Datasets:

eswardivi

/

books

like 0

imum control channel overhead is 21% assuming three symbols and the entire carrier bandwidth used for CORESET, while a more typical overhead is 7% when one-third of the time and frequency New Radio Access Physical Layer Aspects (Part 2) 417 Table 4.1: Various downlink reference signals and their corresponding overhead [73,74]. Reference Description Overhead Signal Type PDSCH DM-RS The DM-RS can occupy 1/3, 1/2 or one full 2.4%-29% OFDM symbol. 1, 2, 3 or 4 symbols per slot can be configured to carry DM-RS. PDSCH PT-RS One resource element in frequency domain 0.2%-0.5% every second or fourth resource block. PT-RS is mainly intended for FR2. CSI-RS One resource element per resource block per 0.25% for eight antenna ports antenna port per CSI-RS periodicity transmitted every 20 ms with 15 kHz subcarrier spacing Two slots with two symbols in each with comb- 0.36% or 0.18% for 20 ms and 40 ms 4 configuration periodicity, respectively and 15 kHz subcarrier spacing resources in the first three symbols of a slot are allocated to PDCCH. The overhead due to the SS/PBCH block is given by the number of SS/PBCH blocks transmitted within the SS/ PBCH block period, the SS/PBCH block periodicity and the subcarrier spacing. Assuming 100 resource blocks across the carrier, the overhead for 20 ms periodicity is in the range of 0.6%-2.3% if the maximum number of SS/PBCH blocks is transmitted [73,74]. 4.1.2.1 Demodulation Reference Signals The main application of DM-RS in NR is to estimate the channel coefficients for coherent detection of the physical channels. In the downlink, the DM-RS is used for channel estima- tion and is subject to the same precoding as PDSCH; thus the (transmit-side) precoding is transparent to the receiver and is viewed as part of the overall channel. There is a trade-off between the channel estimation accuracy and DM-RS density/overhead. If the channel exhi- bits severe frequency-selectivity (i.e., narrower channel coherence bandwidth), the DM-RS density in the frequency-domain should be increased. Similarl

y, if the channel varies faster in time-domain (i.e., shorter channel coherence time), denser DM-RS allocation across time is required. After determining frequency/time-domain DM-RS densities, the DM-RS loca- tions in the time-frequency resource grid should be considered. Assuming stationary channel conditions, uniform DM-RS allocation in both frequency and time-domain is preferred for minimizing interpolation error and reducing implementation complexity. Since no user data is transmitted by DM-RS per se, allocating DM-RS with a proper density is required to maximize the throughput. In NR, a front-loaded DM-RS structure is used as a baseline to achieve low-latency decod- ing (see Fig. 4.2). In the time-frequency resource grid, the front-loaded DM-RS can be located just after the control region, followed by data region. As soon as channel is 418 Chapter 4 Single-symbol Double-symbol Slot duration Slot duration Transmission duration Transmission duration Figure 4.2 Comparison of DM-RS Type A and Type B mappings [6]. estimated based on the front-loaded DM-RS, the receiver can coherently demodulate data in the data region. The front-loaded DM-RS structure is particularly advantageous in decoding-latency reduction for low-mobility scenarios where channel coherence time is lon- ger than the duration of the front-loaded DM-RS. However, allocating only the front-loaded DM-RS can degrade the link performance at higher UE speeds (i.e., channel coherence time becomes shorter). Although the channel information in the data region can be obtained by interpolation, the channel information accuracy diminishes with higher mobility. Therefore, we consider the front-loaded DM-RS patterns with 2 X and 4 X time-domain densities as shown in Fig. 4.3 [70]. To support high-speed scenarios, it is possible to configure up to three additional DM-RS occasions in a slot. The channel estimation in the receiver side can use these additional reference signals for more accurate channel estimation, for example, to perform interpolation between th

e DM-RS occasions within a slot. However, unlike LTE, it is not possible to interpolate channel estimations between slots, or in general different trans- mission occasions, since different slots may be transmitted to different devices and/or in dif- ferent beam directions [14]. New Radio Access Physical Layer Aspects (Part 2) 419 Type 1 single-symbol DM-RS Type 1 double-symbol DM-RS Type 2 single-symbol DM-RS Type 2 double-symbol DM-RS structure structure structure structure Figure 4.3 Various NR PDSCH DM-RS Type A time-frequency patterns [6]. In LoS-dominant channel conditions, the delay spread is expected to be shorter (or equiva- lently the channel coherence bandwidth becomes larger); thus one can consider reducing frequency-domain density of the DM-RS without significant degradation of channel estima- tion precision. By doing so, the overhead due to the DM-RS can be reduced. One example of such low-density DM-RS patterns in frequency-domain is shown in Fig. 4.3. For MIMO trans- mission up to two frequency-domain orthogonal DM-RS ports are supported. The DM-RS is UE-specific, can be beamformed, is confined in the UE scheduled resources, and is transmit- ted only when necessary, both in the downlink and uplink and is used to estimate the commu- nication channel prior to coherent demodulation. To support multi-layer MIMO transmission, multiple orthogonal DM-RS ports can be scheduled, one for each layer. Orthogonality is achieved by means of FDM (comb structure), TDM, and/or CDM (with cyclic shift of the base sequence or orthogonal cover codes) methods. The basic DM-RS pattern is front loaded, as the DM-RS design considers the early decoding requirement to support low-latency appli- cations. For low-speed scenarios, DM-RS uses low density in the time domain. However, for high-speed scenarios, the time density of DM-RS is increased to track fast changes in the 420 Chapter 4 radio channel. The NR defines two time-domain DM-RS structures which differ in the loca- tion of the first DM-RS symbol [14]: Mapping Type A,

where the first DM-RS is located in the second and the third symbol of the slot and the DM-RS is mapped relative to the start of the slot boundary, regardless of where in the slot the actual data transmission occurs. This mapping type is primarily intended for the case where the data occupy (most of) a slot. The reason for the use of the second or the third symbol in the downlink slot is to locate the first DM-RS occasion after a CORESET that is positioned at the beginning of a slot (see Fig. 4.2). Mapping Type B, where the first DM-RS is positioned in the first symbol of the data allocation, that is, the DM-RS location is not given relative to the slot boundary, rather relative to where the data are located. This mapping is intended for transmissions over a small fraction of the slot to support very low latency and other transmissions that cannot wait until a slot boundary starts regardless of the transmission duration. The mapping type for PDSCH transmission can be dynamically signaled as part of the downlink control information (DCI), while for the physical uplink shared channel (PUSCH) the mapping type is semi-statically configured (see Fig. 4.2). The different time-domain locations for (PDSCH) DM-RS mapping types are illustrated in Figs. 4.2 and 4.3, including both single-symbol and double-symbol DM-RS patterns. The purpose of the double-symbol DM-RS is primarily to provide a larger number of antenna ports than what is possible with a single-symbol structure as discussed later. Note that the time-domain location of the DM-RS depends on the scheduled data duration. Multiple orthogonal reference signals can be generated in each DM-RS occasion. Different DM-RS patterns can be configured which are separated in time, frequency, and code domains. The DM-RS has two types: that is, Types 1 and 2, which are distinguished in frequency-domain mapping and the maximum number of orthogonal reference signals. Type 1 can provide up to four orthogonal signals using a single-symbol DM-RS and up to eight orthogonal reference s

ignals using a double-symbol DM-RS, whereas Type 2 can provide 6 and 12 patterns depending on the number of symbols. The DM-RS Type 1 or 2 should not be confused with the mapping Type A or B, since different mapping types can be combined with different reference signal types. Reference signals should preferably have small power variations in the frequency domain to allow a similar channel-estimation quality for all frequencies spanned by the reference signal. Note that this is equivalent to a highly localized time-domain autocorrelation of the transmitted reference signal. New Radio Access Physical Layer Aspects (Part 2) 421 The PDSCH DM-RS sequence r'DM-RS(n) is defined as [`DM-RS(n) = { = {[1 2c(2n) + j[1 - 2c(2n+1)]}/V2, where c(i) is a length-31 Gold sequence generated by the pseudo-random sequence generator defined in [6] and initialized with Cinit latter expression, l denotes the OFDM symbol number within the slot, Nslot is the slot number within a frame, and NSCID = {0,1} is given by the DM-RS sequence initialization field in the DCI associated with the PDSCH transmission, if DCI format 1_1 is used and Nnscid = {0, 1, 65535} is the scrambling identifier when signaled by higher layer parameters scramblingIDC and scramblingIDI; otherwise NSCID = 0 and = Ncell The PDSCH is scheduled by PDCCH using DCI format 1_1 (or DCI format 1_0) with CRC scrambled by C-RNTI, MCS-C- RNTI, or CS-RNTI [6]. The PDSCH DM-RS can be mapped to physical resources in two ways referred to as con- figuration Type 1 or 2, which is determined by RRC parameter dmrs-Type. Prior to resource mapping, the sequence r'DM-RS(m) is multiplied by scaling factor BDMRS H to adjust the trans- mission power and is mapped to resource element (RE) (k,1) as a(k,1) = BPDSCHWf(k)wt(l)rDM-Rs(2n+k) DDMRS where wf(k') and wt(l') are orthogonal cover codes or spreading functions across frequency and time that are given in [6]. In the latter expression, and 1k'=0,1 and k = 4n + 2k + A for configuration Type 1 and + A for configuration Type 2. In DM-RS Type 1,

the underlying pseudo-random sequence is mapped to every other sub- carrier in the frequency domain over the OFDM symbol used for reference signal transmis- sion (see Fig. 4.3). Antenna ports 1000 and 1001 use even-numbered subcarriers in the frequency domain and are separated from each other by multiplying the underlying pseudo- random sequence with different length-2 orthogonal sequences in the frequency domain, resulting in transmission of two orthogonal reference signals for the two antenna ports. If the radio channel can be considered flat across four consecutive subcarriers, the two refer- ence signals will maintain orthogonality at the receiver. Antenna ports 1000 and 1001 are said to belong to CDM group 0, since they use the same subcarriers but are separated in the code-domain using different orthogonal sequences. Reference signals for antenna ports 1002 and 1003 belong to CDM group 1 and are generated in the same way using odd-numbered A Gold sequence is a type of binary sequence that is often used in telecommunication and satellite navigation. Gold sequences have bounded small cross-correlations within a set. A set of Gold sequences consists of 2nd - 1 sequences each with a period of 2nd - 1 A set of Gold sequences can be generated by taking two maximum-length sequences of the same length 2n - 1 such that their absolute cross-correlation is less than or equal to 2(n+2)/2 where n is the size of the linear feedback shift register used to generate the maximum length sequence. The set of 2nd - 1 (logical) exclusive OR of the two sequences in their various phases is a set of Gold codes. The highest absolute cross-correlation in this set of codes is 2nn22)/2 + 1 for even n, and 2(n+1)/2 + 1 for odd n. The exclusive OR of two Gold codes from the same set is another Gold code with arbi- trary phase. Chapter 4 subcarriers and are separated in the code domain within the CDM group and in the fre- quency domain between CDM groups. If more than four orthogonal antenna ports are needed, two consecutive OFDM symbols

are used instead. The aforementioned structure is used over each of the OFDM symbols and a length-2 orthogonal sequence is used to extend the code-domain separation over time, resulting in up to eight orthogonal sequences. The DM-RS Type 2 has a similar structure to Type 1, except some differences with to the number of antenna ports that are supported. Each CDM group for Type 2 consists of two neighboring subcarriers over which a length-2 orthogonal sequence is used to separate the two antenna ports sharing the same set of subcarriers. Four subcarriers are used in each resource block and in each CDM group. Since there are 12 subcarriers in a resource block, up three CDM groups with two orthogonal reference signals can be created using one resource block over one OFDM symbol. If a second OFDM symbol is used along with a length-2 sequence in time-domain, up to 12 orthogonal reference signals can be generated [14]. The location of front-loaded DM-RS symbols, which can be either one or two symbols, is dependent on whether a slot based (DM-RS mapping Type A) or non-slot-based (DM-RS mapping Type B) scheduling is used. In the former type, fixed OFDM symbols regardless of the PDSCH assignment are used to map DM-RS (configurable via parameter lo = {2,3}), whereas in latter type which corresponds to mini-slots, the first OFDM symbol assigned for PDSCH is used to map DM-RS. The reference point for l and the position lo of the first DM-RS symbol depends on the mapping type. Additional DM-RS symbols can be configured (e.g., for high-speed scenarios) as well as for broadcast/multicast PDSCH. In the preceding equations, the reference point for frequency index k depends on PDSCH payload. The reference point for frequency index k is subcarrier 0 of the lowest numbered resource block in CORESET 0, if the corresponding PDCCH is associated with CORESET 0 and Type0-PDCCH common search space and identified by system informa- tion (SI)-RNTI; otherwise, it is the subcarrier 0 in common resource block 0. Furthermore, the reference point

for time index l and the reference position lo of the first DM-RS symbol depends on the mapping type, which for PDSCH mapping Type A, l is defined relative to the start of the slot, that is, lo = 3 if the RRC parameter dmrs-TypeA- Position equals 3; otherwise, lo = 2 and for PDSCH mapping Type B, l is defined relative to the start of the scheduled PDSCH resources, that is, lo = 0. The position of the DM-RS symbols is further dependent on parameter I where for PDSCH mapping Type A, the dura- tion is between the first OFDM symbol of the slot, and the last OFDM symbol of the scheduled PDSCH resources in the slot; and for PDSCH mapping Type B, the duration is the number of OFDM symbols of the scheduled PDSCH resources given by the para- meters specified in [6]. For PDSCH mapping Type B, if the PDSCH duration is 2, 4, or 7 OFDM symbols (i.e., mini-slot scheduling), and if the PDSCH allocation collides with resources reserved for a New Radio Access Physical Layer Aspects (Part 2) 423 CORESET, I is incremented such that the first DM-RS symbol is located immediately fol- lowing the CORESET. If PDSCH duration is 2, 4, or 7 symbols, the UE would not expect to receive a DM-RS symbol beyond the second, third, and fourth symbol, respectively. If one additional single-symbol DM-RS is configured, the UE expects the additional DM-RS to be transmitted on the fifth or sixth symbol when the front-loaded DM-RS symbol is in the first or second symbol, respectively; otherwise, the UE should expect that the additional DM-RS is not transmitted. If PDSCH duration is two or four OFDM symbols, only a single- symbol DM-RS is supported. Furthermore, single-symbol or double-symbol DM-RS is used, if RRC parameter maxLength is equal to 1 or 2, respectively [8,9]. In the absence of CSI-RS configuration, the UE can assume PDSCH DM-RS and SS/PBCH block antenna ports are quasi-co-located with respect to Doppler shift, Doppler spread, aver- age delay, delay spread, and spatial RX 2 parameters. The UE may assume that the PDSCH DM-RS within the same

CDM group are quasi-co-located with respect to Doppler shift, Doppler spread, average delay, delay spread, and spatial RX parameters. Note that the spa- tial RX parameters are meant to describe angular/spatial channel properties at the UE to help the UE select and use one of the beams. The UE can use SS/PBCH block to obtain fre- quency offset, timing offset, Doppler spread, delay spread, and receive beam to process DM-RS. In other words, one can consider the spatial RX parameters as beam indication for the UE, where UE may use the acquired channel parameters from SS/PBCH to receive PDSCH. 4.1.2.2 Phase Tracking Reference Signals The PT-RS was introduced in NR to enable compensation of oscillator phase noise in above-6 GHz frequency bands. Phase noise typically increases as a function of carrier fre- quency. Therefore, PT-RS can be utilized at high carrier frequencies (e.g., mmWave bands) to mitigate the phase noise effect. In the case of OFDM signals, the effect of phase noise is identical phase rotation of all the subcarriers, known as CPE. In NR, the PT-RS is designed To explain the spatial RX concept, let's consider receive antenna diversity where the transmitted signal is received by N antennas which are assumed to have sufficient spatial separation resulting in independent mul- tipath channels between the transmitter and each receiver antenna. Denoting the channel vector, including the amplitude and phase coefficients representing the aggregated effects of multipath propagation, by h = (h1, h2, hn) E CNX1, the received baseband equivalent signal vector reCNX1 can be expressed + n Consequently, by combining the signals from the separate antenna branches with a proper weight- ing, we obtain the combiner output being equal to YRX = WHr = WHh + WH n, where WeCNX1 denotes the weighting vector of the combiner. It is important to note that now the steering vector is replaced with a more generic channel response, including arbitrary amplitude and phase response for each antenna branch. Consequently, the combiner wei

ghts do not match with any particular physical direction anymore, rather they need to be adjusted according to the generic channel response and this form of multi-antenna-based signal combining is referred to as spatial RX processing. The actual weight selection and optimization, in turn, can be implemented by several methods, which differ in complexity and performance. The simplest method is selection combining where only the signal with the highest instantaneous SNR is used for detection. 424 Chapter 4 SO that it has low density in the frequency domain and high density in the time domain, because the phase rotation caused by CPE is identical for all subcarriers within an OFDM symbol; however, it has minimal correlation across OFDM symbols. The PT-RS is UE-specific, confined in a scheduled resource, and can be beamformed. The number of PT-RS ports can be lower than the total number of ports, and orthogonality between PT-RS ports is achieved by means of frequency-division multiplexing. The PT-RS is configurable depending on the quality of the oscillators, carrier frequency, OFDM subcarrier spacing, and modulation and coding schemes used for transmission. The PT-RS introduced in NR is used for time and frequency tracking as well as estimation of delay spread, and Doppler spread at the UE side. They are transmitted in a confined bandwidth for a configurable time duration controlled by RRC parameters. The time-fre- quency structure of PT-RS depends on the waveform. For OFDM, the first reference symbol (prior to applying any orthogonal sequence) in a PDSCH/PUSCH allocation is repeated every LPT-RS E {1, 4} symbol, starting with the first OFDM symbol in the allocation. The repetition counter is reset at each DM-RS position since there is no need for PT-RS inser- tion immediately following a DM-RS occasion. In the frequency domain, PT-RS are trans- mitted in every second or fourth resource block, resulting in a sparse frequency-domain structure. The density in the frequency domain is dependent on the scheduled bandwidt

h in a sense that the higher the bandwidth, the lower the PT-RS density. For the smallest band- widths, no PT-RS is transmitted. To reduce the risk of collision between PT-RS associated with different devices scheduled on overlapping frequency-domain resources, the subcarrier number and the resource blocks used for PT-RS transmission are determined by the C- RNTI of the device. The antenna port used for PT-RS transmission is given by the lowest numbered antenna port in the DM-RS antenna port group. An example time-frequency PT- RS structure is shown in Fig. 4.4. The PT-RS for subcarrier k is given by r PT-RS(k) = I'DM-RS(2m + K') where is the DM-RS at time-domain position lo and subcarrier k. The PT-RS is present only in the resource blocks used for the PDSCH, and only if there is an explicit indication of their pres- ence. In that case, the PT-RS is scaled by a factor of BPT-RS(i) to adjust the transmission power. The PT-RS is mapped to resource elements a(k, 1) = if l is located within the OFDM symbols allocated for the PDSCH transmission and the designated resource element is not used for DM-RS, CSI-RS, SS/PBCH block, PDCCH, or is declared as not available. The time indices l at which PT-RS are allocated are defined relative to the start of the PDSCH allocation lo and are given by l = lo + iLPT-RS where i is incremented as long as the PT-RS occasion falls inside the PDSCH allocation and the aforementioned conditions are met. For PT-RS resource mapping, the resource blocks allocated for PDSCH transmission are numbered from 0 to NRB - 1 from the lowest scheduled resource block to the highest. The corresponding subcarriers in this set of resource blocks are numbered in increasing order start- ing from the lowest frequency to NRB NRB 1. The subcarrier indices that the PT-RS are New Radio Access Physical Layer Aspects (Part 2) 425 PDSCH data PDSCH PDSCH data PT-RS on every OFDM symbol PT-RS on every second OFDM symbol PT-RS on every fourth OFDM symbol PT-RS on every second resource block PT-RS on every fourth resou

rce block Scheduled bandwidth Figure 4.4 Illustration of PT-RS structures in time and frequency domain [6]. mapped to are given by k=kRE + where mod KPT-RS if NRB mod KPT-RS 0; otherwise, mod(NRB mod KPT-RS) [6]. In the latter equation, NRNTI denotes the RNTI associated with the DCI scheduling of the transmission; KPT-RS E {2,4}, and is the DM-RS port associated with the PT-RS port. The density of PT-RS in time and frequency domain is configurable to address different scenarios (e.g., different carrier frequency, modulation and coding scheme, and hardware quality). The PT-RS patterns in time and frequency domains are illustrated in Fig. 4.4. 4.1.2.3 Channel State Information Reference Signals The CSI-RS in NR is used for downlink CSI estimation. It further supports RSRP measure- ments for mobility and beam management (including analog beamforming), time/frequency tracking for demodulation, and uplink reciprocity-based precoding. The CSI-RS is UE-specific; nevertheless, multiple users can share the same CSI-RS resource. The NR defines zero-power and non-zero-power CSI-RS. When a zero-power CSI-RS is configured, the resource elements (designated to CSI-RS) are not used for PDSCH transmission. In this case, the zero-power CSI-RS is used to mask certain resource elements, making them unavailable for PDSCH mapping. This masking not only supports transmission of UE- specific CSI-RS, but also the design allows introduction of new features while maintaining backward compatibility. The NR supports flexible CSI-RS configurations. A CSI resource can be configured with up to 32 antenna ports with configurable density. In the time domain, a CSI-RS resource may start at any OFDM symbol of a slot and span 1, 2, or 4 426 Chapter 4 CDM2 pattern CDM4 pattern CDM8 pattern (OFDM symbols) Table 2-Row10-FD2/TD2/CDM4 12 Ports Table 2-Row11-FD/CDM2 16 Ports Table 2-Row13-FD/CDM2 24 Ports Figure 4.5 Example locations of CSI-RS in time and frequency [6]. OFDM symbols depending on the number of configured antenna ports. The CSI-RS can be p

eriodic, semi-persistent, or aperiodic (DCI triggered). When used for time frequency track- ing, the CSI-RS can be periodic or aperiodic. In this case, a single port is configured, and the signal is transmitted in the form of bursts over two or four symbols that are spread over one or two slots [71]. A configured CSI-RS resource may correspond to up to 32 different antenna ports. In NR, a CSI-RS is always configured on a per-device basis. It must be noted that the UE-specific configuration of CSI-RS does not necessarily mean that a transmitted CSI-RS can only be used by a single device, rather the same set of CSI-RS resources can be separately config- ured for multiple devices, which means that a single CSI-RS can be shared among multiple devices. As illustrated in Fig. 4.5, a single-port CSI-RS occupies a single resource element within a resource block in the frequency domain and one slot in the time domain. While the CSI-RS can be configured to occur anywhere within the resource block, in practice there are some restrictions on CSI-RS resource assignment to avoid collisions with other down- link physical channels and signals. The transmission of a configured CSI-RS is expected not to collide with a CORESET configured for the device; the DM-RS associated with PDSCH transmissions scheduled for the device; and the SS blocks transmissions. A multiport CSI-RS can be viewed as multiple orthogonal per antenna-port CSI-RS sharing the set of resource elements assigned for transmission of the configured multiport CSI-RS. In general, the resource sharing is achieved through a combination of code-domain New Radio Access Physical Layer Aspects (Part 2) 427 (i.e.,different per antenna-port CSI-RS are transmitted on the same set of resource elements with separation achieved by spreading the CSI-RS with different orthogonal codes), frequency-domain (i.e., different per antenna-port CSI-RS are transmitted on different sub- carriers over an OFDM symbol), and time-domain (i.e., different per antenna-port CSI-RS are transmitted on

