data
stringlengths
1
59.1k
In the field of electronic tattoos, a review paper [34] presents the materials and engineering requirements, fabrication developments, and sensing and therapeutic advancements of electronic tattoos. According to the authors, there are three components of a theragnostic-based electronic tattoo i.e., Sensing components for diagnostics, supporting electronics for data transmission, and drug delivery component.
Another survey paper [42] that presents state-of-theart in security and safety issues of the Brain Computer Interaction. This work thoroughly identifies some novel cyberattacks, thier impact on security and safety and their countermeasures. Some of the attacks like misleading stimuli are strictly related to BCI as it indicates alteration of the nuerosignals generated by the patient. Other attacks and counter measures are generic for the fields of IoT, IMD, and even IoBNT. This paper is provides exceptional insights into the security aspect. The main difference in our study and this survey paper is that the purpose of our study is to cover some very important architectural and design aspects of IoBNT, specifically bio cyber interfaces, along with security issues of IoBNT.
To the best of the authors' knowledge, there is no comprehensive review of IoBNT security solutions in the literature; nevertheless, the following research from similar fields such as IoMT and BCI (Brain Computer Interaction) provide some preliminary insights into the security challenges. Keeping in view the recent pandemic and COVID-19 situation a review paper [33] has been published recently, which discusses the technological challenges and solutions for wearable bio-electronics for patient monitoring and domiciliary hospitalization. In the end, they have presented a case study application of the Internet of Medical Things(IoMT) for domiciliary hospitalization of COVID 19 patients.
However, there have been no unified works that have presented a unified view of the protocol stack models [40] . A survey on WBAN (Wireless Body Area Networks) was presented in [28] , which focuses only on the security aspects. The authors proposed a four-tier architecture for WBAN and identified communication technologies for each tier. The communication technologies and devices of WBAN are similar to that of IoBNT.
Molecular communication is the key and most promising enabling technology for the healthcare applications of IoBNT [43] . A comprehensive review paper is presented by authors in [29] . This paper presents a layered architecture of Molecular Communication and identifies the possible attack types and possible counter measures in each layer.
In [31] , the authors studied the security of big 5Goriented IoNT and the potential utility of machine learning to deal with the (big) data generated by the large number of IoNT devices. The authors outlined the potential security attackers and attack types. The comparison of the existing surveys and this study is presented in Table 1 .
This paper provides a comprehensive review of published research on IoBNT and the security and privacy considerations that it raises. To do so, relevant keywords were used to search published papers in databases such as IEEExplore, Science Direct, ACM Digital Library, and Google Scholar. The main keywords or combination of keywords used were "Internet of Bio Nano Things", "Bio cyber Interface", "Electronic Tattoos", "Wearable Internet of Things (IoT)", "Radio Frequency Identification (RFID)", "IoBNT Security, and privacy", "Cyberattacks in IoT", and "Nano networks security", published within 2010-2020 timeline. This resulted in around 80 papers after filtering by categories, topic relevance, time of publication, and contributions.
The distribution of papers across journals shows that the papers were mainly published in journals that cover interdisciplinary topics such as electrochemistry, biotechnology, wireless communication, theragnostics, and optogenetics. While there have been a number of survey and review articles on closely related fields (i.e., IoNT,IoMT and WBAN) (see Section I-A), as well as those on the underlying communication engineering principles (e.g., channel modeling, modulation techniques, enabling and communication technologies, and other networking protocols which are applicable to IoBNT) [30] , [31] , [41] , there has been no survey or review article that focuses on IoBNT challenges and research opportunities.
This paper identifies and highlights the key design issues in the IoBNT implementations (e.g., interface design options between the biological and cyber world) and the related security challenges. A lot of prominent researches in the field of IoBNT and next generation technologies have emphasized on interdisciplinary research efforts to make these novel systems holistic and practical [25] , [27] , [32] . Therefore, this article's main contributions are to bring together insights from disciplines like physics, biology, optogenetics, and electrochemitry to explore more potential bio interfacing mechanisms that are being used in these fields and their applicability to IoBNT domain.
