intra-band non-contiguous carrier Licensed frequency bands, 846 operation, 906-907 aggregation, 664 Licensed spectrum, 846 V2X security, 838-841 IP multimedia subsystem (IMS), Light detection and r

anging LTE-new radio interworking, 124-125 (LIDAR), 790 158-164 IPsec protocol, 145 Limited-buffer rate matching networks, 23-24 IPsec security association (IPsec (LBRM), 490-491 and new radio solutions for SA), 145 Line-of-sight (LoS), 60-62, 288 operation in unlicensed microwave link, 96-97 spectrum, 865-890 microwave systems, 94-96 LAA, 867-877 path loss probability, 301-303 LTE-U and MulteFire, JavaScript Object Notation probability, 302 879-881 (JSON), 26-27 propagation model, 44-45 LTE-Wi-Fi aggregation, JESD204B/C model, 707-708 Linear equalizer, 332 877-879 922 Index Long-term evolution (LTE) Massive IoT, 747, 749-750 Metro ANs, 83-84 (Continued) Massive machine type Metro cell, 91 new radio-unlicensed, communications (mMTC), Micro TRPs, 48 881-890 6, 13 -14, 20, 747 Microcells, 46-47 Rel-8, 53 Massive MTC slice, 36 Microelectromechanical systems Rel-9, 52-53 Master cell group (MCG), (MEMS), 735-736 Rel-10, 53 127-128,202 Microscopic fading, 289-290 Rel-11, 54-55 bearer, 156 Min-sum algorithm, 531 Rel-15, 870 in multi-RAT dual connectivity Minimum mean square error 3GPP LTE, 747 scheme, 153-154 (MMSE), 331, 561 X2 interface, 48-49 split bearer, 156 Mission-critical IoT, 747, Low-complexity baseband Master gNB (MgNB), 128, 153 749-750 decoding process, 783 UE XnAP ID, 132 service, 33-34 Low-density parity check Master information block (MIB), slice, 36 (LDPC), 411 200, 412 mmWave bands, 25, 74, 83 coding principles, 515-516, message, 433-434 channels, 299 529-531 Master node (MN), 154, 277 implementation and operation Low-density spreading (LDS), Matched filter, 357-358 analog front-end design 359-360 Maximum bit rate (MBR), considerations, 719-723 Low-latency communications, 20 172-173, 177-178 antenna design considerations, Low-noise amplifier (LNA), Maximum coupling loss (MCL), 726-727 686-687 47, 748-749 power consumption Low-power wide-area (LPWA), 747 Maximum flow bit rate (MFBR), considerations, 723-726 Low-voltage differential-swing 175-176 semiconductor technologies, (LVDS), 707-708 Maximum length

sequences 716-719 LTE unlicensed Forum (LTE-U (M-sequences), 478-479 large antenna array design and Forum), 845, 879-881 Maximum power reduction implementation in, LTE-Wi-Fi aggregation (LWA), (MPR), 695-696 737-744 845, 865-867, 877-879 Maximum sensitivity degradation Mobile backhaul, 89-90 LTE-Wi-Fi aggregation adaptation (MSD), 697-699 Mobile broadband, 21 protocol (LWAAP), 892 Maximum-a-posteriori estimate slice, 20 LTE-Wi-Fi radio-level integration (MAP estimate), 361-363 Mobile edge (ME), 103 with IP security tunnel Mean path loss, 289 host-level management, (LWIP), 865-867 Measurement gap length (MGL), 104-105 orchestrator, 105-106 Measurement gap repetition period platform, 105 (MGRP), 701 reference architecture, Machine learning (ML), 79, 750 Measurement model, 270-274 106-107 Macro-eNB, 53 Medium access control (MAC), system-level management, Macro-TRPs, 48 39-40, 195-196, 104-105 Macrocells, 3-4, 46-47, 91, 94 198-199, 771-772, Mobile edge computing (MEC), 3, network coverage, 70 808-809, 858-859 79-80,98-109 Macro-extension access node, 47 PDU, 204-205, 208-209 architectural aspects, 101-109 Macroscopic fading, 289 sublayer, 133-135, 137, host, 102 Make-before-break mode, 109 201-215, 285 orchestrator, 103-104 Management and orchestration Memoryless SISO channel, 553 platform manager, 103-104 (MANO), 1-2, 15-16, 23 Message authentication codes, reference points, 103 of NFVI, 25-26 230-231 services, 30-32 Management reference points Message passing algorithm (MPA), and deployment scenarios, (Mm), 103, 107 99-101 Index Mobile networks, 33-34,80-81, Multi-mode devices, 14-15 Narrowband physical downlink 111-112 Multi-operator core network control channel Mobile video optimization (MVO), approach (MOCN (NPDCCH), 763-767 approach), 109 Narrowband physical downlink Mobile-originated data, 767-768 Multi-panel MIMO, 411-412 shared channel (NPDSCH), Mobility, 44, 781 -783 Multipath 763-764, 767-769 function, 114 fading, 287 Narrowband physical random- load balancing, 51-52 multipath-intensity function, 296 ac