different OFDM symbols within a slot) multiplexing schemes. As illus- trated in Fig. 4.5, code-division multiplexing between different (per antenna-port) CSI-RS can be performed across frequency by spreading over two adjacent subcarriers (CDM2) to support two antenna ports; across frequency and time by spreading over two adjacent sub- carriers and two adjacent OFDM symbols (CDM4) to enable four antenna ports; across fre- quency and time by spreading over two adjacent subcarriers and four adjacent OFDM symbols (CDM8) to support up to eight antenna port transmission. Combination of code/fre- quency and time-division multiplexing can be used to configure different multiport CSI-RS structures where, in general, an N-port CSI-RS occupies N resource elements within a resource block or a slot. When CSI-RS supports more than two antenna ports, there are mul- tiple CSI-RS patterns/structures based on different combinations of CDM, TDM, and FDM that can be utilized as shown in Table 4.2. The non-zero-power CSI-RS is mathematically represented by sequence rcci-rs(m) which is defined as r'CSI-RS(m) = 2c(2m)] j[1 - 2c(2m+1)]}/v2 where c(i) is a length-31 Gold sequence generated by the pseudo-random sequence generator defined in [6] and initialized with Cinit=[210(N5)mbol -1) nID] mod 231 at the start of each OFDM symbol. In the latter equation, Nslot is the slot number within a radio frame, l is the OFDM symbol number within a slot, and NID is set by the RRC parameter scramblingID or sequenceGenerationConfig [6,13]. If CSI-RS is configured, the CSI-RS sequence r CSI-RS(m) is mapped to resources elements (k,l) such that a(k,l) = where m' = _na] k + a=por2p = when Np = 1 or Np > 1, respec- tively, provided that the resource element (k,l) is within the resource blocks designated to the CSI-RS resource for which the UE is configured. The reference point for k = 0 is subcarrier 0 in common resource block 0. The value of density P and the number of antenna ports are given by RRC parameters [6,13]. The scaling factor BCSI-RS has a

non-zero value for a non- zero-power CSI-RS to ensure that the power offset specified by the RRC parameter powerControlOffsetSS is satisfied. Other parameters k', l', wf(k'), and wt(l') are given in Table 4.2 where each pair (k,l) corresponds to a CDM group of size 1 (no CDM) or size 2, 4, or 8. The indices k' and l' are used to index the resource elements within a CDM group [6,8]. The time-domain locations lo = {0, 1, 13} and = {2, 3, 12} are defined relative to the start of a slot with the starting positions of a CSI-RS in a slot configured by the RRC para- meters provided in CSI-RS-ResourceMapping information element. The frequency-domain location of CSI-RS is determined by a bitmap signaled via the RRC parameter provided in Table 4.2: Various CSI-RS patterns (locations) within a slot [6]. Number of CSI-RS CDM Type (k,1) CDM Group Index j Ports Np Density p No CDM (ko,1o),(ko+4,1o),(ko+8,0) 0,0,0 1,0.5 No CDM (ko,lo) 1,0.5 FD-CDM2 (ko, 10) FD-CDM2 (ko,lo), (ko + 2,10) FD-CDM2 (ko,lo), ,(ko,lo+1) FD-CDM2 (ko, 10), (k1, (k2,lo), (k3, 10) 0,1,2,3 FD-CDM2 (ko,lo), (k1,10), (ko,1 + 1), (k1,10 + 1) 0,1,2,3 (ko,lo),(k1,10) (FD2,TD2) FD-CDM2 (ko,lo), (k1,10) (k2,10),(k3,10), (k4,lo), (k5,lo) 0,1,2,3,4,5 (ko,lo), (k1,10), (k2,10) 0,1,2 (FD2,TD2) 1,0.5 FD-CDM2 0,1,2,3,4,5,6,7 (ko,1 + 1), (k1,10 + 1), (k2,lo + 1), (k3, 10 +1) 1,0.5 (ko,lo), (k1,lo), (k2,lo), (k3, /0) 0,1,2,3 (FD2,TD2) (ko,lo), (k1,10) (k2,lo), (ko,10+1),(k1,10+ 1), (k2,lo + 1) 1,0.5 FD-CDM2 0,1,2,3,4,5,6,7,8,9,10,11 (ko,/1), (k1,/1), (k2,/1), (ko,/1 +1), (k1,/1 + 1), (k2,/1 + 1) 1,0.5 (ko,lo), (k1,10), (k2,10), (ko,/1), (k1,/1) (k2,/1) 0,1,2,3,4,5 (FD2,TD2) 1,0.5 (ko,10),(k1,10),(k2,10) 0,1,2 0,1,2,3 (FD2,TD4) (ko, 10 + 1), (k1,10 +1), (k2,10 + 1), (k3,10 + 1) 1,0.5 FD-CDM2 0,1,2,3,4,5,6,7, (kg,/1),(k1,1)),(k2,19), (k3,/1), 18,9,10,11,12,13,14,15 (ko,/1 + 1), (k1,/1 + 1), (k2,/1 + 1), (k3,/1 + 1) (ko,1o),(k1,1o),(k2,1o),(k3, 10), 1,0.5 0,1,2,3,4,5,6,7 (ko,/1), (k1,/1), (k2,/1), (k3,/1) (FD2,TD2) 1,0.5 (ko,lo), (k1,lo), (k2,lo), (k3, 10) 0,1,2,3 0,1,2,3 (

New Radio Access Physical Layer Aspects (Part 2) 429 CSI-RS period = 5 slots, offset = 0 CSI-RS period = 5 slots, offset = 3 slots CSI-RS period= 10 slots, offset = 3 slots Time (Slots) Figure 4.6 Example CSI-RS periodicity and offset [6]. CSI-RS-ResourceMapping information element [6,13]. The starting position and number of the resource blocks in which the CSI-RS is transmitted are provided via RRC signaling. The CSI-RS is transmitted on antenna ports p=3000+s+jLcDM where j=0,1,...,Np/Lcom 1 and s=0,1, LCDM 1. In the latter expression, LCDM e{1,2,4,8} is the CDM group size and Np is the number of CSI-RS antenna ports. The CDM groups are numbered in order of increasing frequency-domain allocation first and then increasing time domain-allocation. For a CSI-RS resource configured as periodic or semi-persistent by the RRC parameter resourceType, the CSI-RS is transmitted in slot num- bers satisfying (Nslot framenslot TCSI-RS where the CSI-RS periodicity TCSI-RS (in number of slots) and slot offset Toffset are signaled by the RRC parameter CSI-ResourcePeriodicityAndOffset or slotconfig (see Fig. 4.6). The CSI-RS is transmitted in a slot only if, all OFDM symbols of that slot corresponding to the configured CSI-RS resource are designated for downlink transmission. The antenna ports within a CSI-RS resource are quasi-co-located with quasi-co-location (QCL) Type A, Type D (when applica- ble). In summary, the NR supports periodic, aperiodic, semi-persistent CSI-RS transmission. The NR CSI-RS patterns can be mapped to 1, 2, or 4 OFDM symbols and support CDM2, CDM4, CDM8 spreading functions, for example, CDM8 means that there are eight spread- ing functions wf(k') and wt(l'). For CSI acquisition, the NR supports CSI-RS density =0.5 and 1 RE/RB/port and a PRB-level comb-type transmission. The number of antenna ports can be independently configured for periodic, aperiodic, semi-persistent CSI reporting. A CSI-RS resource configuration up to 32 ports is supported in NR. The UE-specific CSI-RS may be configured to support wide

band CSI-RS and partial band CSI-RS. In order to reduce beam management overhead and latency, the NR supports subtime units of less than one OFDM symbol in a reference numerology [73]. Each CSI-RS resource is configured by the RRC parameter NZP-CSI-RS-Resource. The time-domain locations of the two periodic CSI-RS resources in a slot or four periodic CSI-RS resources in two consecutive slots are given by le {4, 8}, le {5,9}, or le {6, 10} for FR1 and FR2; or le {0, 4}, le {1,5}, le {2, 6}, le {3, 7}, le{7,11}, le{8,12} or le{9,13} for FR2. A single-port 430 Chapter 4 CSI-RS resource with density P = 3 (see Table 4.2) and RRC parameter density is configured by CSI-RS-ResourceMapping. The bandwidth of the CSI-RS resource, given by RRC parameter freqBand configured by CSI-RS-ResourceMapping, is determined as min(52, NBWP(i) RB resource blocks or is equal to RB BWP(i) resource blocks. The UE is not expected to be configured with the periodicity of 2u X 10 slots, if the bandwidth of CSI-RS resource is larger than 52 resource blocks. The periodicity and slot offset, given by RRC parameter periodicityAndOffset configured by NZP-CSI-RS-Resource parameter, is one of 2"Xp slots where Xp = 10,20,40, or 80 [9]. It should be noted that the property of periodic, semi-persistent, or aperiodic is not a property of the CSI-RS per se, rather the property of a CSI-RS resource set. As a result, activation/deactivation and triggering of semi-persistent and aperiodic CSI-RS must be done for a set of CSI-RS within a resource set. In the case of periodic CSI-RS transmission, the UE can assume that a configured CSI-RS transmission occurs every Nth slot, where N ranges from 4 to 640. In addition to the periodicity, the device is also configured with a specific slot offset for the CSI-RS transmission. In the case of semi-persistent CSI-RS transmission, certain CSI-RS periodicity and slot offset are con- figured similar to periodic CSI-RS transmission. However, the CSI-RS transmission can be activated or deactivated via MAC control elements.

Once the CSI-RS transmission has been activated, the device can assume that the CSI-RS transmission will continue according to the configured periodicity until it is deactivated. Similarly, once the CSI-RS transmission has been deactivated, the device can assume that there will be no CSI-RS transmission according to the configuration until it is reactivated. In the case of aperiodic CSI-RS, no periodicity is configured, rather the UE is triggered via signaling in the DCI [9]. The CSI-RS may be further used for RSRP measurements3 and mobility management since NR does not include the cell-specific reference signals that were used in LTE for mobility management. The set of CSI-RS corresponding to a set of beams on which measurements are conducted should be included in the non-zero power (NZP)-CSI-RS resource set associ- ated with the report configuration. Such a resource set may either include a set of config- ured CSI-RS or a set of SS blocks. Measurements for beam management can be carried out on either CSI-RS or SS block. In the case of L1-RSRP measurements based on CSI-RS, the CSI-RS should be limited to single-port or dual-port CSI-RS. In the latter case, the reported L1-RSRP should be a linear average of the L1-RSRP measured on each port. The device can report measurements corresponding to up to four reference signals (CSI-RS or SS blocks), that is, up to four beams, in a single reporting instance. Each report is related to up to four reference signals or beams and includes the measured L1-RSRP for the strongest CSI-RSRP is defined as the linear average over the power contributions (in Watts) of the resource elements that carry CSI-RS configured for RSRP measurements within the identified measurement frequency region in the con- figured CSI-RS occasions. The CSI reference signals are transmitted on specific antenna ports. This measurement is applicable for connected mode only for both intra- and inter-frequency measurements. New Radio Access Physical Layer Aspects (Part 2) 431 beam and the difference between

other beams' L1-RSRP measurements and the measured L1-RSRP of the best beam [14]. If a UE is configured with an NZP-CSI-RS-ResourceSet via RRC parameter repetition set to "on," it may assume that the CSI-RS resources within the NZP-CSI-RS-ResourceSet are trans- mitted with the same downlink spatial domain transmission filter, where the CSI-RS resources in the NZP-CSI-RS-ResourceSet are transmitted on different OFDM symbols. If repetition is set to "off," the CSI-RS resources within the NZP-CSI-RS-ResourceSet are transmitted with the same downlink spatial domain transmission filter. If the UE is configured with a CSI-ReportConfig and parameter reportQuantity is set to "cri-RSRP," or "none" and the CSI-ResourceConfig for channel measurement (RRC parameter resourcesForChannelMeasurement) contains a NZP-CSI-RS-ResourceSet that is configured with the higher layer parameter repetition and without the higher layer parameter trs-Info, the UE can only be configured with the same number (1 or 2) of ports with the higher layer param- eter nrofPorts for all CSI-RS resources within the set. If the UE is configured with the CSI-RS resource on the same OFDM symbol(s) as an SS/PBCH block, the CSI-RS and the SS/PBCH block are quasi-co-located with QCL TypeD, if applicable. Furthermore, the UE will not be configured with the CSI-RS in PRBs that overlap with those of the SS/PBCH block, and the same subcarrier spacing is used for both the CSI-RS and the SS/PBCH block [9]. 4.1.2.4 Tracking Reference Signals A UE must track and compensate time and frequency variations of its local oscillator in order to successfully receive downlink transmissions. The problem is exacerbated in higher radio frequencies. To assist the device in this task, a tracking reference signal can be config- ured. The TRS is not a CSI-RS, rather a TRS is a resource set consisting of multiple peri- odic NZP-CSI-RS. More specifically, a TRS consists of four single-port, density-3 CSI-RS located within two consecutive slots as shown in Fig. 4.7. The CRS-RS within the

resource TRS period: TRS 10/20/40/80 (OFDM symbols) Figure 4.7 Example TRS structure (four single-port, Density-3 CSI-RS over two consecutive slots [9]). 432 Chapter 4 set or the TRS can be configured with a periodicity of 10, 20, 40, or 80 ms. It must be noted that the exact set of time-frequency resource elements used for the TRS may vary; however, the two CSI-RS within a slot are always separated by four symbols in the time domain. This time-domain separation sets a limit for the maximum frequency error that can be com- pensated. Likewise, the frequency-domain separation of four subcarriers sets a limit for the maximum timing error that can be compensated. There is an alternative TRS structure with the same per-slot structure as the TRS structure shown in Fig. 4.7 with only two CSI-RS within a slot, compared to two consecutive slots for the TRS structure shown in the figure. In LTE, the CRS served the same purpose as the TRS; however, the TRS has relatively lower overhead, has one antenna port, and only present in two slots in every TRS period. As we mentioned earlier, the CSI-RS for tracking or TRS is a special configuration of CSI-RS which is specifically configured for a UE. The TRS is used for fine time and fre- quency tracking as well as path delay spread and Doppler spread estimation. A UE in RRC_CONNECTED mode would receive information on a CSI-RS resource set which is configured specifically for the purpose of time/frequency tracking. In that case, the UE is configured with RRC parameter trs-Info and will assume that TRS is transmitted from the antenna port with the same port index of the configured NZP CSI-RS resources in the CSI-RS resource set. In frequency range 1, the UE may be configured with a CSI-RS resource set consisting of four periodic CSI-RS resources in two consecutive slots with two periodic CSI-RS resources in each slot, whereas in frequency range 2, the UE may be con- figured with a CSI-RS resource set of two periodic CSI-RS resources in one slot or with a CSI-RS resource set of four

periodic CSI-RS resources in two consecutive slots with two periodic CSI-RS resources in each slot. The periodic CSI-RS resources in the CSI-RS resource set configured with RRC parameter trs-Info have the same periodicity, bandwidth and subcarrier location. The time-domain location of the TRS is determined by two periodic CSI-RS resources in a slot, or four periodic CSI-RS resources in two consecutive slots and via RRC signaling. The density of the TRS in one physical resource block is three REs per symbol. The TRS can be time-division multiplexed with the synchronization signal/PBCH block (SSB). 4.1.3 Control Channels 4.1.3.1 Physical Broadcast Channel In order to select a PLMN and camp on a cell, a UE must perform a cell search in the sup- ported frequency bands. The procedure requires the UE to achieve time and frequency synchronization with a specific cell. This enables decoding of PBCH which carries the MIB containing the critical system information necessary to decode transmissions on PDSCH. In NR, the SI is divided into minimum SI and other SI. The minimum SI is peri- odically broadcast and comprises basic information required for initial access and New Radio Access Physical Layer Aspects (Part 2) 433 Minimum system information (MIB+RMSI) always present and broadcast periodically Other system information (OSI) optionally present and broadcast periodically System information request On-demand other system information broadcast or dedicated signaling Figure 4.8 Transmission of system information in NR [11]. information for acquiring other SI broadcast periodically or scheduled on-demand. The other SI encompasses everything else not broadcast in the minimum SI message and may be either broadcast or individually transmitted to the UE. In the latter case, the on- demand transmission of other SI can be triggered by the network or based on a request from the UE. The change of SI can only occur at specific radio frames. Note that to ensure coverage and reliability, the SI may be transmitted a number of times with

the same content within a modification period. For the minimum SI delivery, part of mini- mum SI is transmitted in PBCH. The remaining minimum SI (RMSI) is transmitted in the downlink shared channel. The initial BWP information is signaled by PBCH which con- tains the CORESET and PDSCH information for mapping the RMSI. Fig. 4.8 shows the process for transmission of various components of system information. Unlike LTE sys- tem where the minimum SI and a group of SI blocks were broadcast periodically, NR lim- its the amount of SI that is periodically broadcast and instead relies on less-frequent and on-demand transmission of the non-essential SI. The MIB message (in ASN.1 format), which is carried in PBCH, consists of the following components [13]: MIB : : = SEQUENCE { systemFrameNumber BIT STRING (SIZE (6) ) , subCarrierSpacingCommon ENUMERATED {scs15or60, scs30or120}, ssb-SubcarrierOffset INTEGER (0..15), (Continued) 434 Chapter 4 (Continued) dmrs-TypeA-Position ENUMERATED {pos2, pos3}, pdcch-ConfigSIB1 cellBarred ENUMERATED {barred, notBarred}, intraFregReselection ENUMERATED {allowed, notAllowed}, spareBIT STRING (SIZE (1)) PDCCH-ConfigSIB1 : : = SEQUENCE { controlResourceSetZero Control ResourceSetZero, searchSpaceZero SearchSpaceZero In the MIB message which is mapped to BCH logical channel, subCarrierSpacingCommon parameter indicates the subcarrier spacing for SIB1, Msg2/4 for the initial access and the SI messages where values 15 and 30 kHz are applicable to sub-6 GHz and values 60 and 120 kHz are applicable to carrier frequencies above 6 GHz; ssb-subcarrierOffset is the frequency-domain offset between SSB and the overall resource block grid in number of sub- carriers; dmrs-TypeA-Position indicates the position of the first downlink DM-RS; and pdcch-ConfigSIBI determines the bandwidth of PDCCH/SIB1 or the size of the CORESET containing common search space for PDCCH. In other words, the first field of pdcch- ConfigSIB1 determines the common CORESET corresponding to the initial downlink BWP and the second fie

ld identifies the common search space of initial downlink BWP [13]. The RMSI is transmitted via the PDSCH by downlink assignment in an RMSI CORESET. The concept of CORESET was introduced in NR to identify a set of time-frequency resources consisting of multiple resource blocks in the frequency domain and one to three OFDM symbols in the time domain. The NR enables UE to be configured with multiple CORESETs, and each CORESET is associated with a UE-specific configured resource map- ping scheme. The PBCH payload size is 56 bits including 24-bit CRC. In NR, PBCH uses a single- antenna port transmission scheme, using the same antenna port as PSS and SSS within the same SS block. The periodicity of PBCH is 80 ms. The MIB data arrive at the PBCH pro- cessing unit in the form of one transport block (TB) every 80 ms and goes through the fol- lowing steps as shown in Fig. 4.9: payload generation, scrambling, TB CRC calculation and attachment, channel coding, and rate matching. The coded and modulated bits of PBCH are New Radio Access Physical Layer Aspects (Part 2) 435 MIB bits BCH data 1stscrambling stage with PBCH payload generation Scrambling TS 38.212 Section 7.1.1 TS 38.211 Section 7.3.3.1 cell ID m = SFN/2 mod 4 Scrambled bits + CRC scrambling Modulation TS 38.212 Section 7.1.2 TS 38.211 Section 7.3.3.2 Polar coding CRC attachment Resource element mapping TS 38.212 Section 7.1.3 TS 38.211 Section 7.3.3.3/7.4.3 Coded Bits bo,b,, Channel coding 2nd scrambling stage with cell ID TS 38.212 Section 7.1.4 n = (SSB time index) mod 4 n = (SSB time index) mod 8 do,d1,...,NN-1 Scrambled bits Rate matching TS 38.212 Section 7.1.5 fo,f1,...,FE-1 Figure 4.9 Physical layer processing of PBCH and mapping to time-frequency resources [5]. mapped onto the resource elements allocated for PBCH. As we will discuss later, the PBCH content is encoded using the polar code. Two scrambling operations are performed on PBCH which include one before CRC attachment and another one after the polar coding and rate matching. In the first scrambling

stage, initialization based on Cell ID, the sequence is partitioned into four non-overlapping portions. The portion for transmission is selected based on the second and third least significant bits of the SFN. In the second scrambling stage, initialization based on Cell ID, the sequence is partitioned into four or eight non-overlapping portions. The portion for transmission is selected based on the second or third least significant bits of the SS block time index. The physical layer processing of the PBCH is shown in Fig. 4.9. We denote the MIB bits in a TB delivered to the physical layer by ao, a1,..., ANMIB-1' where NMIB = 32 bits is the payload size of the MIB. The lowest order information bit ao is mapped to the most significant bit of the TB payload. Additional timing-related PBCH payload bits ANMIB+7 regen- erated based on the least significant bits of the SFN, half frame, SS block index, and com- bined with the MIB payload (see Fig. 4.10). The NMIB + 8 bits are interleaved according to an interleaving pattern specified in [7] prior to the first scrambling stage. In the first scrambling Chapter 4 BCH Data NMIB-1' 4th, 3rd, 2nd, and the 1st LSB of SFN Half radio frame bit if Lssg = 64 6th, 5th, and 4th bits of SS/PBCH block index ANMIB+5 =5th and 4th bits of SS/PBCH block index Interleaving Figure 4.10 NR PBCH payload generation [7]. stage, the input bits to the scrambling unit ai are scrambled according to = = 0, 1, ,NMIB + 7 where Si is a sequence that is derived from a generic pseudo-random length-31 Gold sequence, which is initialized with Cinit = Ncell at start of each SFN whose value satisfies SFN mod 8 = 0 The Si further depends on the half radio frame index, and the second and third least significant bits of the system frame number [7]. A CRC is calculated for the purpose of error detection on the entire BCH payload. The input bit sequence is denoted by and the parity bits by P0,P1,..., where A = NMIB + 8 is the payload size and L = 24 is the number of CRC parity bits. The parity bits are calculated

and attached to the BCH payload using the generator polynomial gCRC24C' (D) = D24 + D23 D21 + D20 + D17 + D 15 + D13 + D 12 + D8+D4+ D2 + D + resulting in the sequence bo,b1,..., bB-1, where B = A + L. The information bits that are then delivered to the channel coding block are denoted by CO,C1,..., CK-1, where K is the number of input bits. They are encoded using a polar encoder by setting the parameters nmax =9,11L = 1, NPC = 0, and nwm=0. = Detailed PBCH channel encoding and decoding block diagram is shown in Fig. 4.11. The output of polar encoder is denoted by do,d1,...,NN-1 where N is the number of coded bits. The input sequence to the rate matching function is do, d1, dn and the output bit sequence after rate matching is denoted by where the rate matching output sequence length is E : 864 and the rate matching is per- formed by setting the parameter IBIL to zero. The preceding polar coding and rate matching parameters will be explained in Section 4.1.7.1 [7]. New Radio Access Physical Layer Aspects (Part 2) 437 Higher layer parameters PBCH payload PBCH payload PBCH payload CRC24C Frozen bit Subblock generation scrambling attachment insertion Polar encoding Rate matching Multiplexing interleaving interleaving Interleaving onto PBCH PBCH encoding procedure Higher layer Determination parameters of known bits PBCH payload PBCH payload PBCH payload Subblock Distributed CRC-aided successive cancellation list Demultiplexing retrieval deinterleaving descrambling deinterleaving dematching from PBCH PBCH decoding procedure Figure 4.11 NR PBCH channel encoding/decoding block diagram [7,35]. The block of bits bo, b ., bNPBCH-1, where NPBCH denotes the number of bits transmitted on the PBCH, is scrambled prior to modulation, resulting in a block of scrambled bits bNPBCH -1 in which b(i) [b(i) vNPBCH)] mod 2 and c(i) denotes a generic pseudo-random length-31 Gold sequence. The scrambling sequence is initialized with = Ncell at the start of each SS/PBCH block. The parameter V is the two least significant bits of the SS/P