In summary this work aims to make following contributions:
• To provide a detailed design of the potential bio cyber interfacing device for IoBNT applications. • To survey three different bio cyber interfacing technologies namely, biological bio cyber interface, electronic tattoos and RFID based bio cyber interface. • To discuss various security issues and concerns that are specific to each bio cyber interfacing technology. • To provide detailed discussion on security of IoBNT and individuating attacks and threats in IoBNT components i.e., nanonetworks and bio cyber inter- face. • To provide novel and potential mitigation strategies for the security of IoBNT. • To present open issues, challenges, and future research directions involving IoBNT and its security. A detailed elaboration of the contributions and covered topics in this article is depicted in Figure 1 .
The rest of the paper is organized as follows. The architecture of IoBNT is described in the next section. The bio-cyber interface and the taxonomy of bio-cyber interface devices are introduced in Sections III and IV, respectively. The potential security threats and mitigation techniques connected with IoBNT are discussed in Sections V and VI, respectively. The last two sections address the problems and future research prospects, as well as the paper conclusion. Figure 2 depicts the overall taxonomy of this paper.
As previously discussed, the IoBNT network helps to sense biological and chemical changes around the environment and send the aggregated data to the data center for further processing [44] . To realize IoBNT, heterogeneous devices need to collaboratively work together at nano-to macro-scale, via some interfaces between the electrical domain of the Internet and the biochemical domain of the IoBNT networks [25] . For example, as shown in Fig. 3 , the major components in a IoBNT communication network include bio-nano things, nanonetwork, bio-cyber interface, gateway devices, and some medical server(s). The components of IoBNT architecture are elaborated below:
Bio-nano things are nanodevices that are not only a computing machine that can be reduced to a few nanometers in size, but also a device that uses the
Bio-Nano Thing Gateway device Bio cyber interface Medical server Figure 3 : A Typical Internet of Bio-Nano Things Architecture, where a bio-chemical signal from inside the human body is converted into electromagnetic signal via bio cyber interface, and transmitted through Bluetooth or equivalent technology towards medical server for further analysis and processing unique properties of nanocells and nanoparticles to detect and measure new types of phenomena at nanoscale. For example, nanodevices can detect chemical compounds in a fraction of a billion. [45] , [46] , or the presence of different infectious agents such as virus or harmful bacteria [47] , [48] . Nanodevice is essentially comprise of a number of hardware constituents and all the software and programming of the nanodevice is included in the information processing unit. There are two types of Nanodevices, electronic nanodevice, and biological nanodevice. Electronic nanodevices use novel nanotechnology materials like Carbon Nano Tubes (CNT) [49] , [50] and Graphene Nano Ribbons [17] for device construction. Biological nanodevices are built using the tools from nanotechnology and synthetic biology.
Biological nanodevices can be fabricated by reprogramming biological materials [51] , [52] like cells [53] , viruses [54] , [55] , bacteria [56] , bacteriophage [57] , erythrocytes (i.e., a red blood cell), leukocytes (a type of white blood cell) and stem cells [58] , [59] or by artificially synthesizing biomolecules like liposome, nanosphere, nanocapsule, micelle, dendrimer, fullerene and deoxyribonucleic (DNA) capsule. Moreover, hybrid nanodevices can be fabricated by applying both the above-mentioned approaches. The size of nano devices may range from macro molecule to typical biological cell.
The material used to compose nano devices can be only biological [60] (proteins, DNA sequences, lipids, biological cell) or they can be synthesized with non-biological materials such as magnetic particles and gold nanorods. There are a number of naturally occurring nano devices e-g protein motors that bind certain types of molecule on cargo transport them through filaments and unbind them at destination, liposomes capable of storing and releasing certain types of molecules, biological cells coated with non-biological material for non-cell native functions(e-g absorbing mercury). An envsioned bio-nano thing is presented in Figure 4 .
Nano network is comprised of several nano-scale devices such as nano transmitter, nano receiver, nano router, and other specialized nanodevices to perform exclusive tasks like sensing, actuation, monitoring, and control. Intra-body nanonetwork is generally deployed in the environment (e.g., human body orally or through injection) to realize invivo biomedical applications. The nanonetwork consists of devices (e.g., nanomachine, nano router, and nano micro interface) at the nanoscale (1-100 nm), which work collaboratively to achieve sensing and actuation tasks [17] . Nanomachines can be interconnected to execute collaborative tasks in a distributed manner. The size of nanodevices makes them feasible for in vivo applications, where these non-invasive machines can easily be placed in hard-to-access areas (e.g., deep inside tissue) and perform therapeutic operations.