cess channel (NPRACH), management, 52-53, 258-274 propagation, 287 769-772, 777 network aspects of, 164-171 Multiple access (MA), 323-325 Narrowband physical uplink shared pattern, 166-167 schemes, 350-369 channel (NPUSCH), 765, robustness optimization, 51-52 non-orthogonal multiple- 772-775 Mobility management entity access schemes, 353-369 Narrowband primary (MME), 750-751, 805 OFDMA, 351-352 synchronization signal Modification period, 240 SC-FDMA, 353 (NPSS), 757-758 Modified Bessel function, 290 Multiple access channel (MAC), Narrowband reference signal Modularized functional design, 1-2 (NRS), 760 Modulation coding scheme Multiple input and multiple output Narrowband secondary (MCS), 416, 515-539, (MIMO), 59 synchronization signal 718, 808 channels, 552-553 (NSSS), 757-760 determination, 535-539 systems, 291-292 Narrowband system information Molecular oxygen absorption, 300 Multiple SCells, 874-875 block type 1 (SIB1-NB), Monolithic microwave integrated Multiple-entry PHR MAC CE, circuit (MMIC), 716-717 207-208 NAS mobility and session Monte Carlo simulation, 368 Multiple-input-single-output management (NAS-MM MSISDN, 149 (MISO), 552-553 and SM), 149-150 MU-MIMO, 358-359, 396-397, Multi-pole filters, 76-78 NEA algorithms, 128-bit, 229-230 502, 505-506 Multi-port CSI-RS, 426-427 Neighbor cell list (NCL), 248 precoded MU-MIMO system, Multi-radio dual connectivity (MR- Network 569-570 DC), 154, 254, 277-280 energy efficiency, 46 schemes, 551-552,559-560 Multi-user shared access (MUSA), identifiers, 274-276 MulteFire, 879-881 infrastructure, 79-80 Multi-RAT Mutual information, 556 management layer, 43-44 coexistence, 59-60 network-based communication, systems, 14-15 Multi-antenna networked VLC, 82 systems, 577-578 N2 interface, 151 operators, 14 transmission, 846 N26 interface, 159-161 scalability, 36-37 Multicarrier NOMA, 364-365 Narrowband control channel service Multi-cluster PUSCH transmission, element (NCCE), 763-764 behavior, 19 897-898 Narrowband IoT (NB-IoT), 747 chaining, 37-38 Multi-connectivity, 18 pr

otocol stack, 755 sharing, 52, 109-111 schemes, 153-158 uplink transmission, 755 stability, 16 Multi-layer networks, 14-15 Narrowband master information status Multi-layer SU-MIMO schemes, 559 block (MIB-NB), 760 collection process, 30-31 Multimedia broadcast/multicast Narrowband physical broadcast synchronization process, service, 54-55 channel (NPBCH), 30-31 Multimedia messaging, 22 760-762 synchronization, 49 Multimedia priority service (MPS), Narrowband physical cell identity technical domains, 82 59-60 (NB-PCID), 758-760 technology, 749-750 924 Index Network access identifier (NAI), control-plane handover layer 3 procedures, 260-263 functions, 195-201, 236-255 Network allocation vector (NAV), user-plane handover procedures, services, 236-255 858-859 263-265 mobility management, handover, Network architecture, 750-752 Network-controlled power and user equipment modes of operation, 752-755 management, 168-171 measurements, 258-274 and protocol aspects, 891-895 Neutral host network access mode, multi-radio dual connectivity, reference architecture, 277-280 802-807 New 5G NR spectrum, 731 physical layer aspects sidelink and radio access New gNB UE XnAP ID, 132 channel models and protocols, 807-811 New radio (NR), 74, 655, 747, 790 propagation characteristics, Network exposure function (NEF), BS transceiver RF 287-313 characteristics and downlink physical layer Network function repository requirements, 665-691 functions and procedures, function (NRF), 144 direct radio frequency sampling 412-590 Network function virtualization transceivers, 704-716 duplex schemes, 369-379 (NFV), 1, 12-26, 34-37 interface, 14-15 frame structure and architectural aspects, 16-19 large antenna array design and numerology, 383-391 challenges, 14-16 implementation, multiple-access schemes, functional aspects, 19-22 737-744 350-369 legacy support and interworking mmWave transceiver and operating frequency bands, aspects, 24-26 antenna design 379-383 operational aspects, 22-24 considerations, 734-737 time-frequency resources, Network func