BCH block index when Lmax = 4 and the three least significant bits of the SS/PBCH block index when Lmax = 8 or 64 where Lmax denotes the maximum number of SS/PBCH blocks in an SS/PBCH period for a particular band (see Section 4.1.4.3). block scrambled bits bo, b1, bNPBCH~1 is QPSK modulated, resulting in a block of complex-valued modulation symbols dPBCH(0), dPBCH(1), dpBCH(Nsymb mapping of the modulated symbols to the physical resources is described in Section 4.1.4.3. The PBCH exploits a special type of DM-RS that is used for coherent detection and decoding of PBCH [6]. The total number of resource elements used for PBCH transmission per SS block is 576. Note that this number includes the resource elements for PBCH and the resource elements for the DM-RS needed for coherent demodulation of PBCH. Different numerolo- gies can be used for SS/PBCH block transmission. However, to limit the need for devices to simultaneously search for SS/PBCH blocks of different numerologies, in many cases only a single SS block numerology is defined for a given frequency band. The DM-RS sequence r PBCH (m) for an SS/PBCH block is defined by rppch(m) = ([1 - 2c(2m) +j[1 - 2c(2m 1)]/2) where c(n) is a length-31 Gold sequence generated by the pseudo-random sequence generator defined in [6] and initialized at the start of each SS/PBCH block with Cinit = (iSSB + 26 (issb + (Ncell ID mod 4) [6]. L = 4, iSSB = iSSB 4nhf where Nhf denotes the number of the half-frame in which PBCH is transmitted in a frame with Nhf = 0 for the first half-frame in the frame and Nhf = 1 for the second half-frame in the frame, and iSSB is the two least significant bits of the SS/PBCH block index. In the case that L = 8 or L = 64, Nhf = 0, and iSSB = iSSB are the three least sig- nificant bits of the SS/PBCH block index. Note that L denotes the maximum number of SS/ PBCH block beams in an SS/PBCH block period for a particular band [3]. 438 Chapter 4 sequence of complex-valued QPSK-modulated symbols dPBCH(0), dPBCH (1), , dpBCH(Nsymb - 1) containing the PBCH in

formation are scaled by a factor BPBCH to adjust the PBCH transmit power and then mapped in sequence starting with dPBCH(0) to resource elements (k,1) provided that they are not used for PBCH DM-RSs (see Fig. 4.12). The sequence of complex-valued symbols rppch(0), rppch(1), ., rPBCH (143) con- taining the DM-RSs for the SS/PBCH block is scaled by a factor of PBCH in order to DM-RS 4 OFDM symbols PBCH DM-RS Figure 4.12 Structure of PBCH in time/frequency domain 30]. New Radio Access Physical Layer Aspects (Part 2) 439 Table 4.3: Resource mapping within an SS/PBCH block for PSS, SSS, PBCH, and DM-RS for PBCH [6]. Physical Channel/Signal OFDM Symbol Number / Subcarrier Number k Relative to the Start of an Relative to the Start of an SS/PBCH Block SS/PBCH Block 56,57,...,182 56,57,..., 182 155,183,184,.. 48,4 1,55,183,184 0,1,...,239 0,1,...,47,192,193,...,239 DM-RS for PBCH 0+v,4+v,8+v,...,236+ 0+v,4+v,8+v,...,44+ 192 + v, 196 + V,..., 236 + V adjust the PBCH DM-RS transmit power. They are then mapped to resource elements (k,l) in increasing order of first k (frequency index) and then l (time index), within one SS/ PBCH block. As shown in Table 4.3, the location of PBCH DM-RS is dependent on param- eter V = Ncell mod 4 and is shifted in frequency with different Ncell ID values (see Fig. 4.13 for an example). 4.1.3.2 Physical Downlink Control Channel The data transmission in NR in downlink/uplink direction is generally controlled via MAC scheduling. Each device monitors a number of PDCCHs, typically once per slot, although it is possible to configure more frequent monitoring to support traffic requiring very low latency. Upon detection of a valid PDCCH, the device follows the scheduling decision and receives (or transmits) one unit of data, known as a transport block. The PDCCHs are trans- mitted in one or more CORESETs each of length one to three OFDM symbol(s). Unlike LTE, where control channels span the entire carrier bandwidth, the bandwidth of a CORESET can be configured. In NR, a flexible slot format can be con

figured for a UE by cell-specific and/or UE-specific higher layer signaling in a semi-static downlink/uplink assignment manner, or by dynamically signaling via DCI in group-common PDCCH (GC- PDCCH). When the dynamic signaling is configured, a UE should monitor GC-PDCCH which carries dynamic slot format indication (SFI). When a device enters the connected state, it has already obtained the information from PBCH about the CORESET where it can find the control channel used to schedule the RMSI. The CORESET configuration obtained from PBCH also defines and activates the initial bandwidth part in the downlink. The initial active uplink bandwidth part is obtained from the SI scheduled using the downlink PDCCH. 440 Chapter 4 v=0 236 v=2 236 PBCH DM-RS Figure 4.13 PBCH DM-RS location shift by physical cell ID 30]. In NR, the device typically attempts to blindly decode candidate PDCCHs transmitted from the network using one or more search spaces. However, there are some differences com- pared to LTE based on the different design targets for NR as well as the experience learned from LTE deployments. Unlike LTE PDCCH, the PDCCH in NR does not span the entire carrier bandwidth. This is due to the fact that NR devices may not be able to operate over New Radio Access Physical Layer Aspects (Part 2) 441 the full carrier bandwidth of the gNB. The PDCCH in NR is designed to support UE- specific beamforming, which is the result of beam-centric physical-layer design of NR and a requirement when operating in mmWave bands with challenging link budgets. 4.1.3.2.1 Structure and Physical Layer Processing of PDCCH A UE-specific PDCCH is used to schedule downlink and uplink transmissions on PDSCH and PUSCH, respectively. The DCI on PDCCH contains downlink assignments including modulation and coding format, resource allocation, and HARQ information related to DL-SCH; uplink scheduling grants including modulation and coding format, resource allo- cation, and HARQ information related to UL-SCH. The control channels are formed by aggregation

of control channel elements (CCE), each control channel element consisting of a set of resource element groups (REG). Different code rates for the control channels are realized by aggregating different number of control channel elements. Polar code is used for PDCCH channel coding. Each resource element group carrying PDCCH includes its own DM-RS. QPSK modulation is used for PDCCH modulation. A resource element is the smallest unit of the resource grid consisting of one subcarrier in frequency domain and one OFDM symbol in time domain. A PDCCH corresponds to a set of resource elements carrying DCI. Each NR control channel element consists of six REGs where a REG is equivalent to one resource block (12 resource elements in the fre- quency domain) over one OFDM symbol. The CCE size is designed such that at least one UE-specific DCI can be transmitted within one CCE with lower code rates. An NR REG bundle is further defined comprising 2, 3, or 6 REGs, which provides: (1) it determines the precoder cycling granularity (which affects the channel estimation performance) and (2) it is the interleaving unit for the distributed REG mapping. An NR PDCCH candidate consists of a set of CCEs, that is, 1, 2, 4, 8, or 16, corresponding to aggregation levels (ALs) 1, 2, 4, 8, 16, respectively. A control search space consists of a set of PDCCH can- didates and is closely associated with the ALs, the number of decoding candidates for each AL, and the set of CCEs for each decoding candidate. A search spacein NR Rel-15 is associated with a single CORESET. As shown in Fig. 4.14 CORESET is defined as a set of REGs with a given numerology. In the frequency domain, a CORESET is defined as a set of contiguous or distributed physical resource blocks configured using a six-PRB granularity, within which the UE attempts to blindly decode the DCI. There is no restriction on the maximum number of segments for a given CORESET. In the time domain, a CORESET spans 1, 2, or 3 contig- uous OFDM symbol, and the exact duration is signaled to the UE v

ia broadcast SI or UE- specific RRC signaling depending on whether it is a common CORESET or UE-specific CORESET. Compared to LTE PDCCH, the configurability of the CORESETs enable effi- cient resource sharing between downlink control and shared channels, thereby allowing efficient layer-1 signaling overhead management. One of the factors which could impact Chapter 4 PDCCH aggregation level 2 PDCCH REIRE ERERERE One-symbol CORESET REG10 REG11 REG12 REG13 REG14 REG15 REG16 REG17 Two-symbol CORESET REG10 REG12 REG14 REG16 REG18 REG20 REG22 REG24 REG26 REG28 REG30 REG32 REG34 REG11 REG13 REG15 REG17 REG19 REG21 REG23 REG25 REG27 REG29 REG31 REG33 REG35 REG bundle One-symbol CORESET, bundling size 2 REG10 REG11 REG12 REG13 REG14 REG15 REG16 REG17 REG Bundle Two-symbol CORESET, bundling size 2 REG10 REG12 REG14 REG16 REG18 REG20 REG22 REG24 REG26 REG28 REG30 REG32 REG34 REG11 REG13 REG15 REG17 REG19 REG21 REG23 REG25 REG27 REG29 REG31 REG33 REG35 REG bundle Two-symbol CORESET, bundling size 6 REG10 REG12 REG14 REG16 REG18 REG20 REG22 REG24 REG26 REG28 REG30 REG32 REG34 REG11 REG13 REG15 REG17 REG19 REG21 REG23 REG25 REG27 REG29 REG31 REG33 REG35 Frequency Figure 4.14 Illustration of RE, REG, CCE, REG bundle, and CORESET and example mappings of CCE to REG Bundles 14,68]. the time-domain duration of a CORESET is the bandwidth of the corresponding carrier, where more control symbols may be allowed for smaller bandwidths. For example, assum- ing a CORESET consists of 48 PRBs with 2 OFDM symbols, there are 16 CCEs that could accommodate up to 2 PDCCH candidates at AL-8 or a single candidate at AL-16. Furthermore, there can be multiple CORESETs inside the system bandwidth; thus the CORESET may not fully occupy the system bandwidth in the frequency domain. Downlink power adjustment can be applied to CORESETs that occupy narrower fre- quency regions within the carrier bandwidth, depending on desired coverage and link New Radio Access Physical Layer Aspects (Part 2) 443 budget. In that case, one or two OFDM symbols may not be s

ufficient. One-symbol CORESET offers benefits from the perspective of latency and control overhead adjust- ments especially when there are few UEs in the cell or when the coverage target is limited (e.g., small cell deployments). The maximum CORESET duration that may be configured in a cell is implicitly signaled via PBCH. A UE may be configured with one or more CORESETs (using UE-specific or common higher layer signaling) with a maximum of three CORESETs per configured (downlink) BWP. Limiting the maximum number of CORESETs is beneficial for enabling more practical RRC signaling and better UE dimen- sioning. Note that the scheduling flexibility may not be impacted by limiting the maxi- mum number of CORESETs since different monitoring occasions can be flexibly configured associated with the same CORESET. It is important to note that the concept of PDCCH monitoring periodicity is defined per search space set and is not configured at the CORESET level. Every configured search space with a certain monitoring periodicity (in terms of slots and starting symbols within the monitored slots) is associated with a CORESET. For a CORESET configured by UE-specific RRC signaling, some of the con- figured parameters include frequency-domain resources, starting OFDM symbol, CORESET duration, REG bundle size, transmission type (i.e., interleaved or non-inter- leaved), and precoding assumptions for channel estimation filtering [53]. As we mentioned earlier, a PDCCH consists of one or more CCEs. A CORESET consists JCORESET RB resource blocks in the frequency domain, determined by the RRC parameter frequencyDomainResources ControlResourceSet information element, symb E {1, OFDM symbols in the time domain, defined by the RRC parameter duration in the ControlResourceSet information element, where N CORESET = 3 is supported, if the RRC parameter dmrs-TypeA-Position is set to 3 [6]. A control channel element con- sists of six REGs where a REG is equivalent to one resource block over one OFDM symbol. The REGs within a CORESET are numbe

red in increasing order in a time-first manner, starting with 0 for the first OFDM symbol and the lowest numbered resource block in the CORESET. A UE can be configured with multiple CORESETs. Each CORESET is associ- ated with only one CCE-to-REG mapping (see Fig. 4.14). There is a direct correspondence between the number of CCEs and the AL, for example, for ALs 1, 2, 4, 8, and 16, there will be 1, 2, 4, 8, and 16 CCEs, respectively [6]. The time-frequency structure of REG, CCE, REG bundle, and CORESET as well as example mappings of CCE to REG bundles are illustrated Fig. 4.14. The PDCCH processing steps are illustrated in Fig. 4.15. At a high level, the PDCCH pro- cessing in NR is similar to that of LTE ePDCCH than LTE PDCCH in the sense that each PDCCH is processed independently. As shown in the figure, the entire DCI bits are used to calculate the CRC parity bits. Let us assume that ao, a1 ANDCI-1 denote the DCI input bits, and Po, PL-1 represent the parity bits, where NDCI and L 24 are the payload size and the number of parity bits, respectively. Let us assume that a'o, a' anda- Chapter 4 Scheduling information (DCI data) Information element Modulation multiplexing TS 38.211 Section 7.3.2.4 TS 38.212 Section 7.3.1 PDCCH PDCCH CRC attachment Resource element mapping TS 38.212 Section 7.3.2 TS 38.211 Section 7.3.2.5 Channel coding TS 38.212 Section 7.3.3 Rate matching TS 38.212 Section 7.3.4 Scrambling TS 38.211 Section 7.3.2.3 Figure 4.15 Physical layer processing of NR PDCCH [6,7]. sequence such that a' = 1 Vi = 0, 1, L - 1 and a- = L + NDCI - 1. The parity bits are computed with input bit sequence a'o,a ,ANDa+L-1 using the genera- tor polynomial gCRC24C(D) = D24 - D23 + D21 + D20 + D17 +D15+D13 + D12 ++D4+ D2 + D + 1. The output bit sequence is given as bo,b1,..., ,bb-1 where bk=akVk= = 0, 1, NDCI - 1 and bk = Pk-NDCI Ak :=NDcI,NDCI+1,...,NDCI+L-1.Following = the attach- ment of the CRC bits, the sequence is scrambled with the corresponding 16-bit RNTI XRNTI (0), XRNTI(1), XRNTI(15), where XRNTI(0) correspond

s to the MSB of the RNTI binary value, resulting in the sequence of bits C1,..., CK-1 where = bk = 0, 1, NDCI + 7 and Ck = [bk ++XRNTI(k) - NDCI - 8)]mod 2AK = NDCI = + 8, NDCI + 9,..., NDCI + 23 [6,7]. The PDCCH encoding stages are shown in Fig. 4.15. The K scrambled information bits are delivered to the channel coding block and are polar coded, by setting the encoder para- meters to the following nmax = 9,I = 1, NPC = 0, and nwm=0. The encoding process pro- duces N bits which are denoted as do, d1,..., dN-1. The rate matching for polar coded bits is performed on per coded block and consists of subblock interleaving, bit collection, and bit interleaving. Detailed PDCCH channel encoding and decoding block diagram is shown in Fig. 4.16. New Radio Access Physical Layer Aspects (Part 2) 445 Determination Determination of encoded of RNTI bit length DCI bits sequence Ones-initialized Frozen bit Subblock CRC24C attachment Polar encoding Rate matching Multiplexing generation padding scrambling interleaving insertion interleaving onto PDCCH PDCCH encoding procedure Determination Determination of of RNTI information block length Subblock DCI bit sequence extraction Distributed CRC-aided successive cancellation list polar decoding Demultiplexing deinterleaving dematching from PDCCH PDCCH decoding procedure Figure 4.16 NR PDCCH channel encoding/decoding block diagram [7,35]. No puncturing or shortening Puncturing Shotening Repetition Start Subblock Circular Bit-level Polar encoding interleaving buffer interleaving Figure 4.17 Encoding of NR PDCCH and rate matching variants [68]. The input bit sequence to rate matching function is denoted by do, d1 ., dN- and the out- put is denoted by fo,f1,...,FE-1. The input bits to the subblock interleaver are divided into 32 subblocks and the output bits are generated according to yn = dJ(n) where J(n) = mod N/32) An=0,1, ., N - 1 and the subblock interleaver pattern is defined in [7]. The repetition, puncturing, or shortening of polar code is per- formed in the following manner: N = 2n

d coded bits at the output of polar encoder is written into a length-N circular buffer in an order that is predefined for a given value of N. As shown in Fig. 4.17, to obtain M coded bits for transmission, puncturing is realized by select- ing bits from position N - M to position N - 1 from the circular buffer, shortening is real- ized by selecting bits from position 0 to position M - 1 from the circular buffer, and repetition is realized by selecting all bits from the circular buffer, and additionally repeating M - N consecutive bits from the circular buffer [7]. For each CORESET, there is an associated CCE-to-REG mapping based on the REG bundle (see Fig. 4.14). A REG bundle is a set of REGs across which the device can assume the 446 Chapter 4 precoding is constant. This property can be exploited to improve the channel estimation per- formance, which is similar to PRB bundling for PDSCH. The CCE-to-REG mapping can be either interleaved or non-interleaved, depending on the characteristics of the transmission channel, that is, frequency-flat or frequency-selective fading channel. There is only one CCE-to-REG mapping for a given CORESET; however, since the mapping is a property of the CORESET, multiple CORESETs can be configured with different mappings. The CCE- to-REG mapping for a CORESET can be interleaved or non-interleaved, configured by the RRC parameter ace-REG-MappingType in the ControlResourceSet information element and is described by REG bundles. The REG bundle i is defined as REGs {iLREG, iLREG + 1, iLREG + LREG - - 1 where LREG is the size of the REG bundle and REG CORESET = NCORESET the number of REGs in the CORESET. The jth CCE consists of REG bundles {(6j/LREG), (6j/LREG + 1), (6j/LREG + 6/LREG - 1)}, where P(.) denotes an interleaving function. In case of non-interleaved CCE-to-REG mapping LREG = 6 and (j)=j, whereas in the case of interleaved CCE-to-REG mapping LREG = {2,6} for NCORESET symb and LREGE & N CORESET,6} for NCCORESET symb E {2,3} where LREG is configured by the RRC parameter reg-Bundl

eSize. The interleaving function is defined by (j)=rC+c+nshifmodNCORESET/LReG;j = 1, R - 1; = 0, 1, C - 1 andC =NCORESET/LREGR where Re {2, 3, 6} is given by the higher layer parameter interleaverSize. Other parameters are defined as fol- lows: = a PDCCH transmitted in a CORESET configured by the PBCH or SIB1, nshift {0, 1, 274} is given by the RRC parameter shiftIndex. The UE is not expected to monitor configurations for which C is not an integer. For both interleaved and non-interleaved mappings, the same precoding is used within an REG bundle, if the higher layer parameter precoderGranularity equals LREG. The same precoding is used across all REGs within the set of contiguous resource blocks in the CORESET, if the higher layer parameter precoderGranularity equals the size of the CORESET in the frequency domain. For a CORESET configured by PBCH, LREG = 6,R = 2 and the same precoding is used within the REG bundle [6]. Unlike LTE, where the length of the control region can vary dynamically as indicated by PCFICH, a CORESET in NR is of fixed size. This is important from an implementation perspective, both for the UE and the network. From the UE perspective, a pipelined implementation is simpler, if the device can directly start to decode PDCCH without having to first decode another control channel. Various REG-to- CCE mapping options are shown in Fig. 4.18. During the PDCCH detection and decoding process, the UE needs to estimate the channel using the reference signals associated with the PDCCH candidate being decoded. A single antenna port is used for PDCCH transmission which means any transmit diversity or multi- user MIMO scheme is handled in a device-transparent manner. The PDCCH has its own DM-RS, based on the same pseudo-random sequence that is used for PDSCH, that is, the pseudo-random sequence is generated across all the common resource blocks in the REG-to-CCE mapping REG-to-CCE mapping (time REG-to-CCE mapping REG-to-CCE mapping (time (frequency first allocation) first allocation) (frequency first alloca

tion) first allocation) REG 1 REG 1 REG 2 REG 1 REG 1 REG 2 DM-RS REG 2 REG 3 REG 4 REG 2 REG 3 REG 4 REG 3 REG 5 REG 6 REG 3 DM-RS REG 4 REG 4 REG 5 REG 6 REG 5 REG 5 REG 6 DM-RS REG 6 REG composition and DM-RS locations Figure 4.18 REG-to-CCE mapping options and REG composition 6] 448 Chapter 4 frequency domain, but transmitted only in the resource blocks used for PDCCH (with one exception as discussed below). However, during initial access, the location of the common resource blocks is not known to UE as this information is signaled as part of the minimum SI. Therefore, for CORESET 0 configured by PBCH, the sequence is generated starting from the first resource block in the CORESET. The RRC parameters that define a CORESET are as follows [13]: ControlResourceSet:: = SEQUENCE controlResourceSetId ControlResourceSetId, frequencyDomainResources BIT STRING (SIZE (45)), duration INTEGER ...maxCoReSetDuration) , , //maxCoReSetDuration = 3 ice-REG-MappingTyp CHOICE { interleaved SEQUENCE { reg-BundleSize ENUMERATED {n2, n3, n6}, interleaverSize ENUMERATED {n2, n3, n6}, shiftIndex INTEGER(0. maxNrofPhysicalResourceBlocks-1 nonInterleaved precoderGranularity ENUMERATED {sameAsREG-bundle, allContiguousRBs} tci-StatesPDCCH SEQUENCE (SIZE (1. .maxNrofTCI-StatesPDCCH) OF TCI-StateId tci-PresentInDCI ENUMERATED {enabled} OPTIONAL pdcch-DMRS-ScramblingID BIT STRING (SIZE (16)) OPTIONAL In the preceding definition [6,9], controlResourceSetId corresponds to L1 parameter CORESET-ID whose value 0 identifies the common CORESET configured in MIB and in ServingCellConfigCommon and values 1, 2, ,maxNrofControlResourceSets-1 identify the CORESETs configured by dedicated signaling. The controlResourceSetId is unique among the BWPs of a serving cell. frequencyDomainResources corresponds to the L1 parameter CORESET-freg-dom. Each bit corresponds a group of six RBs, with the grouping start from PRB 0, which is fully New Radio Access Physical Layer Aspects (Part 2) 449 contained in the bandwidth part within which the CORESET is configured