Nano micro interface is mediating device between nano and micro scale communication; it is referred to as biocyber interface and bio electronics device throughout this article to keep the generality. Basically, this device converts electrical signals received as commands from the medical server of healthcare providers into biochemical signals understandable by intra-body nanonetworks and vice versa [61] . A detailed description of bio cyber interface design is presented in Section III
Gateway devices are now an integral part of most IoT applications, where they function as a relay device between sensors and the Internet. Gateway devices are computing devices such as smartphones, tablets, laptops, or mini-computers. A gateway device supports the mobility of patients and ensures efficient signal reception of low transmission range technologies (NFC, RFID, Bluetooth LE, etc). The wearable bio-electronic devices are usually designed to be compatible with the operating system requirements of smartphones [35] , [62] . For RFID enabled applications,smartphones integrated with the RFID readout unit, which eliminates the need for external RFID tag readers. Similarly, for bio-electronic devices enabled with Bluetooth LE as the communication modality, smartphones are the ideal choice. Access point can be a cellular base station or a WiFi access point, which helps to route the body sensor's traffic to the medical server.
Medical server acts as a repository in which all the sensory data collected and sent from the patient's body is stored, analyzed and processed. This server may act as a terminal for real-time and continuous health monitoring, where emergency situations can be mitigated by sending alert messages. Only authorized entities can access this server to send commands and receive collected data, due to the critical nature of biomedical applications.
Bio-cyber interface is the hybrid and most sophisticated device, that is capable of communicating between nanoscale and microscale devices [22] , [25] , [44] . This device receives aggregated data from nanonetwork, processes nanoscale data in its transduction unit to convert it into a format suitable for the conventional network (e-g Internet) and sends it microscale devices. The components of envisioned bio-cyber interface are described in detail below:
The transduction unit in the bio-cyber interface performs the operation of converting electrical signals from external devices into biochemical signals readable by Bio-Nano Things [63] . Transduction properties in biocyber interface can be achieved by engineering devices using biomaterials, artificially synthesized biomaterials, non-bio materials or by combining all the approaches to create hybrid design [64] . The transduction unit is further divided into two constituent parts, Electro-bio transduction unit, and Bioelectro transduction unit.
Electro -Bio Transduction Unit converts the electrical signals, received from external devices into biochemical signals, and transmits them to intra-body nanonetwork for further processing. This unit consists of a decoder, a drug storage unit, an external physical effect source, and an injection chamber. The decoder receives the signal transmitted from an external device and derives logic gates. The logic gates are binary commands that produce some physical effect in the environment like thermal, optical or magnetic radio-frequency (RF) signals. The physical effect source (thermal, optical or magnetic field) placed around the drug storage unit, stimulates the nanomachines to release their content in response to external changes in the environment. The injection chamber injects the released molecules into the blood vessel network.
The drug storage compartment of the bio-cyber interface contains nanomachines that are fabricated with materials sensitive to external physical effects like changes in light, temperature, pH or enzymes [65] . The injected nanomachines traverse the blood vessel network and are anchored at the targeted site due to high affinity. The nanomachines are equipped with ligands (i.e., signaling molecules in the MC channel) that bind to reciprocal receptors (i.e., receiving molecule) and are only expressed at the targeted site [63] . The process of electromagnetic to bio chemical signal conversion is depicted in Fig 5 [63] . Bio and non-bio materials can be used to engineer nanomachine, like liposomes that are drug nanocarriersfabricated with a coating that is sensitive to external environmental factors [64] .
Photosensitive materials release encapsulated molecules when stimulated by light at certain wavelength emitted from an external laser source [66] . The process of photoisomerization destabilizes the bilayer membrane of liposome upon light illumination and allows the release of photoresponsive molecules. For example, caged compounds release molecules by bond breaking [67] and gold nanorods generate heat as a response to their conformational change [63] , [68] .
Temperature-sensitive materials release their contents upon a nonlinear sharp change in temperature of the environment such as temperature-sensitive liposomes [69] , [70] and dendrimers [71] . Such a sharp change in temperature triggers the thermo-responsive liposomes to release encapsulated molecules in the environment. Thermoresponsive liposomes should ideally maintain their load at body temperature ( 37°C) and deliver the encapsulated drug only upon an increase intra-body temperature [63] .