. The most significant bit corresponds to the group of the lowest frequency which is fully contained in the bandwidth part within which the CORESET is configured, each subsequent lower significant bit corresponds to the next lowest frequency group that are fully contained within the bandwidth part in which the CORESET is configured. The bits corresponding to a group not fully contained within the bandwidth part in which the CORESET is con- figured are set to zero. duration is the contiguous time duration of the CORESET in number of symbols. cce-reg-MappingType identifies the mapping method of CCE-to-REG. reg-BundleSize is the number of REGs within an REG bundle corresponding to L1 parameter CORESET-REG-bundle-size. interleaveSize corresponds to L1 parameter CORESET-interleaver-size. shiftIndex corresponds to CORESET-shift-index. precoderGranularity denotes the precoder granularity in frequency domain. It corre- sponds to L1 parameter CORESET-precoder-granularity. tci-StatesPDCCH is a reference to a configured transmission configuration indication (TCI) state providing QCL configuration/indication for PDCCH. Downlink beamforming is typically transparent to the UE, that is, the device does not need to know which beam is used at the transmitter. However, NR also supports beam indication, which implies that the UE is informed of a certain PDSCH and/or PDCCH transmission using the same transmit beam as a configured ref- erence signal (CSI-RS or SS block). The beam indication is based on the (downlink signaling of) transmission configuration indication (TCI) states. Each TCI state includes information about a reference signal, for exam- ple, a CSI-RS or an SS block. By associating a certain downlink transmission (PDCCH or PDSCH) with a certain TCI, the network informs the UE that it can assume the upcoming downlink transmission uses the same spatial filter as the reference signal associated with that TCI. A device can be configured with up to 64 candidate TCI states. The beam indication for PDCCH is done by assigning a

subset of the M configured can- didate states via RRC signaling to each configured CORESET. Using MAC signaling, the network can dynamically indicate a specific TCI state, within the per-CORESET-configured subset, to be valid. When monitoring PDCCH within a certain CORESET, the device can assume that the PDCCH transmission uses the same spatial filter as the reference signal associated with the MAC-indicated TCI. In other words, if the device has determined a suitable receiver-side beam direction for reception of the reference signal, it can assume that the same beam direction is suitable for reception of the PDCCH. For PDSCH beam indication, there are two alternatives depending on the scheduling offset, that is, depending on the transmission timing of PDSCH relative to the corresponding PDCCH carrying scheduling information for that PDSCH. If this sched- uling offset is larger than N symbols, the DCI of the scheduling assignment may explicitly indicate the TCI state for the PDSCH transmission. To enable this, the device is initially configured with a set of up to eight TCI states from the originally configured set of candidate TCI states. A three-bit indicator within the DCI then indicates the exact TCI state which is valid for the scheduled PDSCH transmission. If the scheduling off- set is smaller or equal to N symbols, the device should instead assume that the PDSCH transmission is QCL with the corresponding PDCCH transmission. In other words, the TCI state for the PDCCH state indicated by MAC signaling should be assumed to be valid for the corresponding scheduled PDSCH transmission. The rea- son for limiting the dynamic TCI selection based on DCI signaling to the scenarios where the scheduling off- set is larger than a certain value is that for shorter scheduling offsets, there will not be sufficient time for the UE to successfully decode the TCI information within the DCI and to adjust the receive beam accordingly before the PDSCH is received [14]. 450 Chapter 4 tci-PresentInDCI corresponds to L1 parameter

CORESET-precoder-granularity. pdcch-DMRS-ScramblingID is the PDCCH DM-RS scrambling initialization. The DM-RSs associated with a given PDCCH candidate are mapped to every fourth subcar- rier in a REG, that is, the reference signal overhead is one-fourth. This is a denser reference signal pattern relative to LTE, which has a reference signal overhead of one-sixth; however, an LTE device can interpolate channel estimates across time and frequency as a result of cell-specific reference signals common to all devices and present regardless of control chan- nel transmission. The use of a dedicated reference signal per PDCCH candidate is advanta- geous, despite the slightly higher overhead, since it allows different type of device- transparent beamforming schemes. By using a beamformed control channel, the coverage and performance can be enhanced compared to the non-beamformed control channels in LTE. This is an essential part of the beam-centric design of NR [14]. When attempting to decode a PDCCH candidate occupying certain number of CCEs, the device can compute the REG bundles that constitute the PDCCH candidate. Channel esti- mation must be performed per REG bundle as the network may change precoding across REG bundles. In general, this results in sufficiently accurate channel estimates for PDCCH detection. However, it is also possible to configure the device to assume the same precoding across contiguous resource blocks in a CORESET, thereby allowing the device to perform frequency-domain interpolation of the channel estimates. This also implies that the device may use reference signals referred to as wideband reference signals outside the PDCCH region that it is trying to detect (see Fig. 4.19). The QCL concept is also applicable to the reference signals. If the UE has a priori knowledge about the QCL of two reference signals, it can exploit this property to improve the channel estimation and to manage different receive beams at the device. If no QCL is configured for a CORESET, the UE assumes that PDCCH candidate

s are quasi-co-located with the SS/ PBCH block with respect to delay spread, Doppler spread, Doppler shift, average delay, and spatial RX parameters. This is a reasonable assumption as the device has been able to receive and decode the PBCH in order to access the system. The block of bits b(0), b(1), b(NPDCCH - 1), where NPDCCH denotes the number of bits transmitted on PDCCH, is scrambled prior to modulation, resulting in a block of scram- bled bits b(0), b(1), b(NPDCCH - 1) where b(i) = [b(i) + c(i) |mod 2, in which c(i) is a length-31 Gold sequence generated by the pseudo-random sequence generator defined in [6] and initialized with Cinit = (nRNTI216 nID) mod 231. For a UE-specific search space, NIDE{0,1,. 65535} is set by the RRC parameter pdcch-DMRS-ScramblingID; other- wise, N I D = Ncell and NRNTI is determined by the C-RNTI for a PDCCH in a UE-specific search space, when the RRC parameter pdcch-DMRS-ScramblingID is configured; otherwise NRNTI 0. The block of scrambled bits b(0), b(1), b(NPDCCH QPSK modulated, resulting in a block of complex-valued modulation symbols New Radio Access Physical Layer Aspects (Part 2) 451 CORESET CORESET DM-RS for PDCCH 2 DM-RS for all PDCCHs in the CORESET DM-RS for PDCCH 1 Wideband reference DM-RS per PDCCH signal Figure 4.19 Illustration of the regular and wideband reference signals for PDCCH [68]. d(0), d(1), ., d(Nsymb - 1). The block of complex-valued symbols d(0), d(1), d(Nsymb - 1) is scaled by a factor of BPDCCH and is mapped to resource elements (k,1), which are designated to PDCCH to be monitored by the UE and are not used for the associated PDCCH DM-RS, in increasing order of first k (frequency index) and l (time index) [6]. Fig. 4.20 illustrates the CORESET structures in overlapped and non- overlapped BWPs. 452 Chapter 4 NRB = = max(nRB) = max(nRB) CORESET 1 CORESET 1 CORESET CORESET 2 Overlap Overlap CORESET CORESET NRB =0 NRB=0 Figure 4.20 Illustration of the CORESET structures in overlapped and non-overlapped BWPs [71]. The PDCCH DM-RS sequence 'PDCCH(1,m) on O

CORESET is configured by the PBCH or by the controlResourceSetZero field in the PDCCH-ConfigCommon information element; otherwise, subcarrier 0 in common resource block 0. The parameter l is the OFDM symbol number within the slot. In the absence of CSI- RS configuration, the PDCCH DM-RS and SS/PBCH blocks are quasi-co-located with respect to Doppler shift, Doppler spread, average delay, delay spread, and spatial RX parameters. New Radio Access Physical Layer Aspects (Part 2) 453 The notion of wideband DM-RS has been introduced to assist the NR UE in channel estima- tion for the control channel detection (see Fig. 4.19). For each CORESET, the precoder granularity in the frequency domain is configurable between REG bundle size and the num- ber of contiguous RBs in the frequency domain within the CORESET. The UE may assume DM-RS is present in all REGs within the set of contiguous RBs of the CORESET where and when at least one REG of a candidate is mapped [6]. 4.1.3.2.2 UE Group-Common Signaling An NR gNB can simultaneously support diverse service categories and thus mitigation of the performance degradation of interrupted services is an important issue in the physical layer design. While the flexible frame structure may alleviate this problem, due to the implementation complexity and unpredictability of URLLC packet arrival times, a more elaborate solution is needed in real-world deployment scenarios. 3GPP has considered a number of solutions in the course of NR development. For infrequent URLLC transmis- sions, one can give priority to URLLC transmissions while ensuring the reliability of the other transmissions interrupted by URLLC traffic. A preemption indicator transmitted by the base station indicates which resources are used for the URLLC transmission. If the URLLC packet is stretched in the frequency domain, the URLLC transmission will interrupt the entire system bandwidth and thus degrade all data channels in use. To notify scheduled users of this event, the base station broadcasts a preemption indicator con

sisting of time and/or frequency information of the interrupted resources. This indicator helps users identify the reason for packet errors and what part of the packet is unaffected from the interruption. Retransmission of selected code blocks when the ongoing service is interrupted by the URLLC transmission is another solution, where part of the code block that has been affected by URLLC transmission is retransmitted. By transmitting a combining indicator or flush-out indicator, the receiver can perform the soft symbol combining of the transmitted and retransmitted code blocks. One can further achieve better coding gain by lowering the code rate of the retransmitted code block. An efficient scheduling scheme in terms of resource allocation and latency is to multiplex data with different transmission time lengths (i.e., mini-slots and slots) and in case of resource limitations to let the high-priority service use resources from lower priority ser- vice. This type of multiplexing is also referred to as preemption. For example, in NR downlink, a mini-slot carrying high-priority or delay-sensitive data can preempt an already ongoing slot-based transmission on the first available OFDM symbols without waiting for the next free resource. This operation enables ultra-low latency for mini-slot- based transmission, especially in the scenario where a long slot-based transmission has already been scheduled. A similar concept is also considered for the uplink and in general for LTE. At the cost of degrading the longer transmission, no additional resources need to be reserved in advance for the URLLC service. The impacted longer transmission is then 454 Chapter 4 promptly repaired with a transmission containing a subset of the code block groups in a later transmission time, after providing the essential information to clean the corrupted soft values in the receive buffer from the preempted data. If the URLLC transmission occurs frequently, the efficiency of the above approach will be reduced due to the fre- quent retransmissi

ons. To ensure the reliability of the ongoing services while supporting the URLLC transmission, robustness improvement and service sharing strategies may be adopted [68]. An NR UE can be configured to monitor group-common signaling via DCI format 2_1 which carries preemption indication related to multiplexing eMBB and URLLC traffic with different transmission durations in the downlink. Upon reception of preemption indication, a UE should apply an appropriate HARQ combining mechanism to retrieve the data despite the missing portion due to preemption. Group-common PDCCH is designed for the purpose of signaling to a group of UEs. It carries dynamic SFI, that is, via DCI format 2_0, to indi- cate which symbols in a slot are designated as downlink, uplink, or flexible symbols. The SFI carries an index to a UE-specific table containing permissible slot configurations [6]. The downlink preemption indicator is signaled via DCI format 2_1; however, whether a UE needs to monitor preemption indication is configured through RRC signaling. The UE is additionally configured with a set of serving cells; a mapping for each serving cell in the set of serving cells to the corresponding fields in DCI format 2_1; an information payload size for DCI format 2_1; and a bitmap for identification of punctured time-frequency resources via higher layer signaling. If the UE detects a DCI format 2_1 for a serving cell from the configured set of serving cells, the UE may assume that there is no transmission assigned to the UE in PRBs and symbols within the active downlink BWP, from a set of PRBs and a set of symbols of the last monitoring period, that are indicated by DCI format 2_1. Note that DCI format 2_1 indication is not applicable to the reception of SS/PBCH blocks [8]. A UE needs to monitor preemption indication carried by DCI format 2_1, if it is provided with RRC parameter DownlinkPreemption and it is configured with an INT-RNTI provided by RRC parameter int-RNTI. In that case, if the UE detects DCI format 2_1, the set of sym- bols i

ndicated by a field in DCI format 2_1 includes the last slot symb TINT21-HINI symbols prior to the first symbol of the CORESET in the slot. The parameter TINT is the PDCCH monitor- ing periodicity provided by a higher layer parameter, N slot symb is the number of symbols per slot, u is the subcarrier spacing configuration for a serving cell with mapping to a respective field in the DCI format 2_1, and HINT is the subcarrier spacing configuration of the down- link BWP where the UE receives the PDCCH conveying the DCI format 2_1. If the UE is configured with RRC parameter TDD-UL-DL-ConfigurationCommon the symbols desig- nated as uplink by the latter parameter are excluded from the last jsymb slot TINT2 HINT symbols prior to the first symbol of the CORESET in the slot. The resulting set of symbols includes a number of symbols that is denoted as NINT [8]. New Radio Access Physical Layer Aspects (Part 2) 455 The UE is further provided with the (preemption) indication granularity for the set of PRBs and OFDM symbols (within a slot) that might be preempted through RRC parameter timeFrequencySet. If the value of latter parameter is zero (time-domain preemption region boundary), the 14-bit bitmap in DCI format 2_1 has a one-to-one correspondence with 14 groups of consecutive symbols from the set of symbols where each of the first NINT - NINT/14 symbol groups includes NINT/14 symbols, each of the last 14 - NINT 14 NINT/14 symbol groups includes NINT/14 symbols, where bitmap, bit value of "0" indicates transmission to the UE in the corresponding symbol group and a bit value of "1" indicates no transmission to the UE in the corresponding symbol group [8,13]. Interpretation of the bitmap is configurable such that each bit represents one OFDM symbol in the time domain and the full bandwidth part, or two OFDM symbols in the time domain and one half of the bandwidth part. Furthermore, the monitoring periodicity of the preemption indi- cator is configured in the device every nth slot. An example is shown in Fig. 4.21 where UE1 ha

s been scheduled with a downlink transmission spanning one slot. During the transmission to UE1, delay-sensitive data for UE2 arrives at the gNB, which immediately schedules a trans- mission to UE2. Typically, if there are frequency resources available, the transmission to UE2 is scheduled using resources not overlapping with the ongoing transmission to UE1. However, in a heavy-loaded network, this may not be possible and there is no option but to use some of the resources originally allocated to UE1 for the delay-sensitive transmission to UE2. We refer to this case as the transmission to UE2 preempting the transmission to UE1, which would experience temporary performance degradation because some of the resources that UE1 assumes to contain its own data contain data for UE2. If the value of the RRC parameter timeFrequencySet is one (frequency-domain preemption region boundary) then seven pairs of bits of a field in the DCI format 2_1 have a one-to- one mapping with seven groups of consecutive symbols where each of the first NINT - 7 NINT/7 symbol groups includes NINT/7 symbols, 7 - NINT 7 NINT/7 symbol groups includes NINT/7 symbols (as shown in Fig. 4.21). The first bit in a pair of bits for a symbol group is applicable to the subset of first BINT/2 PRBs from the set of BINT PRBs, and the second bit in the pair of bits for the symbol group is applicable to the subset of last BINT/2 PRBs from the set of BINT PRBs, a bit value of "0" indicates transmission to the UE in the corresponding symbol group and subset of PRBs, and a bit value of "1" indicates no transmission to the UE in the corresponding sym- bol group and subset of PRBs [8,13]. In relation to the dynamic slot configuration, if a UE is configured by RRC parameter SlotFormatIndicator, it will be provided with SFI-RNTI via higher layer parameter sfi- RNTI, and the payload size of DCI format 2_0 via RRC parameter dci-PayloadSize. The UE is also provided, in one or more serving cells, with a configuration for search space set S and the corresponding CORESET

P for monitoring the first Mps PDCCH candidates 456 Chapter 4 PDCCH with DCI Preempted format 2_1 resources (containing Downlink allocated preemption resources to UE2 indication) allocated to Uplink UE1 in slot n resources Slot n Slot n+1 Preemption indication monitoring period {1,2,4} slots 00001000100000 Preemption indication bitmap in DCI format 2_1 when timeFrequencySet = 0 PDCCH with DCI Preempted format 2_1 resources Downlink allocated (containing resources to UE2 preemption indication) allocated to Uplink UE1 in slot n resources 0000100 Preemption indication bitmap in DCI format 2_1 when timeFrequencySet= Slot n Slot n+1 Preemption indication monitoring period {1,2,4} slots Figure 4.21 Illustration of downlink preemption indication (interrupted transmission indication) [8]. related to DCI format 2_0 at aggregation level of LSFI CCEs to locate Type3-PDCCH com- mon search space. The DCI format 2_0 indicates a slot format corresponding to each slot in the downlink or uplink BWPs of a serving cell. The indication is communicated by setting the value of the SFI index in DCI format 2_0 for the serving cell to a combination of slot New Radio Access Physical Layer Aspects (Part 2) 457 formats for a number of slots. A slot format is identified by its corresponding index and the mapping between values of the SFI index and combinations of slot formats is signaled via RRC parameters [8]. In TDD operation mode, the UE is provided, via RRC signaling, with a reference subcarrier spacing configuration USFI for each slot format in a set of slot formats indicated by the SFI index in DCI format 2_0. An active downlink and uplink BWP pair are associated with sub- carrier spacing configuration HZUSFI Each slot format in the combination of slot formats is applicable to 2(H-HSFI) consecutive slots in the active downlink and uplink BWP pair where the first slot starts at the same time as the first slot for the reference subcarrier spac- ing configuration and each downlink, uplink, or flexible symbol for the reference subcarrier s

pacing configuration corresponds to 2(H-HSFI) consecutive downlink, uplink, or flexible symbols for the subcarrier spacing configuration u [8]. In FDD operation mode, the SFI index field in DCI format 2_0 indicates an assortment of slot formats that include (separate) combinations of slot formats for reference downlink and uplink BWPs of the serving cell. The UE is provided via RRC signaling with the reference subcarrier spacing configurations USFI-DL or USFI-UL for the combination of slot formats indicated by the SFI index field value in DCI format 2_0 for the reference downlink and uplink BWPs of the serving cell, respectively. 4.1.3.2.3 Downlink Control Information Formats The PDCCH payload is known as DCI to which a 24-bit CRC is attached to detect transmis- sion errors and to aid the decoder in the UE receiver. Compared to LTE, the CRC size has been increased to reduce the risk of incorrectly received control information and to assist early termination of the decoding operation in the receiver. A DCI transports downlink and uplink scheduling information, requests for aperiodic channel quality indicator (CQI) reports, or uplink power control commands for one cell and one RNTI. Depending on the content and purpose of each DCI, different formats are defined. Each DCI payload is pro- cessed by information element multiplexing, CRC attachment, channel coding, and rate matching. The DCI formats defined in NR are shown in Table 4.4 along with their use cases. Similar to LTE, the UE identity modifies the CRC transmitted through a scrambling operation. When the DCI is received, the UE will calculate a scrambled CRC on the pay- load part using the same transmit-side procedure and compares it against the received CRC. If the CRC matches, the message is declared to be correctly received and intended for the UE. Therefore, the identity of the UE that is supposed to receive the DCI message is implic- itly encoded in the CRC, which reduces the number of bits necessary to transmit on the PDCCH. Note that the RNTI which is f

urther scrambled with the DCI CRC is not necessar- ily the identity of the device (in the case of C-RNTI), rather it can be different type of group or common RNTIs used to indicate paging or a random-access response. The information fields in the DCI formats shown in Table 4.4 are mapped to the informa- tion bits ao to aNDCI~1 such that the first field is mapped to the lowest order information bit 458 Chapter 4 Table 4.4: NR downlink control information (DCI) formats [7]. Format Purpose Application Uplink Scheduling of PUSCH in one cell Scheduling Scheduling of PUSCH in one cell Downlink Scheduling of PDSCH in one cell scheduling DCI format 1_0 with CRC scrambled by C-RNTI DCI format 1_0 with CRC scrambled by RA-RNTI DCI format 1_0 with CRC scrambled by TC-RNTI DCI format 1_0 with CRC scrambled by SI-RNTI DCI format 1_0 with CRC scrambled by P-RNTI Scheduling of PDSCH in one cell Other Notifying a group of UEs of the slot format purposes Notifying a group of UEs of the PRB(s) and OFDM symbol(s) where UE may assume that no transmission is intended for the UE Transmission of TPC commands for PUCCH and PUSCH Transmission of a group of TPC commands for SRS transmissions by one or more UEs ao, and each successive field is mapped to higher order information bits. If the number of information bits in a DCI format is less than 12 bits, zero padding is used until the payload size is equal to 12. The DCI size of the uplink DCI format 0_1 and downlink DCI format 1_1 are made equal with padding bits added to the smaller of the two in order to reduce the number of blind decoding. It may appear that parts of the DCI content are the same for the different formats; however, there are differences due to different capabilities supported by each DCI format. The content of various DCI formats is described in the following [7]. Note that the information fields and their interpretation may change according to the RNTI value that is used in conjunction with the DCI value. DCI Format 0_0 Identifier for DCI formats (1 bit): It is a bit t

o indicate whether the DCI is a downlink assignment or an uplink grant. The value of this bit is always set to 0, indicating an uplink DCI format. Frequency domain resource assignment: The number of bits for this field is determined by following formula log2(NUL.,BWP(NUL,BWP+1)/2) The meaning of UL,BWP varies depending on the search space where DCI format 0_0 is expected to be detected. When transmitted in common search space, it indicates the size of the initial bandwidth part, whereas in UE-specific search space, it would indicate the size of the active bandwidth part, if the following criteria are satisfied: the total number of different DCI sizes moni- tored per slot is less than 4 and the total number of different DCI sizes with C-RNTI New Radio Access Physical Layer Aspects (Part 2) 459 monitored per slot is less than 3. The value of this field is determined in two cases: If PUSCH hopping is enabled and the resource allocation is Type 1, NUL-hop MSB bits are used to indicate the frequency offset. If PUSCH hopping is disabled and the resource allocation is type 1, the entire bits of this field would indicate PUSCH RIV. Time-domain resource assignment (4 bits): It carries the row index of the items in pusch-TimedomainAllocationList in RRC message for PUSCH configuration, where the indexed row defines the slot offset K2, the start and length indicator SLIV, and the PUSCH mapping type to be applied in the PUSCH transmission. Frequency hopping flag (1 bit): It is used to handle frequency hopping for resource allo- cation Type 1. Modulation and coding scheme (5 bits): It is used to provide the device with informa- tion about the modulation scheme, the code rate, and the transport block size (TBS) (see Table 4.13). New data indicator (1 bit): It is used to indicate whether the grant relates to retransmis- sion of a TB or transmission of a new TB. Redundancy version (RV) (2 bits): This field determines the value of the parameter rvid =0,1,2,3 which is used to indicate the redundant information sent in a HARQ retran

smission. HARQ process number (4 bits): It informs the device about the HARQ process to be used for soft combining. Transmit power control (TPC) command for scheduled PUSCH (2 bits): It is used to adjust the PUSCH transmission power. Uplink/Supplementary uplink (SUL) indicator (0 or 1 bit): One bit to indicate whether the grant relates to the SUL or the ordinary uplink, for UEs configured with SUL in the cell. It is only present if a SUL is configured as part of the SI. DCI Format 0_1 Identifier for DCI formats (1 bit): A bit to indicate whether the DCI is a downlink assignment or an uplink grant. The value of this bit is always set to 0, indicating an uplink DCI format. Carrier indicator (0 or 3 bits): This field is present if cross-carrier scheduling is config- ured and is used to indicate to which component carrier the DCI is related. Bandwidth part indicator (0, 1, 2 bits): It is used to activate one of up to four bandwidth parts configured by higher layer signaling. It is determined by the number of uplink BWPs NBWP configured via RRC signaling, excluding the initial uplink bandwidth part. The size of this field is [log2(nBWP) bits. Frequency domain resource assignment: This field indicates the resource blocks on one component carrier over which the device should transmit PUSCH. The number of bits is variable and dependent on the resource allocation type. The number of bits is equal to NRBG bits, if resource allocation Type 0 is configured for the UE. For resource 460 Chapter 4 allocation Type 1, the number of bits is equal to bits or 1 bits if both resource allocation Type 0 and 1 are configured. Note that if both resource allocation Type 0 and 1 are configured, the MSB bit is used to distinguish resource allocation Type 0 from Type 1. Time domain resource assignment (0, 1, 2, 3, or 4 bits): This field indicates the resource allocation in the time domain. The number of bits is determined as log2(I) bits, where I denotes the number of entries in RRC parameter pusch-TimeDomainAllocationList VRB-to-PRB mapping

(1 bit): It is used to indicate whether interleaved or non- interleaved VRB-to-PRB mapping should be used. Frequency hopping flag (1 bit): It is used to handle frequency hopping for resource allocation Type 1. Modulation and coding scheme (5 bits): It is used to provide the device with informa- tion about the modulation scheme, the code rate, and the TBS (see Table 4.13). New data indicator (1 bit): It is used to indicate whether the grant relates to retransmis- sion of a TB or transmission of a new TB. Redundancy version (2 bits): This field determines the value of the parameter rvid = 0, 1, 2, 3 which is used to indicate the redundant information sent in a HARQ retransmission. HARQ process number (4 bits): It informs the device about the HARQ process to be used for soft combining. First downlink assignment index (DAI) (1 or 2 bits): The DAI is used for handling of HARQ codebooks when UCI is transmitted on PUSCH. This field would be one bit for semi-static HARQ-ACK codebook and 2 bits for dynamic HARQ-ACK codebook. Second DAI (0 or 2 bits): This field would be 2 bits for dynamic HARQ-ACK codebook with two HARQ-ACK subcodebooks and zero bit otherwise. TPC command for scheduled PUSCH (2 bits): It is used to adjust the PUSCH transmis- sion power. Sounding reference signals (SRS) resource indicator: The SRI is used to determine the antenna ports and uplink transmission beam to use for PUSCH transmission. The num- ber of bits depends on the number of SRS groups configured and whether codebook- based or non-codebook-based precoding is used. Precoding information and number of layers (0, 2, 3, 4, 5, or 6 bits): This field is used to select the precoding matrix W and the number of layers for codebook-based precod- ing. The number of bits depends on the number of antenna ports and the maximum rank supported by the UE. Antenna ports (2, 3, 4, or 5 bits): This field indicates the antenna ports on which the data are transmitted as well as antenna ports that are scheduled for other users. SRS request (2 bits): This field is

used to request transmission of a sounding RS. New Radio Access Physical Layer Aspects (Part 2 461 CSI request (0, 1, 2, 3, 4, 5, or 6 bits): This field is used to request transmission of a CSI report. Code block group (CBG) transmission information (0, 2, 4, 6, or 8 bits): This field indi- cates the code block groups for retransmission. PT-RS-DM-RS association (0 or 2 bits): This field is used to indicate the association between the DM-RS and PT-RS ports. beta_offset Indicator (0 or 2 bits): This information is used to control the amount of resources used by UCI on PUSCH in case dynamic beta offset signaling is configured for DCI format 0_1. DM-RS sequence initialization (0 or 1 bit): This information is used to select between two preconfigured initialization values for the DM-RS sequence. It would be zero bit, if the transform precoder is enabled and 1 bit, otherwise. Uplink/SUL indicator (0 or 1 bit): It would be zero bit for UEs that are not configured with SUL in the cell or UEs configured with SUL in the cell but only physical uplink control channel (PUCCH) carrier in the cell is configured for PUSCH transmission; oth- erwise, 0 bit for UEs configured with SUL in the cell. DCI Format 1_0 Identifier for DCI formats (1 bit): A bit to indicate whether the DCI is a downlink assignment or an uplink grant. The value of this field is always set to 1, indicating a downlink DCI format. Frequency-domain resource assignment: This field indicates the resource blocks on one component carrier on which the UE receives PDSCH. The size of this field depends on the size of the bandwidth part and on the resource allocation type, that is, Type 0 only, Type 1 only, or dynamic switching between the two types. The number of bits is where the interpretation of NDL,BWP depends on the search space where DCI format 1_0 is monitored. The parameter NDL,BWP indicates the size of the active downlink bandwidth part, if DCI format 1_0 is monitored in the UE- specific search space and if the total number of different DCI sizes configured t

o moni- tor is less than 4 and the total number of different DCI sizes with C-RNTI configured to monitor is less than 3 for the cell; otherwise, NDL,BWP would denote the size of CORESET 0. Time-domain resource assignment (4 bits): This field indicates the resource allocation in the time domain. When the UE is scheduled to receive PDSCH by a DCI, the time- domain resource assignment field of the DCI provides a row index of a higher layer- configured table pdsch-symbolAllocation, where the indexed row defines the slot offset K0, the start and length indicator SLIV, and the PDSCH mapping type for PDSCH reception. VRB-to-PRB mapping (1 bit): It indicates whether interleaved or non-interleaved VRB-to-PRB mapping should be used and only presents for resource allocation Type 1. 462 Chapter 4 Modulation and coding scheme (5 bits): It is used to provide the UE with information about the modulation scheme, code rate, and TBS (see Table 4.13). New data indicator (1 bit): It is used to clear the UE soft buffer for initial transmissions. Redundancy version (2 bits): This field determines the value of the parameter rvid=0, 1, 2, 3 which is used to indicate the redundant information sent in a HARQ retransmission. HARQ process number (4 bits): This informs the device about the HARQ process use for soft combining. Downlink assignment index (2 bits): The DAI only presents when a dynamic HARQ codebook is used. The DCI format 1_1 supports 0, 2, or 4 bits, while DCI format 1_0 uses 2 bits. TPC command for scheduled PUCCH (2 bits): It is used to adjust the PUCCH transmis- sion power. PUCCH resource indicator (3 bits): It is used to select PUCCH resources from a set of configured resources. PDSCH-to-HARQ feedback timing indicator (3 bits): It provides information on when the HARQ acknowledgment should be transmitted relative to the PDSCH transmission. DCI Format 1_1 Identifier for DCI formats (1 bit): The value of this bit is always set to 1, indicating a downlink DCI format. Carrier indicator (0 or 3 bits): This field is present if cr

oss-carrier scheduling is config- ured and is used to indicate the component carrier that the DCI corresponds to. Bandwidth part indicator (0, 1, or 2 bits): The number of bits is determined by the num- ber of downlink BWPs NBWP,RRC configured by higher layers, excluding the initial downlink bandwidth part. The size of this field is equal to log2(nBWP) bits, where NBWP = NBWP,RRC + 1, if NBWP,RRC 3 in which case the bandwidth part indicator is equivalent to the higher layer parameter BWP-Id; otherwise NBWP,RRC. Frequency-domain resource assignment: This field indicates the resource blocks on one component carrier on which the device should receive PDSCH. The number of bits is variable and dependent on the resource allocation type. The number of bits is equal to NRBG bits, if resource allocation Type 0 is configured for the UE. For resource alloca- tion Type 1, the number of bits is equal to bits 1 bits, f both resource allocation Types 0 and 1 are configured. Note that if both resource allocation Types 0 and 1 are config- ured, the MSB bit is used to distinguish resource allocation Type 0 from Type 1. Time-domain resource assignment (0, 1, 2, 3, or 4 bits): The size of this field is deter- mined as log2(I) bits, where I is the number of entries in the higher layer parameter pdsch-TimeDomainAllocationList When the UE is scheduled to receive PDSCH by a New Radio Access Physical Layer Aspects (Part 2) 463 DCI, the time-domain resource assignment field of the DCI provides a row index of a higher layer-configured table pdsch-symbolAllocation, where the indexed row defines the slot offset K0, the start and length indicator SLIV, and the PDSCH mapping type for PDSCH reception. VRB-to-PRB mapping (0 or 1 bit): It indicates whether interleaved or non-interleaved VRB-to-PRB mapping should be used and only presents for resource allocation Type 1. PRB bundling size indicator (0 or 1 bit): It is used to indicate the PDSCH bundling size. It is zero bit if the RRC parameter prb-BundlingType is not configured or is set to "stati

c." It is one bit, if the higher layer parameter prb-BundlingType is set to "dynamic." Rate matching indicator (0, 1, or 2 bits): The number of bits is determined according to RRC parameters rateMatchPatternGroup1 and rateMatchPatternGroup2. Modulation and coding scheme (TB 1) (5 bits): It is used to provide the UE with infor- mation about the modulation scheme, code rate, and TBS (see Table 4.13) related to the first transport block. New data indicator (TB 1) (1 bit): It is used to clear the UE soft buffer for initial trans- missions related to the first transport block. Redundancy version (TB 1) (2 bits): This field determines the value of the parameter which is used to indicate the redundant information sent in a HARQ retransmission related to the first transport block. Modulation and coding scheme (TB 2)5 (5 bits): It is used to provide the UE with infor- mation about the modulation scheme, code rate, and TBS related to the second transport block (see Table 4.13). New data indicator (TB 2) bit): It is used to clear the UE soft buffer for initial trans- missions related to the second transport block. Redundancy version (TB 2) (2 bits): This field determines the value of the parameter which is used to indicate the redundant information sent in a HARQ retransmission related to the second transport block. HARQ process number (4 bits): This informs the device about the HARQ process to use for soft combining. Downlink assignment index (0, 2, or 4 bits): The number of bits is 4, if more than one serving cell is configured in the downlink and the RRC parameter pdsch-HARQ-ACK- Codebook = dynamic, where the two MSB bits are the counter DAI and the two LSB bits are the total DAI. The number of bits is 2, if only one serving cell is configured in the downlink and the RRC parameter pdsch-HARQ-ACK-Codebook = dynamic, where the 2 bits are the counter DAI; zero bits otherwise. If a second transport block is present (only if more than four layers of spatial multiplexing are supported in DCI format 1_1), the three fields above

are repeated for the second transport block. 464 Chapter 4 TPC command for scheduled PUCCH (2 bits): It is used to adjust the PUCCH transmis- sion power. PUCCH resource indicator (2 bits): It is used to select PUCCH resources from a set of configured resources. PDSCH-to-HARQ_Feedback timing indicator (3 bits): It provides information on when the HARQ acknowledgment should be transmitted relative to the PDSCH transmission. Antenna port(s) and number of layers (4, 5, or 6 bits): The antenna ports PNA-1} are determined according to the ordering of DM-RS port(s). If a UE is configured with both dmrs-DownlinkForPDSCH-MappingTypeA and dmrs-DownlinkForPDSCH- MappingTypeB, the size of this field is equal to max(XA,XB), where and XB are the "antenna ports" bit sizes derived from dmrs-DownlinkForPDSCH-MappingTypeA and dmrs-DownlinkForPDSCH-MappingTypeB respectively. A number of zeros are inserted in the XB MSB positions of this field, if the mapping type of the PDSCH corre- sponds to the smaller value of or XB. Transmission configuration indication (0 or 3 bits): The size of this field is zero bit, if RRC parameter tci-PresentInDCI is not enabled; otherwise it would carry 3 bits. SRS request (2 or 3 bits): It is used to request transmission of a sounding reference sig- nals in the uplink. For UEs not configured with SUL in the cell, 2 bits are used, whereas for UEs that are configured with SUL in the cell, 3 bits are used where the first bit is the non-SUL/SUL indicator and the second and third bits are used to request periodic or aperiodic SRS transmission. This bit field may also indicate the associated CSI-RS. CBG transmission information (0, 2, 4, 6, or 8 bits): If CBG retransmissions are config- ured, this field indicates the code block groups that are retransmitted. CBG flushing out information (0 or 1): If CBG retransmissions are configured, the con- tent of this field indicates the soft buffer flushing, which is determined by RRC parame- ter codeBlockGroupFlushIndicator. DM-RS sequence initialization (1 bit): This

information is used to select between two preconfigured initialization values for the DM-RS sequence. It would be zero bit, if the transform precoder is enabled and 1 bit otherwise. DCI Format 2_0 DCI format 2_0 is used for notifying the UE of slot format. The SFI is transmitted using regular PDCCH structure and SFI-RNTI, which is common to a group of UEs. To assist the device in the blind decoding, the device is configured with information about the up to two PDCCH candidates on which the SFI can be transmitted. DCI format 2_0 with CRC scrambled with SFI-RNTI carries Slot format indicator 1, Slot format indicator 2, Slot format indicator N. The size of DCI format 2_0 is configurable by higher layers up to 128 bits. DCI Format 2_1 DCI format 2_1 is used to signal the preemption indication to the device. It is transmitted using the regular PDCCH structure, using INT-RNTI which can be com- mon to multiple devices. In other words, DCI format 2_1 is used for notifying the UEs of New Radio Access Physical Layer Aspects (Part 2) 465 the PRB(s) and OFDM symbol(s) that are preempted and have no transmission intended for the UE. DCI format 2_1 with CRC scrambled by INT-RNTI carries Preemption indication 1, Preemption indication 2, Preemption indication N. The size of DCI format 2_1 is configurable by higher layers up to 126 bits where each preemption indi- cation is 14 bits. DCI Format 2_2 The main purpose of DCI format 2_2 is to support power control for semi- persistent scheduling (SPS) since there is no dynamic scheduling assignment or scheduling grant which can include the power control information for PUCCH and PUSCH in this case. The power-control message is addressed to a group of UEs using an RNTI specific for that group and each UE is configured with the power control bits in the message. DCI format 2_2 is further aligned with the size of DCI formats 0_0/1_0 to reduce the blind decoding complex- ity. DCI format 2_2 with CRC scrambled by TPC-PUSCH-RNTI or TPC-PUCCH-RNTI car- ries block number 1, block number 2, b

lock number N. The RRC parameters tpc-PUSCH or tpc-PUCCH determine the index to the block number for a cell uplink, with the following fields defined for each block: (1) closed-loop indicator (0 or 1 bit) and (2) TPC command (2 bits). DCI Format 2_3 DCI format 2_3 is used for power control of uplink sounding reference signals for the UEs which have not linked the SRS power control to the PUSCH power con- trol, either because independent control was desirable, or the UE was configured without PUCCH and PUSCH. DCI format 2_3 structure is similar to DCI format 2_2, with the pos- sibility to individually configure 2 bits for SRS request in addition to the two power control bits. DCI format 2_3 is aligned with the size of DCI formats 0_0/1_0 to reduce the blind decoding complexity. DCI format 2_3 with CRC scrambled by TPC-SRS-RNTI carries block number 1, block number 2, block number N where the starting position of a block is determined by the parameter startingBitOfFormat2-3 provided by the higher layers for the UE configured with the block. If the UE is configured with RRC parameter srs-TPC- PDCCH-Group = typeA for an uplink carrier without PUCCH and PUSCH or when the SRS power control is not linked to PUSCH power control, in DCI format 2_3 one block is configured for the UE containing 0 or 2 bits of SRS request. The TPC commands TPC command number 1, TPC command number 2, TPC command number N apply to the respective carriers. If the UE is configured with RRC parameter srs-TPC-PDCCH- Group = typeB for an uplink carrier without PUCCH and PUSCH or an uplink carrier on which the SRS power control is not tied to PUSCH power control, one or more blocks are configured for the UE by the higher layers. In that case, each block applies to an uplink car- rier and DCI format 2_3 contains 0 or 2 bits of SRS request and 2 bits of TPC command. 4.1.3.2.4 Common and UE-Specific Search Spaces A UE may be configured with one or more CORESETs (using UE-specific or common sig- naling) with a maximum of three CORESETs per configured dow

nlink BWP. Note that the 466 Chapter 4 scheduling flexibility may not be impacted by limiting the maximum number of CORESETs since different monitoring occasions can be configured flexibly even in associa- tion with the same CORESET. It is important to further note that the concept of PDCCH monitoring periodicity is defined per search space set and is not configured at the CORESET-level. Every configured search space with a certain monitoring periodicity (in terms of slots and starting symbols within the monitored slots) is associated with a CORESET. In LTE, the DCI format was closely coupled with the DCI size and monitoring for a certain DCI format in most cases implied monitoring for a new DCI size. In NR, the DCI formats and DCI sizes are decoupled. Different formats can have different DCI sizes, but several formats can share the same DCI size. This allows adding more formats in the future without increasing the number of blind decoding attempts. An NR device needs to monitor up to four different DCI sizes: one size used for the fallback DCI formats, one for downlink scheduling assignments, and unless the uplink downlink non-fallback formats are size-aligned, one for uplink scheduling grant. In addition, a device may need to monitor SFI and/or preemption indication DCIs using a fourth size, depending on the configuration. An NR UE needs to monitor the PDCCH candidates at multiple aggregation levels for the detection and reception of PDCCH. Inside a configured CORESET, NR SS defines the PDCCH candidates of each AL [8]. In NR, PDCCH employs DM-RS-based transmission. Unlike LTE PDCCH where cell-specific reference signals were used for coherent demodu- lation, the NR channel estimation complexity scales with the number of CCEs being monitored. Thus it is important to balance the scheduling flexibility against UE imple- mentation burden to facilitate cost-efficient UE implementation. For PDCCH DM-RS in a CORESET, the antenna port QCL configuration relating to the SS/PBCH block antenna port(s) or configured CSI-RS a

ntenna port(s), is on a per-CORESET basis. This implies that in mmWave deployments, which rely on beam-sweeping operations, different CORESET and search space configurations corresponding to different received beams are necessary [8,53]. The CCE structure described in the previous section helps reduce the number of blind decod- ing attempts; however, it is required to have mechanisms to limit the number of PDCCH can- didates that the device is expected to monitor. From a scheduling point of view, restrictions in the allowed aggregations are undesirable as they may reduce the scheduling flexibility and require additional processing at the transmitter side. At the same time, requiring the device to monitor all possible CCE aggregations in all configured CORESETs significantly increases device complexity and power consumption. A search space is a set of candidate control chan- nels comprising a set of CCEs at a given aggregation level, which the device is supposed to monitor and decode. Due to multiple aggregation levels, a device can have multiple search New Radio Access Physical Layer Aspects (Part 2) 467 spaces. There can be multiple SSs using the same CORESET or multiple CORESETs config- ured for a device. A device is not expected to monitor PDCCH outside its active bandwidth part. At a configured monitoring occasion for a search space, the devices will attempt to decode the candidate PDCCHs for that search space. Five different aggregation levels corre- sponding to 1, 2, 4, 8, and 16 CCEs can be configured. The highest aggregation level is meant to support extreme coverage requirements [6,8,14]. The number of PDCCH candidates can be configured per search space and per aggregation level. When the UE attempts to decode a candidate PDCCH, the content of the control channel is declared as valid, if the CRC checks and the device can successfully process the contained information, that is, scheduling assignment, and uplink grants. If the CRC does not pass, the information is either subject to uncorrectable transmissi

on errors or intended for another UE and in either case the device ignores that PDCCH transmission. The gNB can only address a UE, if the corresponding control information is transmitted on a PDCCH formed by the CCEs in one of the UE's search spaces. Therefore, for efficient utilization of the CCEs in the system, the UE should be associated with different search spaces. Each device in the system can be configured with one or more UE-specific search spaces. Since a UE-specific search space is typically smaller than the number of PDCCHs that the network can transmit at the corresponding aggregation level, there must be a mechanism to deter- mine a set of CCEs in UE-specific search space. One option is to allow the network to figure the UE-specific search space for each device, in the same way that CORESETs are configured. However, this would require explicit signaling exchange with each device and possibly reconfiguration at handover. Instead, the UE-specific search spaces for PDCCH are defined without explicit signaling and based on the device unique identity in the cell in the connected mode, that is, C-RNTI. Furthermore, the set of CCEs that the device should mon- itor at a certain aggregation level varies as a function of time to avoid two devices con- stantly blocking each other. If they collide at one time instant, they are not likely to collide at the next time instant. In each of these search spaces, the UE attempts to decode the PDCCHs using the UE-specific C-RNTI. There is also information intended for a group of UEs in the cell. These messages are scheduled with different predefined RNTIs, for exam- ple, SI-RNTI for scheduling system information, P-RNTI for transmission of a paging mes- sage, RA-RNTI for transmission of the random-access response, TPC-RNTI for uplink power control, INT-RNTI for preemption indication, and SFI-RNTI for slot format configu- ration. As part of random-access procedure, it is necessary to transmit information to a device before it is assigned a unique identity. These types of

information cannot rely on a UE-specific search space as different devices would monitor different CCEs despite the message being intended for all of them. Thus common search spaces are defined, where a common search space is similar in structure to a UE-specific search space with the differ- ence that the set of CCEs is predefined and known to all devices irrespective of their own identity [14]. Chapter 4 Common search space Common search space UE-specific search space (monitoring periodicity (monitoring periodicity Dedicated RRC [RRC configured per serving cell] and offset and number of and offset and number of configuration (monitoring periodicity and offset and candidates per candidates per number of candidates per aggregation level) aggregation level) aggregation level) CORESET CORESET 1 CORESET One slot OFDM symbol No monitoring in Aggregation Level 4 is No monitoring in this occasion of monitored this occasion of the search space the search space CORESETs containing search spaces with aggregation level 2 Figure 4.22 Procedure for PDCCH search space configuration and example search spaces [68]. The number of blind decoding attempts is proportional to subcarrier spacing and the slot duration. For 15/30/60/120 kHz subcarrier spacing, up to 44/36/22/20 blind decoding attempts per slot can be supported across all DCI payload sizes, respectively. It must be noted that the number of blind decoding attempts is not the only measure of UE complexity, channel estimation efforts also need to be taken in consideration. The number of channel estimations for subcarrier spacings of 15/30/60/120 kHz has been limited to 56/56/48/32 CCEs across all CORESETs in a slot [8]. Depending on the configuration, the number of PDCCH candidates may be limited either by the number of blind decoding attempts, or by the number of channel estimates. In order to minimize the device complexity, a device monitors a maximum of three different DCI sizes using C-RNTI and one DCI size using other RNTIs. In carrier aggregation scenarios, the gener

al blind decoding operation described earlier is applied per component carrier. While the total number of channel esti- mates and blind decoding attempts increases compared to the single carrier case, there is no direct proportionality between the number of aggregated carriers and blind decoding attempts [14]. The procedure for PDCCH search space configuration and example search spaces are shown in Fig. 4.22. As we stated earlier, a set of PDCCH candidates are defined for each UE to monitor, which are referred to as PDCCH search spaces. A search space can be categorized as common or UE-specific. In other words, a search space is defined by the PDCCH candidates that need to be monitored. These candidates are determined by a hashing function that operates within New Radio Access Physical Layer Aspects (Part 2) 469 a set of CCEs in a particular CORESET and the monitoring periodicity and offsets that determine when the search space should be monitored (see Fig. 4.22). The UE is required to monitor PDCCH candidates in one or more of the following search spaces [8]: Type0-PDCCH common search space set configured by pdcch-ConfigSIBI in MasterInformationBlock or by searchSpaceSIBI in PDCCH-ConfigCommon or by searchSpaceZero in PDCCH-ConfigCommon for a DCI format, the CRC of which is scrambled with SI-RNTI in the primary cell. Type0A-PDCCH common search space configured searchSpaceOtherSystemInformation in PDCCH-ConfigCommon for a DCI format, the CRC of which is scrambled with SI-RNTI in the primary cell. Type1-PDCCH common search space set configured by ra-SearchSpace in PDCCH- ConfigCommon for a DCI format, the CRC of which is scrambled with RA-RNTI, TC-RNTI, or C-RNTI in the primary cell. Type2-PDCCH common search space set configured by pagingSearchSpace in PDCCH-ConfigCommon for a DCI format, the CRC of which is scrambled with P- RNTI in the primary cell. Type3-PDCCH common search space set configured by SearchSpace in PDCCH-Config with searchSpaceType = common for a DCI format, the CRC of which is scrambled with INT