Magnetic particles release their contents in response to magnetic radio-frequency signals generated by an external source. For example, gold nanocrystals attached to DNA molecules induce the hybridization of the DNA molecules [72] , generating double-stranded or signal stranded DNA molecules. An architecture for bio-cyber interface has been proposed in [63] which uses two kinds of liposomes that react to variations in light and temperature. According to the proposal when the decoded binary command from the external device is 011, thermal responsive liposomes release their contents. When the decoded command is 111 photosensitive liposomes release their contents. Another proposal [64] has performed a wet laboratory experiment to engineer artificially synthesized materials (ART) using polystyrene bead, to operate as in-messaging nanomachine to forward messages to intra-bodynanonetwork The functionality in the ART based nano device is not yet implemented and is reserved for future research work.
An information-theoretic model for glucose monitoring as an application of IoBNT has been proposed in [73] . The model utilizes pancreatic beta-cell to transmit insulin molecules into the intra-body network via blood vessel channel according to commands received from the healthcare provider. The feedback of glucose level is transmitted by muscle cells towards the beta cell to be transmitted towards the healthcare provider. The model is currently designed for intra-body nanonetworks in this proposal and can be made online for continuous health monitoring by using cyber-interfaced beta cells. The bio electro transduction unit detects the concentration of received molecules and converts the molecular signals into electrical form. According to [63] , this unit employs a cell structure to be used as a bioluminescence bioreporter. The bioreporter works on the principle bioluminescence reaction that produces a reporter protein upon excitation from an analyte.
The analyte here can be transmitted information molecules, moving towards the receiving nanomachines and the reporter protein is Luciferase (LU). Bioluminescence is the production and emission of light by living organisms as a result of a chemical reaction during which chemical energy is converted into light energy. When the two chemical enzymes Luciferin (L) and Luciferase (LU), catalyze oxygen in the presence of ATP, a chemical reaction releases energy in the form of light. Fluorescence molecules such as rhodamine derivatives are another form of luminescent materials that can be used to produce light energy [64] .
These materials emit fluorescence upon excitation from microbiological conditions, which can be detected by external devices with the help of fluorescence microscopy. The emitted light energy is detected by a nanoscale light-sensitive sensor, which produces an equivalent electric signal in response. The electrical signal drives the transmitter to send the derived information through the wireless channel. Research contributions towards the realization of bio-cyber interface have resulted in some excellent theoretical frameworks for the Bio electrotransduction unit. The transduction unit for bio-cyber interface can be designed using natural/synthesized biological materials or novel nanotechnology materials like carbon and graphene.
The connection of the human body with the Internet is the basic notion behind the IoBNT paradigm. biocyber interface can be used as a module for connecting the human body to the Internet. bio-cyber interface must possess the capability to wirelessly transmit the aggregated intra-body data to the healthcare provider. Bio cyber interface is responsible for communication with the intra body nano networks as well as with the internet. Therefore, it must contain two communication units i.e.,one for intra body communication and one for external communication with the internet.
As discussed earlier the intrabody nano network communication cannot be realized through the traditional wireless communication technologies due to the minute size of nano transceivers inside the human body. Therefore, novel enabling technologies have been investigated that are feasible for communication inside the human body and provides an interface to access the human body. These technologies include molecular communication [43] and nano-electromagnetic communication at the THz band. Both of these technologies are discussed in detail below:
Molecular communication performs the exchange of messages between bio-nano machines through information encoded molecules [20] , [74] . To generalize MC into a communication engineering perspective, researchers have defined an architecture of MC that consists of key communication concepts and processes [74] . The basic components of MC architecture include the sender bio nanomachine, receiver bio nanomachine, and the propagation medium. The sender bio nanomachine encodes the information molecules, usually in the form of molecular concentration (i.e., the number of information molecules per unit volume of solvent molecules) modulated over time or some other form according to application. The encoded information is then released into the environment through unbinding from the sender bio nanomachine.
The propagation medium for in-vivo applications is an aqueous medium that allows the flow of molecules towards the destination. The process of molecule propagation can be active or passive. In active or walkway based propagation, molecules are encapsulated in vesicles and are transported through molecular motors towards the destination. Whereas in passive or diffusionbased propagation [75] , information molecules diffuse away into the environment and are transported towards the destination through guide molecules. The receiver bio nanomachine captures the information molecules form the propagation medium and decodes them into some chemical reaction. Chemical reactions may include the production of some signal for other molecules, performing some simple task or producing other molecules.