-RNTI, SFI-RNTI, TPC-PUSCH-RNTI, TPC-PUCCH-RNTI, TPC-SRS-RNTI, C-RNTI, CS-RNTI(s), or SP-CSI-RNTI. UE-specific search space set configured by SearchSpace in PDCCH-Config with searchSpaceType = ue-Specific for a DCI format, the CRC of which is scrambled with C-RNTI, CS-RNTI(s), or SP-CSI-RNTI. An example search space configuration for two devices is shown in Fig. 4.23. The UE determines a CORESET and PDCCH monitoring occasions for Type0-PDCCH common search space set, if it is not provided with RRC parameter searchSpace-SIBI. The Type0- PDCCH common search space set is defined by the CCE aggregation levels and the number of PDCCH candidates per CCE aggregation level. The CORESET configured for this search space set has CORESET index 0 and search space set index 0. If the UE is not provided with a CORESET for any of Type0A-PDCCH/Type1-PDCCH/Type2-PDCCH common search spaces, the corresponding CORESET would be the same as the CORESET for Type0-PDCCH common search space. The CCE aggregation levels and the number of PDCCH candidates per CCE aggregation level for Type0-PDCCH, Type0A-PDCCH, and Type2-PDCCH common search space are given in Table 4.5 [8]. The DM-RS antenna port associated with PDCCH reception in the Type0-PDCCH/Type0A- PDCCH/Type2-PDCCH common search spaces and for the corresponding PDSCH recep- tions as well as the DM-RS antenna port associated with SS/PBCH block reception are quasi-co-located with respect to delay spread, Doppler spread, Doppler shift, average delay, 470 Chapter 4 Physical CCE index Interleaving Logical CCE index 0123456789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Aggregation level 2 Aggregation level 8 PDDCH candidate A (Fully overlapped) Logical CCE index 0123456789 Aggregation level 2 Aggregation level 8 PDDCH candidate B Cyclic shift (Fully overlapped) Aggregation level 2 Aggregation level 8 PDDCH candidate C (Partially overlapped) 36 CCEs in the CORESET, REG bundle is 6, and number of interleaver row is 3 (colored boxes are UE1 and UE2 search spaces) Fi

gure 4.23 Example search space configuration for two devices [ 14]. Table 4.5: CCE aggregation levels and maximum number of PDCCH candidates per CCE aggregation level for Type0/ Type0A/Type2-PDCCH common search space [8]. CCE Aggregation Level Number of Candidates and spatial RX parameters. The value for the DM-RS scrambling sequence initialization is the cell ID. The subcarrier spacing and the cyclic prefix length for PDCCH reception with ype0A-PDCCH/Type1-PDCCH/Type2-PDCCH common search spaces are the same as for PDCCH reception with Type0-PDCCH common search space. The DM-RS antenna port associated with PDCCH reception and the associated PDSCH reception in Type1-PDCCH common search space are quasi-co-located with the SS/PBCH block identified in initial access procedure or with a received CSI-RS with respect to delay spread, Doppler spread, Doppler shift, average delay, and spatial RX parameters [8]. For each downlink BWP configured for a UE in the serving cell, the UE can be provided with NCORESET < 3 CORESETs. For each CORESET, the RRC signaling provides the UE New Radio Access Physical Layer Aspects (Part 2) 471 with a CORESET index 0sp<12; a DM-RS scrambling sequence initialization value; a precoder granularity for a number of REGs in the frequency domain where the UE can assume use of a same DM-RS precoder; a number of consecutive symbols; a set of resource blocks; CCE-to-REG mapping parameters; an antenna port QCL, from a set of antenna port QCLs, indicating QCL information of the DM-RS antenna port for PDCCH reception in a respective CORESET; and an indication for presence or absence of TCI field in DCI format 1_1 transmitted by PDCCH in the CORESET [8]. For each CORESET in a downlink BWP of a serving cell, the RRC parameter frequencyDomainResources provides a bitmap, whose bits have one-to-one correspondence with non-overlapping groups of six PRBs, in ascending order of the PRB index in the down- link BWP bandwidth of N RB BWP PRBs with starting position N start BWP where the first PRB of the first grou

p of six PRBs is indexed as 6 NEW start A group of six PRBs are allocated to a CORESET, if the corresponding bit value in the bitmap is set to one. If the UE receives the initial configuration of more than one TCI state through RRC parameter TCI-States but has not received a MAC CE activation command for at least one of the TCI states, the UE can assume that the DM-RS antenna port associated with PDCCH reception in the UE-specific search space is quasi-co-located with the SS/PBCH block that the UE has identified during the initial access procedure. If the UE has received a MAC CE activation command at least for one of the TCI states, it applies the activation command 3 ms after a slot where it transmits HARQ-ACK information for the PDSCH providing the activation command [8]. Table 4.6 provides the maximum number of PDCCH candidates max (MPDCCH(H)) across all CCE aggregation levels and across all DCI formats with different size in the same search space that the UE is expected to monitor per slot and per serving cell as a function of the subcarrier spacing. The table further provides the maximum number of non-overlapped CCEs that a UE is expected to monitor per slot and per serving cell as a function of the sub- carrier spacing. The CCEs are considered non-overlapped, if they correspond to different CORESET indices or different first symbols for the reception of the respective PDCCH candidates [8]. Table 4.6: Maximum number of PDCCH candidates per slot and per serving cell as a function of subcarrier spacing [8]. Maximum Number of Monitored PDCCH Maximum Number of Non-overlapped Candidates Per Slot and Serving Cell CCEs Per Slot and Serving Cell max(MPDCCH(H)) max(CPDCCH(H)) 472 Chapter 4 For each downlink BWP that is configured for a UE in a serving cell, the UE is provided via RRC signaling with S V 10 search space sets. For each of those search space sets, the UE is provided with an search space set index 0 S < 40; an association between the search space set S and a CORESET p; a PDCCH monitoring periodicity of k

p,s slots and a PDCCH monitoring offset of 8p,s slots; a PDCCH monitoring pattern within a slot, indicating first symbol(s) of the CORESET within a slot for PDCCH monitoring; a number of PDCCH can- didates Mb p,s per CCE aggregation level L; and an indication that search space set S is either a common or a UE-specific search space set via RRC signaling [8]. Alternative PDCCH mapping rules are illustrated in Fig. 4.24. The UE can also be provided via RRC signaling with a time interval consisting of slots indicating a number of slots where the search space set S could exist. The information on the first symbol and the number of consecutive symbols for a CORESET, which results in a PDCCH candidate mapping to symbols of different slots, is not provided to the UE. The UE cannot assume that two PDCCH monitoring occasions, for the same search space set or for different search space sets, within the same CORESET are separated by a number of symbols that are less than the CORESET duration. The UE determines the PDCCH monitoring occasion from the PDCCH monitoring periodic- ity, offset, and pattern within a slot. For search space set S in CORESET p, the UE deter- mines the PDCCH monitoring occasion(s) in a slot with number Nslot in a frame with number nframe, if nslot - 8ps) mod kp,s 0. If the UE is informed in advance of the duration via RRC signaling, it would monitor PDCCH for search space set S in CORESET p for Tp,s consecutive slots, starting from slot nslot and would not monitor PDCCH for search space set S in CORESET p for the next kp,s-Tp,s consecutive slots [8]. A UE-specific search space at CCE aggregation level Le {1, 2, 4, 8, 16} is defined by a set of PDCCH candidates for CCE aggregation level L. For search space set S associated with CORESET p, the CCE indices for aggregation level L corresponding to PDCCH candidate ms,nct of the search space set in slot nslot for an active BWP in the serving cell corresponding to carrier indicator field value NCI are given as follows [8]: Yp,nslot For common search spaces Yp,

nslot whereas for a UE-specific search spaces Yp,nslot = (ApY),, mod D, Yp,-1 NRNTI, Ap = 39827, 39829, or 39839 p mod 3 = 0, 1, or 2, respectively. In the preceding expression, D = 65537; NCCE,p is the number of CCEs, numbered from 0 to NCCE,P - 1, in CORESET p; and NCI denotes the car- rier indicator field value, if the UE is configured via RRC signaling with a carrier indicator field in the serving cell on which the PDCCH is monitored; otherwise, including for any common search space NCI = 0. Furthermore, ms,nc = 0, M(L) p,s,ncl - 1, where M(L) p,s,ncl New Radio Access Physical Layer Aspects (Part 2) 473 CCE 0 CCE 0 REG bundle 0 REG bundle 0 REG bundle 1 CCE 1 REG bundle 1 REG bundle 6 REG bundle 6 REG bundle 7 REG bundle 7 CORESET CORESET REG 0REG 1REG 2 CCE 0 CCE 0 REG bundle 0 REG bundle 0 REG bundle 1 REG bundle 1 REG 3 REG bundle 6 REG bundle 6 REG bundle 7 REG bundle 7 CORESET CORESET Figure 4.24 Illustration of alternative PDCCH mapping rules: (A) PDCCH candidate-level, (B) CCE-level, (C) REG bundle-level, and (D) REG-level. number of PDCCH candidates that the UE is supposed to monitor at aggregation level L in a search space set S in a serving cell corresponding to NCI. For any common search space, max(M(D) p,s,ncl )=MM = whereas for a UE-specific search space, denotes the max- imum of M over the configured values of NCI for a CCE aggregation level L of search p,s,ncl space set S in CORESET p; the RNTI value used for NRNTI is the C-RNTI [8]. Example PDCCH search spaces at various aggregation levels are shown in Fig. 4.25. Physical CCE index 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 2324 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 4 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 Virtual CCE index 4 5 6 7 8 9 10 11 12 13 14 15 16 17.18-19 20 21 22 23 24 25 26 27 28 29 30 31 Aggregation level 16 Aggregation level 8 Aggregation level 4 Aggregation level 2 Aggregation level 1 Figure 4.25 Example PDCCH search spaces at various aggregation levels. New Radio Access Ph

ysical Layer Aspects (Part 2) 475 When the total number of PDCCH candidates that a UE is configured to monitor in a slot exceeds the blind decoding limit of the UE, it must drop some of the candidates. While the number of blind decoding attempts can be controlled by the network through RRC configu- ration, the specification does not define a rule for dropping the PDCCH candidates, when the number of PDCCH candidates exceeds the blind decoding limit of the UE. As an exam- ple, the candidates can be prioritized according to the search space type and then according to search space set number within a search space type and finally according to aggregation level within a search space set. Once the blind decoding limit is reached, the remaining can- didates can be discarded. If a UE has a limit of 56 CCEs and it is configured with a single CORESET in the slot that spans only one OFDM symbol then the CCE limit should not be an issue since the maximum number of PRBs on an RF carrier is 275, and this corresponds to less than 56 CCEs. However, if a UE is configured to monitor two-symbol or three- symbol CORESETs, or multiple CORESETs in a slot and the number of CCEs for all CORESETs is large, the CCE processing constraint can potentially limit the number of PDCCH candidates for which decoding can be attempted, considering that the CCE limit of 56 CCEs must be shared among many CORESETs in the slot. 4.1.3.2.5 Dynamic and Semi-persistent Scheduling The MAC sublayer in a gNB includes dynamic resource schedulers that manage and allo- cate radio resources to active users in the downlink and uplink. Scheduling is performed in either dynamic or semi-static manner. Dynamic scheduling is the default mode-of-operation where the scheduler for each time interval, that is, a slot, determines which devices are going to transmit or receive and further configures the transmission parameters based on the measurement reports from the UEs. Since scheduling decisions are made frequently, it is possible to track fast variations of the user tra

ffic as well as the channel quality, thus effi- ciently utilizing the available resources in order to maximize the network capacity. Semi- static scheduling implies that the transmission parameters are provided to the devices in advance and are not changed on a dynamic basis. The scheduler operation takes into account the UE buffer status and the QoS requirements of each UE as well as the associated radio bearers when assigning radio resources among active UEs (see Fig. 4.26). The schedulers assign network resources to the UEs by consider- ing the radio conditions as seen by the UEs, identified through measurements made at the gNB and/or reported by the UE. The schedulers assign radio resources in a unit of slot, for example, one mini-slot, one slot, or multiple slots and resource assignments consist of radio resources (time, frequency, code, space, power). The UEs identify the allocated resources by receiving a scheduling decision (resource assignment) through PDCCH. The UE periodi- cally or on-demand basis conducts measurements to support scheduler operation. The uplink buffer status reports (measuring the data that is buffered in the logical channel queues in the UE) are used to provide support for QoS-aware packet scheduling. Power headroom reports Chapter 4 Measurements Buffer status Associated radio Scheduling (UE/Network) reports requirements bearer request Scheduler Resource allocation/ Monitor PDCCH DCI (PDCCH)/C-RNTI CS (RRC signaling) CS (RRC signaling) CS (RRC signaling) PDSCH [periodicity] [periodicity] [periodicity] Monitor PDCCH Monitor PDCCH Monitor PDCCH Activating CS DCI (PDCCH)/CS-RNTI PUSCH DCI (PDCCH)/CS-RNTI PDSCH Activating CS No PDCCH PUSCH PUSCH DCI (PDCCH)/C-RNTI PDSCH PDSCH No PDCCH PUSCH PUSCH PDSCH PUSCH DCI (PDCCH)/C-RNTI DCI (PDCCH)/C-RNTI DCI (PDCCH)/C-RNTI DCI (PDCCH)/C-RNTI PDSCH If received overwrites CS If received overwrites CS If received overwrites CS PDSCH Without configured scheduling With configured scheduling (CS) With type 1 configured With type 2 configured scheduling

addressed to a UE using its CS-RNTI indicates that the downlink assignment can be implic- itly reused according to the periodicity defined by the RRC signaling, until deactivated. When a configured downlink assignment is active, if the UE cannot find its C-RNTI on the PDCCH(s), a downlink transmission according to the configured downlink assignment is assumed; otherwise, if the UE finds its C-RNTI on the PDCCH(s), the PDCCH allocation overrides the configured downlink assignment. When carrier aggregation is configured, one configured downlink assignment can be signaled per serving cell. When bandwidth New Radio Access Physical Layer Aspects (Part 2) 477 adaptation is configured, one configured downlink assignment can be signaled per BWP. On each serving cell, there can be only one configured downlink assignment active at a time, and multiple configured downlink assignment can be simultaneously active on differ- ent serving cells. Activation and deactivation of configured downlink assignments are inde- pendent among the serving cells [11]. In the uplink, the gNB can dynamically allocate resources to the UEs by scrambling their respective C-RNTI with PDCCH(s) CRCs. A UE always monitors the PDCCH(s) in order to find possible grants for uplink transmission when its downlink reception is enabled where the activity is synchronized with the UE DRX cycles. When carrier aggregation is config- ured, the same C-RNTI applies to all serving cells. In addition, with configured grants, the gNB can allocate uplink resources for the initial HARQ transmissions to the UEs. Two types of configured uplink grants are defined in NR: Type 1, where the RRC signaling directly provides the configured uplink grant (including the periodicity); and Type 2, where the RRC signaling defines the periodicity of the configured uplink grant while PDCCH addressed to the UE using its CS-RNTI can either signal and activate the configured uplink grant, or deac- tivate it, that is, a PDCCH addressed to the UE using its CS-RNTI would indicate that the upl

ink grant can be implicitly reused according to the periodicity defined by RRC, until deactivated (see Fig. 4.26). When a configured uplink grant is active, if the UE cannot find its C-RNTI/CS-RNTI on the PDCCH(s), an uplink transmission according to the configured uplink grant can be attempted. Otherwise, if the UE finds its C-RNTI/CS-RNTI on the PDCCH(s), the PDCCH allocation overrides the configured uplink grant. Retransmissions other than repetitions are explicitly allocated via PDCCH(s). When carrier aggregation is configured, one configured uplink grant can be signaled per serving cell. Similarly, when bandwidth adaptation is configured, one configured uplink grant can be signaled per BWP. In each serving cell, there can be only one configured uplink grant active at a time. A config- ured uplink-grant for one serving cell can either be of Type 1 or Type 2. For Type 2, activa- tion and deactivation of configured uplink grants are independent among the serving cells. When SUL is configured, a configured uplink grant can only be signaled for one of the two uplink carriers of the cell [11]. 4.1.4 Synchronization Signals In order to connect/attach to the network, a UE must perform initial cell search and down- link synchronization. The objective of initial cell search is to find a strong cell signal for connection establishment, to obtain an estimate of frame timing, to obtain cell identification, and to find the reference signals for coherent demodulation of PBCH and PDCCH. For this purpose, the PSS and SSS are used. The PSS and SSS are transmitted in SSBs together with PBCH. The blocks are transmitted per slot at a fixed slot location. During initial cell search, the UE correlates the received signals and the synchronization signal sequences by means 478 Chapter 4 Secondary synchronization signal Primary synchronization signal Physical layer cell identity (PCI) Figure 4.27 Derivation of PCI based on PSS and SSS sequences [6]. of matched filtering and attempts to locate the PSS in order to obtain symbol and hal

f- frame timing. It then attempts to find the SSS in order to detect the cyclic prefix length as well as the duplexing scheme and to obtain the exact frame timing based on matched filter results for the PSS and SSS. It then proceeds to detect the cell identity from the reference signals sequence index and to decode the PBCH for the purpose of obtaining the minimum SI. The synchronization signal (SS) blocks are organized into SS bursts and SS bursts are organized into SS burst sets that are periodically transmitted in order to support beamform- ing operation. In NR, there are 1008 unique physical-layer cell identities, that is, an increased number compared to 504 in LTE in order to provide sufficient deployment flexibility in dense net- work topologies. As shown in Fig. 4.27, the NR physical-layer cell identities are in 336 unique physical-layer cell-identity groups, each group containing three unique identities. Each NR cell ID can be jointly represented by a PSS/SSS combination. The PSS consists of three frequency-domain binary BPSK length-127 M-sequences, and the SSS corresponds to 336 Gold sequences with length-127. Both of these signals are mapped into 127 contiguous subcarriers. A physical-layer cell identity is uniquely defined by a number in the range of 0-335, representing the physical-layer cell-identity group, and a number in the range of 0-2, representing the physical-layer identity within the physical-layer cell-identity group as Maximum length sequences are pseudo-random binary sequences that are generated using maximal linear feedback shift registers. The M-sequences are periodic and reproduce every binary sequence that can be reproduced by the shift registers (i.e., for length-m registers, they produce a sequence of length 2m - 1). An M-sequence is spectrally flat with the exception of a near-zero DC term. Since M-sequences are periodic and shift registers cycle through every possible binary value with the exception of the zero vectors, the registers can be initialized to any state with the excepti

on of the zero vectors. A binary polynomial over GF(2) can be associated with the linear feedback shift register. The degree of polynomial is equal to the length of the shift register and the coefficients that are either 0 or 1 correspond to the taps of the register. New Radio Access Physical Layer Aspects (Part 2) 479 PSS/SSS sequence 8 Subcarriers 9 Subcarriers Frequency 132133134 135 136 137 138 139 140 141 142143 Figure 4.28 Illustration of PSS/SSS sequence mapping to the resource elements [6]. in the following formula where Nin e{0,1,..., 335} and The cell number information is carried in PSS, whereas the cell group number is carried in SSS. It is worthwhile mentioning that the physical beams associated with an SS blocks are transparent to the UE, since the latter only sees the equivalent synchronization signals and the PBCH after precoding and/or beamforming operations that are implementation specific. 4.1.4.1 Primary Synchronization Sequence The PSS sequence dpss(n) is defined by dpss(n) = 2x(m) where 127 and 0<n<127. In the latter expression, e(i+7)=[x(i+4)+x(i)]mod 2 and [x(6) x(5) x(4) x(3) x(2) x(1) x(0)] = [ 110110]. The sequence symbols (126) containing the PSS is scaled by a factor of BPSS in order to adjust its transmission power and is mapped to resource elements (k,l) in increasing order of k where k and l represent the frequency and time indices, respectively (see Table 4.3), within one SS/PBCH block. The PSS sequence mapping to resource elements in the fre- quency domain is illustrated in Fig. 4.28. Furthermore, the time-frequency structure and timing of the PSS transmission are depicted in Fig. 4.29. 4.1.4.2 Secondary Synchronization Sequence The secondary synchronization signal sequence dsss(n) is defined as dsss(n) = and 0n<127. In the latter expression, x1(i+7)= [x1(i+1)+xx(i)]m where xo(6) xo(5) xo(4) x1(6) x1(5) x1(4) x1(3) x1(2) x1(1) x1(0) sequence of symbols dsss(0), dsss(126) containing the secondary synchronization signal is scaled by a factor of BSSS and is mapped to resource elemen

ts (k,l) in increasing order of k where k and l represent the frequency and time indices, respectively (see Table 4.3), New Radio Access Physical Layer Aspects (Part 2) 481 SS burst set periodicity (20 ms) SS burst set SS burst SS burst SS burst SS burst 5ms [Half frame] window Figure 4.30 SS block structure and timing [6,8]. within one SS/PBCH block [6]. The SSS sequence mapping to resource elements in the fre- quency domain is illustrated in Fig. 4.28. Furthermore, the time-frequency structure and timing of the SSS transmission are depicted in Fig. 4.29. 4.1.4.3 Synchronization Signal Blocks In NR, the primary and secondary synchronization signals are used by the UE for initial cell search and to obtain frame timing, Cell ID, and to find the reference signals for coherent demodulation of other channels. The PSS, SSS, and PBCH are time-multiplexed and trans- mitted in an SSB with the same numerology. One or more SS block(s) constitute an SS burst, and one or more SS bursts form an SS burst set as illustrated in Fig. 4.30. The SS burst sets are transmitted periodically. An SS block consists of four consecutive OFDM symbols. Regardless of the SS burst set composition, the transmission of SSBs within an SS burst set is confined to a 5 ms window to help the UEs reduce power consumption and com- plexity for radio resource management-related measurements. Fig. 4.31 compares the syn- chronization signals and broadcast channel transmission timings of LTE and NR. The SS block is transmitted periodically with a period which may be configured between 5 and 160 ms. However, the UEs performing initial cell search or handover can assume that the SS block is repeated every 20 ms. Each SS burst set is always confined to a 5 ms win- dow located either in the first or the second half of a 10 ms radio frame. This allows a UE that is searching for an SS block in the frequency domain to know the time duration it should pause at each frequency before retuning to the next frequency within the synchroni- zation raster, concluding that

there is no PSS/SSS present at that frequency. The 20 ms SS block periodicity is four times longer than the corresponding 5 ms periodicity of LTE PSS/ SSS transmission (see Fig. 4.31). The longer SS block period was selected to improve the NR network energy efficiency and to reduce the layer-1 overhead. The disadvantage of a NR SS/PBCH block locations in time and frequency (configurable locations based on subcarrier spacing, frequency range, and other higher layer parameters) SS block periodicity Offset relative to center frequency RF carrier center frequency Frame = 10 ms Frame = 10 ms Frame = 10 ms Frame = 10 ms LTE PSS, SSS, and PBCH locations in time and frequency (fixed timing, periodicity, and frequency Locations) Frame = 10 ms Frame = 10 ms- Frame = 10 ms Frame = 10 ms RF carrier center frequency PSS/SSS periodicity (5 ms) PBCH periodicity (40 ms) Figure 4.31 Comparison of LTE and NR synchronization and broadcast channel transmission timing. New Radio Access Physical Layer Aspects (Part 2) 483 OFDM symbol Case A (SCS = 15 kHz) Case B (SCS= 30 kHz) Case C (SCS 30 kHz) Case B (SCS 30 kHz) Coexistence 0.125ms Case D (SCS= 120 kHz) Case E (SCS 240 kHz) Figure 4.32 Structure and timing of SS/PBCH block transmission with various numerologies [6,31]. longer SS block period is that a device must pause at each frequency for a longer time in order to conclude that there is no PSS/SSS at the frequency. However, this is compensated by the sparse synchronization raster relative to LTE, which reduces the number of frequency-domain locations at which a device must search for an SS block. The maximum number of SSBs within an SS burst set is 4 for frequency ranges up to 3 GHz, 8 for 3-6 GHz, and 64 for above 6 GHz in order to achieve a trade-off between coverage and layer-1 overhead. Furthermore, the number of actually transmitted SSBs could be less than the maximum number. The position(s) of actually transmitted SSBs can be sig- naled to the UEs in order to assist their RRC_CONNECTED or RRC_IDLE mode measure- ments and to

help the UEs in RRC_CONNECTED and potentially RRC_IDLE mode to receive downlink data/control in unused SSBs. The structure and timing of SSB transmis- sion with various numerologies is illustrated in Fig. 4.32, where a number of symbols are reserved for downlink control at the beginning of the slot, and some symbols are reserved for guard period and uplink control to allow UL/DL switching and fast uplink feedback. The SSB pattern corresponding to 15 and 30 kHz subcarrier spacing can provide more UL/ DL switching opportunities in TDD mode. The SSB pattern for 30 kHz subcarrier spacing can be used to facilitate LTE-NR coexistence in the downlink in an FDD system, consider- ing the locations of LTE PDCCH and cell-specific reference signals in the symbols with LTE default 15 kHz subcarrier spacing. 484 Chapter 4 Within a PBCH transmission time interval update period of 80 ms, there are 16 possible positions for an SS burst set, if we consider the minimum period for an SS burst set to be 5 ms. The 16 possible positions of an SS burst set can be identified by the three least sig- nificant bits of the SFN and one-bit half radio frame index. The SSBs can be repeated within an SS burst set. When the UE detects an SSB, it will acquire the timing informa- tion from its PBCH, from which the UE is able to identify the radio frame number, the slot index in a radio frame, and OFDM symbol index in a slot. The timing information includes 10 bits for SFN, 1 bit for half radio frame index, and 2, 3, or 6 bits for SSB time index for frequency ranges up to 3, 3-6, and 6-52.6 GHz, respectively. Within the SSB indices, two or three LSBs are carried by changing the DM-RS sequence of PBCH. Thus, for the sub-6 GHz frequency range, the UE can acquire the SSB index without decoding the PBCH. It also facilitates PBCH soft combining over multiple SSBs as these SSBs with different indices carry the same PBCH payload [72]. As shown in Fig. 4.32, for a half frame with SS/PBCH blocks, the first symbol indices of the candidate SS/PBCH blocks are

determined according to the subcarrier spacing of SS/PBCH blocks as follows, where index 0 corresponds to the first symbol of the first slot in a half-frame [8]: Case A (SCS = 15 kHz): The first symbols of the candidate SS/PBCH blocks have indi- ces of {2,8} + 14n where n = 0, 1 for carrier frequencies fc < 3 GHz and n = 0, 1, 2, 3 for carrier frequencies 3 GHz V fc < 6 GHz. Case B (SCS = 30 kHz): The first symbols of the candidate SS/PBCH blocks have indi- ces {4, 16, 20} + 28n where n = 0 for carrier frequencies fc < 3 GHz, n = 0, 1 for car- rier frequencies 3 GHz VI fc < 6 GHz. Case C (SCS = kHz): The first symbols of the candidate SS/PBCH blocks have indi- ces {2,8} + 14n where n = 0, 1 for paired spectrum operation and carrier frequencies fc < 3GHz and n = 0, 1, 2, 3 for carrier frequencies 3 GHz fc < 6 GHz. For unpaired spectrum operation, = 0, 1 for carrier frequencies fc < 2.4 GHz, n = 0, 1, 2, 3 for car- rier frequencies 2.4 GHz fc < 6 GHz. Case D (SCS = kHz): The first symbols of the candidate SS/PBCH blocks have indices {4, 8, 16, 20} + 28n where n = 0, 1, 2, 3, 5, 6, 7, 8, 10, 11, 12, 13, 15, 16, 17, 18 for carrier frequencies fc 6 GHz. Case E (SCS = 240 kHz): The first symbols of the candidate SS/PBCH blocks have indi- ces {8, 12, 16, 20, 32, 36, 40, 44} + 56n where n = 0, 1, 2, 3, 5, 6, 7, 8 for carrier frequen- cies fc 6 GHz. In order to support multi-beam operation, particularly in high-frequency scenarios, NR introduced the SSB, which comprises primary and secondary synchronization signals and PBCH. As illustrated in Fig. 4.33, a given SSB is repeated within an SS burst set, which is potentially used for gNB beam-sweeping transmission. The SS burst set is confined to a 5 ms window and transmitted periodically. For initial cell selection, the UE assumes a default SS burst set periodicity of 20 ms. The main advantage of SS burst set is that SS block periodicity Frame = 10 ms Frame = 10 ms Frame = 10 ms Frame = 10 ms SSB Index (mapped to each beam) 0 Beams in space as UEs perceive them Signal streng

th at UE1 Signal strength at UE2 Figure 4.33 Example beam sweeping and correspondence to the transmission of SS burst set [30]. 486 Chapter 4 time-division multiplexing beam-sweeping allows for low-cost analog antenna array archi- tectures. Frequency-division multiplexing is another approach that could have been poten- tially adopted in NR; however, it would have precluded use of analog antenna array architectures. This feature, although particularly useful for mmWave operation, can also be leveraged at lower frequency bands. In LTE, the PSS/SSS and PBCH are always located at the center of the RF carrier. Thus once an LTE device detects the PSS/SSS, it has already found the center frequency of the carrier. The drawback of this approach is that a device with no a priori knowledge of the frequency-domain carrier position must search for PSS/SSS at all possible carrier positions. To allow a faster cell search in NR, the possible SS block locations for each frequency band are a limited set of frequencies referred to as the synchronization raster. Therefore, instead of searching for an SS block at each carrier raster, a UE only needs to search for an SS block within the sparse set of synchronization raster. Since NR carriers can still be located at an arbitrary position on the carrier raster, the SS block may not be necessarily located at the center of a carrier, and it may not be aligned with the resource block grid due to different numerologies. Thus once the SS block has been detected, the device must be explicitly informed about the exact SS block frequency-domain position on the carrier. This is achieved by means of information partly within PBCH and partly within the RMSI. As we mentioned earlier, an SS/PBCH block consists of four OFDM symbols in the time domain, numbered in increasing order from 0 to 3 within the SS/PBCH block, where PSS, SSS, and PBCH with the associated DM-RS are mapped to symbols as shown in Fig. 4.29. In the frequency domain, an SS/PBCH block consists of 240 contiguous subcarriers with the

subcarriers numbered in increasing order from 0 to 239 within the SS/PBCH block. There are two types of SS/PBCH block, that is, Type A and Type B, where the former is specified for operation in sub-6 GHz frequency range and the latter is defined for mm Wave bands. The frequency-domain location of SS/PBCH block is defined by parameter kssb which provides the subcarrier offset from subcarrier 0 in common resource block NSSB CRB subcarrier 0 of the SS/PBCH block. The common resource block parameter NSSB CRB is derived from the RRC parameter offsetToPointA. The four LSBs of kssb are derived from the RRC parameter ssb-SubcarrierOffset where, for SS/PBCH block Type A, the most significant bit of kssb is given by anmin+5 in the PBCH payload [6,7]. If ssb-SubcarrierOffset is not pro- vided, kssb is derived from the frequency difference between the SS/PBCH block and Point A. The complex-valued symbols corresponding to resource elements that are part of a com- mon resource block partially or fully overlap with an SS/PBCH block and are not used for SS/PBCH transmission. For an SS/PBCH block, a single-antenna port and the same cyclic prefix length and subcarrier spacing are used for transmission of PSS, SSS, PBCH, and DM-RS for PBCH. For SS/PBCH block Type A, UE {0, 1} and kssb E {0, 1, 2, ., 23} with the quantities kssb and N CRB SSB expressed in terms of 15 kHz subcarrier spacing. For New Radio Access Physical Layer Aspects (Part 2) 487 SS/PBCH block Type B, HE {3,4} and kssb E {0, 1, 2, 11} where the quantity kssb expressed in terms of the subcarrier spacing provided by the RRC parameter subCarrierSpacingCommor and N CRB SSB is defined in terms of 60 kHz subcarrier spacing. The center of subcarrier 0 of resource block N CRB SSB coincides with the center of subcarrier 0 of a common resource block with the subcarrier spacing provided by the RRC parameter subCarrierSpacingCommon. This common resource block overlaps with subcarrier 0 of the first resource block of the SS/PBCH block. The SS/PBCH blocks are transmitted with the

same block index on the same center frequency location which are quasi-co-located with respect to Doppler spread, Doppler shift, average gain, average delay, delay spread, and spa- tial RX parameters (when applicable) [6]. 4.1.5 Physical Downlink Shared Channel The downlink physical layer processing of transport channels consists of several steps as shown in Fig. 4.34 including CRC calculation and attachment to the TBs where a 24-bit CRC for payloads larger than 3824 bits or otherwise 16-bit CRC is attached; code block segmentation and code block CRC attachment; channel coding based on LDPC codes; physical-layer HARQ processing and rate matching; bit-interleaving; modulation; layer mapping and precoding; and mapping to assigned resources and antenna ports. At least one symbol with DM-RSs is present on each layer in which PDSCH is transmitted to a UE. The number of DM-RS symbols and RE mapping is configured by the RRC parameters. The PT-RS may be transmitted on additional symbols to aid receiver phase tracking. As shown in Fig. 4.34, in the first stage of PDSCH processing, the entire TB ao, a1 ANTB-1 is used to calculate the CRC parity bits Po, p1, PLcRc-1 where and LCRC denote the (TB) payload size and the number of CRC parity bits, respectively. The number of parity bits is set to 24 and the CRC generator polynomial gCRC24A(D) = [D24 D23 + D18+D17 D14 + D11 D10 + D7 + D6 + D5 + D4 + D3 D + 1] is used, however, if NTB 3 8824, LCRC is set to 16 bits and the generator polynomial gCRC16(D) = [D16 + D12 + D5 + is used. The output bits following the CRC attachment denoted by bo, b1, bB-1 where B = NTB + LCRC. For initial transmission of a TB with coding rate R, which is determined by the MCS index contained in the DCI, and the retrans- missions of the same TB, each code block of the TB is encoded with LDPC base graph 2 (see the section on channel coding), if NTB VI 292, if NTB < 3824 and R VI 0.67 or if R V 0.25; otherwise, LDPC base graph 1 is used as depicted in Fig. 4.35. As shown in Fig. 4.35, the maximum size of

a code block Kcb is 8448 bits for LDPC base graph 1, 3840 for LDPC base graph 2. The code blocks whose size exceeds these limits would be segmented and appended with an additional CRC of length LCRC = 24 bits. The input bits to the code block segmentation denoted as bo, b1 where number of bits in the TB (including the CRC), are then processed by code block segmentation 488 Chapter 4 PDSCH data (transport block) b(cw) Transport block CRC attachment Scrambling TS 38.212 Section 7.2.1 TS 38.211 Section 7.3.1.1 bo,b,...BB-1 b (cw) (0),b (cw) (1), (N LCW) LDPC base graph selection Modulation TS 38.212 Section 7.2.2 TS 38.211 Section 7.3.1.2 d(cw) (0), Code block segmentation and code block CRC attachment Layer mapping TS 38.211 Section 7.3.1.3 TS 38.212 Section 7.2.3 Channel coding Antenna mapping TS 38.212 Section 7.2.4 TS 38.211 Section 7.3.1.4 Rate matching Resource mapping TS 38.212 Section 7.2.5 mapping to VRB TS 38.211 Section 7.3.1.5 Code block concatenation Resource mapping TS 38.212 Section 7.2.6 VRB-to-PRB mapping TS 38.211 Section 7.3.1.6 go, ,81 8G-1 Figure 4.34 Physical layer processing of PDSCH [6,30]. R(bit/s) Base graph 1 Base graph 2 K(bit) Figure 4.35 Usage of NR LDPC base graphs [55]. New Radio Access Physical Layer Aspects (Part 2) 489 Table 4.7: NR low-density parity check code lifting factors [7]. Set Indexiis Set of Lifting Sizes Z {2,4,8,16,32,64,128,256} {3,6,12,24,48,96,192,384} {5,10,20,40,80,160,320} {7,14,28,56,112,224} {9,18,36,72,144,288} {11,22,44,88,176,352} {13,26,52,104,208} {15,30,60,120,240} Table 4.8: NR low-density parity check base graphs parameters [7]. Parameter Base Graph 1 Base Graph 2 Minimum code rate Rmin Base matrix size 46 X 68 42 X 52 Number of systematic columns Maximum information block size Kcb 8448( = 22 (384) 3840( = X384) Number of non-zero elements followed by code block CRC attachment, resulting in the output bits where index l <0 represents the code block number and K denotes the number of bits for 1th code block. The total number of code blocks C is determin

ed C = B/(Kcb-LCRC)]. The code blocks are then fed into the channel coding unit. The LDPC encoded bits are denoted by dio, du d(N,-1) where the value of N1 is calculated as follows: If the bit sequence input for a given code block to channel coding is denoted by co, C1,..., CK-1 where K is the number of bits to encode, and the LDPC encoded bits are denoted by do, d1,...,dn-1 then N = 66Z for LDPC base graph 1 and N = 50Z for LDPC base graph 2, where the lifting factor Z is given in Table 4.7 [7]. The NR LDPC base graphs parameters are shown in Table 4.8. The rate matching for LDPC code is performed on code block basis and consists of bit selection and bit-level interleaving. The input bit sequence to rate matching block is denoted as do, d1,..., dn which is written into a circular buffer of length Ncb for code block l, where code length N was defined earlier. Let us assume Ncb = N for the 1th code block, if ILBRM = =07 and in other cases Ncb = min(N, NREF), in which NREF = TBSLBRM/(RLBRMC), C is the number of code blocks of the transport block, RLBRM = 2/3 and TBSLBRM for DL- SCH/PCH is obtained from Table 4.16, taking into consideration the maximum number of layers for one TB supported by the UE in the serving cell; the maximum modulation order Limited-buffer rate matching (LBRM) is a technique to process HARQ with reduced requirements for soft buffer sizes while maintaining the peak data rates. LBRM shortens the length of the virtual circular buffer of the code block segments for certain large sizes of transport blocks, thus sets a lower bound on the code rate. 490 Chapter 4 configured for the serving cell, if configured by higher layers; otherwise, a maximum modu- lation order of Qm = 6 is assumed for DL-SCH; and the maximum coding rate of 948/1024. Due to unequal amplitude of demodulated log likelihood ratios (LLRs) for 16QAM/ 64QAM/256QAM modulated symbols, it is necessary to consider a bit interleaving scheme for high-order modulations (see Fig. 4.37) in order to enhance the performance of the LDPC codes. T

he output bit sequence of the bit-interleaving function is the input to code block concatenation. The code block concatenation consists of sequentially concatenating the rate-matched outputs of different code blocks. The output bit sequence of code block concatenation is denoted by go,g1, gG-1 where G is the total number of coded bits for transmission [7]. Rate matching is performed separately for each code block by puncturing a fraction of sys- tematic bits. Depending on the code block size, the fraction of punctured systematic bits can be up to one-third of the systematic bits. The remaining coded bits are written into a circular buffer, starting with the non-punctured systematic bits and continuing with parity bits as shown in Fig. 4.36. The selection of the bits for transmission is based on reading the Transport block (TB) Code block segmentation Code block (CB) Code block (CB) Code block (CB) CRC addition Code block (CB) Code block (CB) Code block (CB) Zero padding Code block (CB) LDPC encoding Code block (CB) Parity bits Punctured Rate matching Systematic Systematic Parity bits Parity bits Transmitted bits Puncturing Repetition Systematic Systematic Systematic Systematic Parity bits Parity bits Parity bits Parity bits 1st Transmission 2nd Transmission 3rd Transmission 4th Transmission Figure 4.36 Example of rate-matching and code block concatenation processes 14]. New Radio Access Physical Layer Aspects (Part 2) 491 Circular Bit-level Code block encoded buffer interleaver concatenation LDPC encoded bits Bits for the 1st Bits for the 2nd Bits for the 3rd Bits for the 4th 1st 2Z group group group group 256QAM 256QAM 256QAM 256QAM symbol 0 symbol 0 symbol 0 symbol 0 Figure 4.37 Example bit-level interleaving for 256QAM modulation. Table 4.9: Starting position of different redundancy versions [7]. LDPC Base Graph 1 LDPC Base Graph 2 17Ncb 13N cb 25Nc6 43Nch required number of bits from the circular buffer where the exact set of bits to transmit depends on the RV corresponding to different starting positions in

the circular buffer. Thus by selecting different RVs, different sets of coded bits representing the same set of informa- tion bits can be generated, which is used when implementing HARQ with incremental redundancy. The starting points in the circular buffer (RVO,RV1,RV2,RV3) are defined such that both RV0 and RV3 codes are self-decodable which means that they include the systematic bits under typical conditions. The RV index of the incremental redundancy HARQ in NR is derived differently compared to LTE. Unlike LTE that RV index positions are sequentially incremented in the circular buffer, in NR, if rvid = 0, 1, 2, 3 denotes the RV number for the current transmission, the rate matching output bit sequence {eklk = 0, 1, , E - 1} is generated as ek = d(ko+j)mod Ncb where ko is given by Table 4.9 according to the value of rvid and the LDPC base graph lifting factor [7,14]. 492 Chapter 4 Transport block (TB) Code block segmentation Code block (CB) Code block (CB) Code block (CB) Code block (CB) Code block grouping Code block group 0 Code block group 1 Code block group n Figure 4.38 CBG-based retransmissions [32]. It is possible to perform HARQ retransmissions with a code block granularity. In that case the information included in the DCI would determine the code block group (CBG) which is (re)transmitted, and information about handling the CBGs for soft-buffer/HARQ combining purposes (see Fig. 4.38). The NR supports code block group-based transmission with single or multi-bit HARQ-ACK feedback. For the case of CBG-based retransmission, HARQ- ACK multiplexing is supported. The motivation for CBG-based retransmission is to improve the spectrum efficiency because if CBG-based retransmission is configured, the HARQ feedback is provided per CBG and only the erroneously received code block groups are retransmitted. This can consume less radio resources than retransmitting the entire TB. If the retransmission is caused due to low SNR then combining in the soft-buffer would improve the decoding quality during retransmission

s; however, if the retransmitted code block was affected by preemption, the buffer content is not correct, and it is better to dis- card the content of the buffer and to request a fresh transmission. If a UE is configured to receive CBG-based transmissions when the RRC parameter codeBlockGroupTransmission set for PDSCH, it determines the number of CBGs for a PDSCH transmission by calculating M = min(NCBG) C) where NCBG denotes the maximum number of CBGs per TB which is configured by RRC parameter maxCodeBlockGroups PerTransportBlock for PDSCH, and C is the number of code blocks. We define M1 = mod M, K = C/M and K2 = C/M If M1 > 0, the mth CBG when m=0,1, ., M1 - 1 consists of code blocks with indices mK + k, k = 0, 1, K1 The mth CBG when m : M1, M1 + 1, M - 1 consists of code blocks with indices M K + ( m - M1)K + k, = 0, 1, K2 - 1. If a UE is configured with CBG-based retransmissions, the scheduling assignment for the UE would contain the necessary HARQ-related control signaling including process number, new-data indicator, CBG trans- mit indicator (CBGTI), and the CBG flush indicator (CBGFI) as well as information to han- dle the transmission of the HARQ acknowledgment in the uplink such as timing and New Radio Access Physical Layer Aspects (Part 2) 493 resource indication information. Upon receiving a scheduling assignment in the DCI, the receiver would attempt to decode the TB. Since transmissions and retransmissions are scheduled using the same framework, the UE needs to know whether this is a new transmis- sion, in that case the soft buffer should be flushed, or a retransmission, where soft combin- ing should be performed. Therefore, a single-bit new data indicator is included as part of the scheduling information. The new data indicator operates at TB level. However, if CBG- based retransmissions are configured, the device needs to know which CBGs are retrans- mitted and whether the corresponding soft buffer should be flushed. This is handled by additional information fields in the DCI when CBG-based retr

ansmissions are configured, that is, CBGTI and CBGFI fields. The CBGTI is a bitmap indicating whether a certain CBG is present in the downlink transmission. The CBGFI is a single bit field, indicating whether the CBGs identified by CBGTI should be flushed or whether soft combining should be performed. The decoding operation results in either a positive acknowledgment in the case of a successful decoding or a negative acknowledgment in the case of unsuccessful decoding, and it is fed back to the gNB as part of the uplink control information. If CBG- based retransmissions are configured, a bitmap with one bit per CBG is fed back instead of a single bit representing the entire TB. The uplink uses the same asynchronous HARQ pro- tocol as the downlink. The necessary HARQ-related information including process number, new-data indicator, and CBGTI (when configured) are included in the scheduling grant [9]. CBG-based retransmissions are transparent to the MAC sublayer and are handled in the physical layer despite being part of the HARQ mechanism. From the MAC perspective, the TB is not correctly received until all CBGs are correctly received and decoded. It is not pos- sible to combine transmission of new CBGs associated with another TB with retransmis- sions of CBGs belonging to the incorrectly received TB in the same HARQ process [14]. The bit sequence fo,f1, fe-1 is generated by interleaving bit sequence CO,C1,...,EE-1 according to the value of the modulation order Qm as follows Aj=0,1,... E/Qm 1;i=0,1, Qm - 1 [7]. The new radio supports up to two code- words in the downlink transmission. For each codeword cw, the block of bits b(cw)(N(cw) where N(cw) denotes the number of bits in codeword CW transmitted on the physical shared channel, is scrambled prior to modulation, resulting in a block scrambled (N(cw) such 6(cm) (i) = +c(c))(i) mod 2. The scrambling sequence a generic pseudo- random length-31 Gold sequence that is initialized by setting Cinit nRNTI215 cw214. where NIDE {0, 1, 1023}, if configured through RRC sign

aling; otherwise NID Ncell NRNTI corresponds to the RNTI associated with the PDSCH transmission [6]. The scrambled bit sequence (N(CW) 1) is modulated using one of the modulation schemes, for example, QPSK, 16QAM, 64QAM, or 256QAM. The complex- valued modulation symbols for each of the codewords that are going to be transmitted are mapped to one or several layers for spatial multiplexing. As shown in Table 4.10, the (E+!)() (!)(E)* (!E)(O)==()()) (Z+!E)(1)P: = (1)(s)* (L+!E)(U)P=()() (!E)(L)P=(1)(E)X = (L+!E)(OP=()(1) = (1)(+) (!E)(1)P =(1)(z) = (L+!E)(OP=()(1) (!E)(OP=(!)()) Buiddew [9] leneds Buiddew 01 PoMay th the New Radio Access Physical Layer Aspects (Part 2) 495 complex-valued modulation symbols corresponding to codeword CW are mapped to layers x(i) = where V denotes the number of layers and is the number of modulation symbols per layer [6]. The block of vectors 1, mapped antenna ports as follows: The set of antenna ports {po,P1,..., Pv-1} are determined according to the procedure specified in [7]. For each antenna port that is used for transmission of the physical shared channel, the prop- erly scaled block of complex-valued symbols are sequen- tially mapped to the virtual resource elements (k',1) allocated for transmission of PDSCH that have not been designated for reference signals. Any common resource block partially or fully overlapping with an SS/PBCH block is considered occupied and is not used for transmission. The virtual resource elements are mapped in frequency-first manner as illus- trated in Fig. 4.39. The virtual resource blocks are mapped to physical resource blocks in Noninterleaved y(Po'(i) Resource block VRB bundle 1(Pvi) (i) Interleaved f(j)==rc+c; j=cR+r; r = 0,1,..., R - 1; c = 0,1,...,C - 1; Figure 4.39 Mapping to VRBs and from VRBs to PRBs [6]. 496 Chapter 4 form of either non-interleaved or interleaved, wherein the non-interleaved is the default mapping scheme. For non-interleaved VRB-to-PRB mapping, the virtual resource block n is mapped to physi- cal resource block n, except for

PDSCH transmissions scheduled with DCI format 1_0 in a common search space where virtual resource block n is mapped to physical resource block Nstart where CORESET' Nstart CORESET is the lowest numbered physical resource block in the CORESET where the corresponding DCI was received [6]. In interleaved mapping scheme, the mapping process is defined in terms of resource block bundles. The set of Nowp(i) resource blocks in the ith bandwidth part with starting position are divided into Nbundle + (Nowp(i)mod L))]/L resource-block bundles in increasing order of the resource-block number and bundle number where Li is the bundle size for the ith bandwidth part defined by RRC parameter vrb-ToPRB-Interleaver and resource block bundle 0 consists of L - (Nowp(i)mod Li) resource blocks, resource block bundle Nbundle - 1 consists of >0 resource blocks; otherwise, all resource block bundles consists of L resource blocks (except for PDSCH transmissions scheduled with DCI format 1_0 with the CRC scrambled by SI-RNTI in Type0-PDCCH common search space in CORESET 0 and in any common search space other than Type0-PDCCH common search space). The virtual resource blocks in the region je {0, 1, Nbundle - 2} are mapped to the physical resource blocks such that virtual resource block bundle Nbundle - 1 is mapped to physical resource block bundle Nbundle - 1 and virtual resource block bundle {0, 1,..., 2} is mapped to physical resource block bundle where f(j) + c;j = cR + r = 0, 1, ., R - 1;c = 0, 1, ., C - 1;R = 2; and C = Nbundle/R If no bundle size is configured, the UE will assume L 2 with a precoding resource block group (PRG) size of 4 (see Fig. 4.39). 4.1.6 CSI Measurement and Reporting and Beam Management 4.1.6.1 CSI Measurement and Reporting In wireless communications, the channel state information refers to channel properties of a wireless communication link. This information describes how a signal propagates from the transmitter to the receiver and represents the combined effect of scattering, multipath fad- ing, signal power a

ttenuation with distance, etc. The knowledge of CSI at the transmitter and/or the receiver makes it possible to adapt data transmission to current channel condi- tions, which is crucial for achieving reliable and robust communication with high data rates in multi-antenna systems. The CSI is often required to be estimated at the receiver, and usu- ally quantized and fed back to the transmitter. The downlink channel can be estimated from uplink reference signals in TDD systems under certain conditions due to reciprocity. New Radio Access Physical Layer Aspects (Part 2) 497 In general, the transmitter and receiver can observe different CSI. There are two types of CSI, that is, instantaneous CSI and statistical CSI. In instantaneous CSI or short-term CSI the current channel conditions are known, which can be interpreted as knowing the impulse response of a digital filter. This provides an opportunity to adapt the transmit signal to the impulse response and thereby to optimize the received signal for spatial multiplexing or to achieve low bit-error-rates. In statistical CSI or long-term CSI, the statistical characteristics or statistics of the channel are known. The latter information may include the type of fading distribution, the average channel gain, the line-of-sight component, and the spatial correla- tion. Similar to the instantaneous CSI, this information can be used for optimization of transmission parameters. The CSI estimation accuracy is practically limited by how fast the channel conditions are varying. In fast-fading channels where the channel conditions may vary rapidly during transmission of a single information symbol, only statistical CSI is reasonable. On the other hand, in slow-fading scenarios, the instantaneous CSI can be estimated with reasonable precision and used for transmission adaptation for a period of time before becoming obsolete. In practical scenarios, the available CSI is often manifested as instantaneous CSI with some estimation/quantization error combined with some statistical infor

mation. To support diverse use cases, NR features a highly flexible and unified CSI framework, in which there is reduced coupling between CSI measurements, CSI reporting and the actual downlink transmission compared to LTE. The CSI framework can be seen as a toolbox, where different CSI reporting settings and CSI-RS resource settings for channel and inter- ference measurements can be selected, SO that they correspond to the antenna configuration and transmission scheme in use such that the CSI reports on different beams can be dynami- cally triggered. The framework also supports more advanced schemes such as multi-point transmission and coordination. The control and data transmissions follow a self-contained principle, where all information required to decode the transmission (such as accompanying DM-RS) is contained within the transmission itself. As a result, the network can seamlessly change the transmission point or beam as the UE moves in the network. The CSI-RS refer- ence signals are used for CSI acquisition and beam management. The CSI-RS resources for a UE are configured by RRC information elements and can be dynamically activated/deacti- vated via MAC control elements or DCI [57]. The configuration and use of CSI-RS in NR can be defined via the CSI framework. As shown in Fig. 4.40, the basic units of CSI framework in NR are CSI reporting setting and CSI resource setting. The CSI reporting setting is linked to M resource settings for channel and interference measurements (CM and IM), where M = 1 indicates resource setting for channel measurement and beam management; M = 2 indicates resource settings for channel measurements and CSI-interference measurement (CSI-IM) or NZP CSI-RS for interference measurement; and M = 3 is an indication for resource settings for channel measurements and two resource settings for CSI-IM and NZP CSI-RS-based interference measurement. 498 Chapter 4 CSI resource CSI resource sets settings for CM CSI reporting CSI resource CSI resource sets settings 1 settings for IM CSI resour

ce settings for NZP CSI resource sets CSI-RS-based IM CSI reporting settings N CSI resource set 1 CSI-RS resource 1 CSI resource set S CSI-RS resource K Figure 4.40 CSI framework in NR. The above-mentioned resource settings are linked to S resource sets each resource set comprises SS/PBCH block resources for beam management and is linked to CSI-RS resources [9]. The time and frequency resources that can be used by the UE to report CSI are controlled by the gNB. The CSI may consist of CQI, PMI, CSI-RS resource indicator (CRI), SS block resource indicator, layer indication (LI), rank indicator (RI), and/or and L1-RSRP measurements. For CQI, PMI, CRI, LI, RI, L1-RSRP, the UE is configured via RRC signaling with more than one CSI-ReportConfig reporting settings, CSI- ResourceConfig resource settings, and one or two lists of trigger states, indicating the resource set IDs for channel and optionally for interference measurement. Each trigger state contains an associated CSI-ReportConfig [9]. Each reporting setting CSI-ReportConfig is associated with a single downlink BWP and contains the reported parameter(s) for one CSI reporting band including CSI Type-I or II, codebook configuration comprising codebook subset restriction, time-domain behavior, New Radio Access Physical Layer Aspects (Part 2) 499 frequency granularity for CQI and PMI, measurement restriction configurations, LI, reported L1-RSRP parameter(s), CRI, and the SSB resource indicator. The time-domain behavior of the CSI-ReportConfig is determined by RRC signaling and can be set to aperiodic, semi- persistent, or periodic. For periodic and semi-persistent CSI reporting, the configured peri- odicity and slot offset applies in the numerology of the uplink BWP in which the CSI report is configured to be transmitted. The higher layer parameter ReportQuantity identifies the CSI-related or L1-RSRP-related quantities to report. Another RRC parameter indicates the reporting granularity in the frequency domain including the CSI reporting band and whether PMI/CQI repo

rting is wideband or subband. The CSI-ReportConfig can also contain CodebookConfig, which contains configuration parameters for Type-I or Type-II CSI including codebook subset restriction, and configurations of group-based reporting [9]. Each CSI resource setting contains a configuration of more than one CSI resource sets, each consisting of CSI-RS resources (either NZP CSI-RS or CSI-IM) and SS/PBCH block resources used for L1-RSRP computation. Each CSI resource setting located in the down- link BWP is defined by RRC signaling, and all CSI resource settings are linked to a CSI report setting within the same downlink BWP. The reporting configuration for CSI can be aperiodic (using PUSCH), periodic (using PUCCH), or semi-persistent (using PUCCH and DCI activated PUSCH). The CSI-RS resources can be periodic, semipersistent, or aperiodic. The supported combinations of CSI reporting configurations and CSI-RS resource in NR are shown in Tables 4.11 and 4.12. If interference measurement is performed using CSI-IM, each CSI-RS resource for CM is resource-wise associated with a CSI-IM resource based on the ordering of the CSI-RS resource and CSI-IM resource in the corresponding resource sets. The number of CSI-RS resources for channel measurement equals to the number of CSI-IM resources [9]. The CSI reports are used to provide the gNB with an estimate of the downlink communica- tion channel observed by the UE in order to assist channel-dependent scheduling. The new Table 4.11: Triggering/activation of CSI reporting for CSI-RS configurations [9]. CSI-RS Periodic CSI Semi-persistent CSI Reporting Aperiodic CSI Configuration Reporting Reporting Periodic No dynamic For reporting on PUCCH, the UE receives an Triggered by DCI or CSI-RS triggering/ activation command and for reporting on PUSCH, activation activation the UE receives triggering on DCI command Semi- Not supported For reporting on PUCCH, the UE receives an Triggered by DCI or persistent activation command, whereas for reporting on activation CSI-RS PUSCH, the UE rece

ives the trigger on DCI command Aperiodic Not supported Not supported Triggered by DCI or CSI-RS activation command 500 Chapter 4 Table 4.12: Major components of NR CSI framework [9]. CSI Report Settings CSI Resource Settings CSI Trigger States It defines what CSI to It defines what signals to use to It associates "what CSI to report report and when to compute CSI. and when to report it" with "what report it. A resource setting configures more signals to use to compute CSI." Quantities to report: than one CSI resource sets where Links report settings with CSI related or L1-RSRP each CSI resource set consists of resource settings related CSI-RS resources (either NZP CSI- Contains the list of associated Time-domain RS or CSI-IM); and SS/PBCH CSI-ReportConfig behavior: aperiodic, Block Resources that are used for semi-persistent, L1-RSRP calculation periodic Time-domain behavior: aperiodic, Frequency-domain semi-persistent, periodic as well as granularity: reporting periodicity and slot offset band, wideband, Note: The number of CSI-RS Resource Sets subband is limited to one, if CSI Resource Setting is Time-domain periodic or semi-persistent restrictions: For channel and interference measurements Codebook configuration parameters: Type-l and Type-II radio supports analog beamforming and high-resolution CSI feedback through beam manage- ment, where a UE measures a set of analog beams for each digital port and reports the beam quality. The gNB then assigns a number of analog beams to the UE. As the downlink channel experienced by the UE varies, the gNB can change this assignment when the link associated with an assigned beam deteriorates. While beam management is especially instrumental in above 6 GHz frequency bands, it can also be applied to sub-6 GHz bands. Furthermore, NR supports a modular and scalable CSI framework, where high-resolution spatial channel infor- mation is provided via two-stage precoding. The first stage involves the choice of a basis sub- set, and the second stage incorporates a set of coefficien

ts for approximating the channel eigenvector with a linear combination of the basis subsets. It must be noted that while beam management and CSI acquisition can be independently operated, they can be used together to support mobile UEs. For UEs in RRC_CONNECTED state, in addition to the SS block, UE-specific CSI-RS can be configured in order to improve the quality of UE measurements and to provide better user- centric mobility experience. For example, in high-frequency bands, narrow-beam CSI-RS can be configured for the UEs at the edge of the cell in order to achieve better signal-to-interfer- ence-plus-noise ratio (SINR) range and measurement accuracy. As shown in Fig. 4.41, New Radio Access Physical Layer Aspects (Part 2) 501 Cell 2 Cell 1 NR SS block handover CSI-RS boundary SS block Figure 4.41 Configurable CSI-RS for downlink mobility measurements [72]. assuming the same energy per resource element (EPRE) is applied to CSI-RS and SSB resources for transmission, narrow-beam CSI-RS measurement can provide better SINR range which can improve RSRP measurement accuracy compared to wide-beam SSB measurements. The CSI-RS properties of the serving and neighboring cells for the mobility measurements can include NR cell ID, slot configuration used to obtain the slot offset for CSI-RS and the periodicity, for example, 5, 10, 20, 40 ms, configurable measurement bandwidth of CSI-RS, configurable parameter for CSI-RS scrambling sequence, configurable numerologies, and association between CSI-RS for mobility measurement and SSB, such as spatial QCL information. The above CSI-RS properties are signaled to the UE via dedicated RRC signaling [72]. Two antenna ports are said to be quasi co-located, if the properties of the channel over which a symbol on one antenna port is transmitted can be inferred from the channel over which a symbol on the other antenna port is transmitted. The QCL supports beam management (spatial parameter), frequency/timing offset estimation (Doppler/delay), and RRM measurements (average gain). The refe

rence signal set contains a reference to either one or two downlink reference signals and an associated quasi co-location type (QCL-Type) for each one con- figured by an RRC parameter. The quasi co-location relationship is configured by the RRC parameter qcl-Type1 for the first downlink reference signal, and qcl-Type2 for the second downlink reference signal (if configured). The quasi co-location types corresponding to each reference signal are given by the RRC parameter qcl-Type in QCL-Info and may take one of the following values [9]: QCL-TypeA': {Doppler shift, Doppler spread, average delay, delay spread} QCL-TypeB': {Doppler shift, Doppler spread} QCL-TypeC': {average delay, Doppler shift} QCL-TypeD': {Spatial RX parameter} 502 Chapter 4 The NR supports two types of spatial-resolution CSI: standard-resolution (Type I) and high-- resolution (Type II). The low-resolution CSI is targeted for SU-MIMO transmission since it relies on the UE receiver to suppress the inter-layer interference. This is possible since the number of received layers is less than the number of receiver antennas for a given UE. For MU-MIMO transmission, the number of received layers is typically larger than the number of receive antennas for the UE. The base station exploits beamforming/precoding to suppress inter-UE interference. Thus a higher resolution CSI, capturing more propagation paths of the channel, is needed to provide sufficient degrees of freedom at the transmitter [57]. In LTE, the UEs are configured with a transmission mode and a number of CSI reporting modes which are limited by complexity and scalability, whereas in NR, a modular frame- work is specified where a UE can be configured with one measurement setting, which includes N 1 CSI reporting and M > 1 CSI resource settings. A CSI resource setting can be associated with one or more reporting settings to flexibly support beam management and CSI acquisition, resulting in L 1 links. A UE can be dynamically assigned one or more reporting settings or links to generate the desir

ed CSI report, which may include CRI, which is used to indicate a preferred CSI-RS resource from a configured set since different CSI- RS resources in the set can be differently precoded, rank indicator (RI), CQI, and PMI. The CRI, RI, CQI, and PMI are associated with resource selection (when a UE measures multi- ple CSI-RS resources), the number of dominant downlink channel directions, sustained spectral efficiency or related SINR values, and the dominant channel directions chosen from a codebook of vectors or matrices. Since CSI requirements for different operational modes are different, the CSI reporting settings can include different CSI components for CSI acquisition (see Fig. 4.40 and Table 4.12). The UE measures the spatial channel between itself and the serving base station using the CSI-RS transmitted from the gNB transmit antenna ports in order to generate a CSI report. The UE then calculates the CSI-related metrics and reports the CSI to the gNB. Using the reported CSIs from all UEs, the gNB performs link adaptation and scheduling. The goal of CSI measurement and reporting is to obtain an approximation of the CSI. This can be achieved when the reported PMI accurately represents the dominant channel eigenvector(s), thereby enabling accurate beamforming. The standard-resolution (Type I) CSI utilizes a dual-stage codebook with precoding matrix W = W1W2 incorporating a wideband W1 matrix that is common for all subbands, captur- ing long-term channel characteristics, and a subband W2 matrix representing fast fading properties of the channel. In this context a subband comprises multiple consecutive resource blocks. Type I codebooks support up to rank 8, that is, the rank indicator RI {1, 2, Designed for Npanel panels of dual-polarized arrays, the W1 matrix factor is constructed from 2Npanel blocks of two-dimensional DFT matrices. The W2 matrix selects a subset of DFT vectors from W1 and applies phase shifts (taken from the phase shift keying alphabet) New Radio Access Physical Layer Aspects (Part 2) 503 acro

coefficients in W2 is also included in W1. Therefore, the amplitude component of the linear combination coefficients comprises wideband and subband components. The phase compo- nent is per subband and configurable as QPSK or 8-PSK. Due to large degrees of freedom offered by Type II CSI, the number of precoder hypotheses is large. However, exhaustive codebook search, which is prohibitively complex, is not needed. Due to high spatial resolu- tion, the precoder can be efficiently determined by performing scalar quantization of each of the channel eigenvector coefficients. Since Type II CSI is configurable in terms of its basis set size Le {2, 3, 4}, the amplitude frequency granularity, that is, wideband-only or wideband + subband, and phase shift (QPSK or 8-PSK), a range of performance-overhead trade-offs would be possible [57]. There are two subtypes of Type I CSI that are referred to as Type I single-panel CSI and Type I multi-panel CSI, corresponding to different codebooks. These codebooks are designed assuming different antenna configurations on the gNB transmitter. The codebooks for Type I single-panel CSI are designed assuming a single antenna panel with NH X Nv cross-polarized antenna elements. In general, the precoder matrix W for Type I single-panel CSI can be constructed as the product of two matrices W1 and W2 where the information 504 Chapter 4 selection Co-phase selection b0,b1,b2,b3 Type-I CSI: The UE selects beam and co-phase; i.e., the relative phase difference between cross-polarized antennas, coefficients P2 = e.jup Bupbi Type-II CSI: The UE selects multiple beams, amplitude scaling, and phase coefficients for linear combination among the beams Beam group selection Amplitude scaling Co-phasing and linear combination Wideband Per subband Figure 4.42 Illustration of Type-l and Type-II CSI in NR (57,69]. about the selected W1 and W2 is reported separately in different parts of the PMI. The matrix W1 captures long-term frequency-independen characteristic of the channel. A single W1 is selected and repo

rted for the entire reporting bandwidth (wideband feedback). In con- trast, the matrix W2 encompasses short-term frequency-dependent characteristic of the chan- nel. Thus the precoder matrix can be selected and reported on a subband basis where a subband covers a fraction of the overall reporting bandwidth. Alternatively, the device may decide not to report W2 when subsequently selecting CQI. In that case, it should assume that the network randomly selects W2 on a per physical resource block group basis. New Radio Access Physical Layer Aspects (Part 2) 505 Note that this does not impose any restrictions on the actual precoding applied at the gNB side, rather it is only about the assumptions that the device would make when selecting CQI. The matrix W1 can be considered as defining a beam or a group of beams pointing toward a specific direction. More specifically, the matrix W1 can be written as where each column of the matrix b defines beam. The 2 of a structure the matrix corresponds to two polarizations. Note that, as the matrix W1 is assumed to rep- resent long-term frequency-independent channel characteristics, the same beam direction can be assumed to fit both polarization directions. Selecting matrix W1 or equivalently b can be seen as selecting a specific beam direction from a large set of possible beam direc- tions defined by the full set of W1 matrices within the codebook. In the case of rank 1 or rank 2 transmissions, either a single beam or four adjacent beams are defined by the matrix W1 (see Fig. 4.42). In the case of four adjacent beams corresponding to four columns in matrix b, matrix W2 would select the exact beam to be used for the transmission. Since W2 can be reported on a subband basis, it is possible to adjust the beam direction per subband. In addition, W2 provides co-phasing between the two polarizations. In the case where W1 only defines a single beam corresponding to b being a vector, matrix W2 would only pro- vide co-phasing between the two polarizations. For transmission ranks R>2 the ma

trix W1 defines Nbeams orthogonal beams where Nbeams = [R/2]. The Nbeams beams, together with the two polarization directions in each beam, are then used for transmission of the R layers, with the matrix W2 only providing co-phasing between the two polarizations. The NR sup- ports transmission of up to eight layers to the same device [14]. In contrast to single-panel CSI, codebooks for Type I multi-panel CSI are designed assum- ing the joint use of multiple antenna panels at the network side considering that it may be difficult to ensure coherence between transmissions from different panels. More specifically, the design of the multi-panel codebooks assumes an antenna configuration with two or four two-dimensional panels, each with NH X Nv cross-polarized antenna elements. The opera- tion principles of Type I multi-panel and single-panel CSI are similar, except that the matrix W1 defines one beam per polarization and panel, whereas matrix W2 provides per-subband co-phasing between polarizations as well as panels. The Type I multi-panel CSI supports spatial multiplexing with up to four layers. Type II CSI provides channel information with significantly higher spatial granularity com- pared to Type I CSI. Similar to Type I CSI, Type II CSI is based on wideband selection and reporting of beams from a large set of beams. However, while Type I CSI selects and reports a single beam, Type II CSI may select and report up to four orthogonal beams. For each selected beam and each of the two polarizations, the reported PMI then provides an amplitude value (partly wideband and partly subband) and a phase value (subband). This allows constructing a more detailed model of the channel, capturing the main rays and their 506 Chapter 4 respective amplitudes and phases. At the network side, the PMI received from multiple devices can be used to identify a set of devices with which transmission can be done simul- taneously on a set of time/frequency resources, that is, MU-MIMO, and what precoder to use for each transmission. Since Ty