In [20] state-of-the-art in MC have been presented, discussing architecture, theoretical and physical modeling and challenges and opportunities in the development of MC based systems. Some researchers have also related MC with traditional networking protocols such as [19] proposed a TCP like molecular communication which is a connectionoriented protocol. Moreover, another proposal [74] presents an OSI like layered architecture for MC. The functionality of each layer and relevant research challenges and opportunities according to each have been discussed in detail.
In [76] , a comprehensive survey on recent advancements on MC according to communication, perspective has been presented, which provides a detail discussion on the transmitter, receiver and propagation medium of MC systems. Some other works on the development of MC systems include the transmitter and receiver design [77] - [79] modulation techniques [80] - [83] , channel mod-eling and noise analysis [84] - [88] . The most prominent application area of MC is healthcare and nanomedicine. In this direction, a comprehensive survey paper [89] on the medical applications of MC has been proposed that discusses a taxonomy of potential medical applications of MC, challenges, and opportunities in realizing them and future directions. Among the medical applications of MC, Targeted Drug Delivery (TDD) has got the most attention from the research community. A very comprehensive survey paper on the MC based TDD that discusses system models and design requirements/ challenges for MC based TDD systems have been conducted by authors in [90] .
Other work on MC based TDD can be found in [91] - [96] . MC is the most promising IoBNT technology due to its bio-inspired nature [25] and suitability for bio medical applications. Therefore most of the work in nanocommunication and IoBNT domain is dominanted by MC.
Nano EM communication at THz band is another communication option for intr-abody communication, whose feasibility was first discussed in [17] . In this technology, EM waves are used as an information carrier between source and destination. The advancement in novel nanomaterials such as Carbon nanotubes (CNT) and graphene derivatives such as graphene nanoribbons (GNR) has opened up doors for EM communication at the nanoscale [21] . The THz spectrum is less vulnerable to propagation effects such as scattering, which makes it ideal for intrabody communication due to its safety for biological tissue i.e., non ionization [97] .
The state-of-the-art in EM nano communication was proposed by Akyildiz et al [21] , which discusses the components, architecture and manufacturing possibilities of nanosensors for nano EM communication. Moreover, this paper also discusses the potential application areas of nano EM networks and challenges in realizing them. A new channel modeling scheme for nano EM networks based on radiative transfer theory is proposed in [98] . This paper also investigated the channel capacity of THz band based EM communication using different power allocation schemes. A state-of-the-art review paper that discusses possible biomedical applications of THz EM communication, current models and possible antenna designs, is proposed by Abbassi et al [99] .
For better signal reception and to increase the mobility of patients with bio-cyber implants, smartphone devices are used as gateway devices to the internet [63] . Smartphone devices have now become an integral part of many IoT applications due to its ubiquity, advanced computational capabilities and opensource [100] . Wireless Body Area Networks (WBANs), Wireless Chemical Sensors (WCS) and other body area sensors are now being developed to be compatible with the operating system of the smartphone.
This section includes wireless technologies that are feasible for the connection between bio-cyber interface and smartphone device. Wireless Body Area Networks (WBANs) is a similar field of IoBNT which utilize implanted sensor devices to wirelessly transmit, measured human body parameters to healthcare applications via the Internet [101] . Research in WBANs is relatively mature as compared to IoBNT, therefore, a literature survey from WBANs has been included in this section. Moreover, researchers in the field of chemical engineering have been working on wireless chemical sensors (WCS) for more than a decade now [26] .
WCSs are wearable patches that are capable of transmitting sensed bodily chemical values to the connected smartphone for analysis and processing by healthcare applications. The wireless communication technologies that are suitable with bio-cyber interface design are identified through a literature review of relevant domains such as WBANs and WCSs. This work also go through some of the important factors to consider when designing a bio-cyber interface, such as size, power supply requirements, data rate, real-time vs. on-demand data transfer, transmission range, and warning capability. Selection of suitable wireless technology depends upon the bio-cyber interface application system requirements and the underlying mode of operation of the wireless technology.
There are a number of active and passive wireless technologies available today to establish human-Internet connectivity. Active wireless technologies contain transponders able to transmit and receive radio frequency waves at high data rates and long-distance. Active wireless technologies require a continuous battery supply for transmission and to power the circuitry. Examples of active wireless technologies include Zigbee and Bluetooth, and IEEE 802.15.6. While passive wireless technologies include RFID and Near Field Communication (NFC). These technologies are briefly explained below and are illustrated in Table 2 .
Bluetooth (LE) Low Energy technology was introduced to wirelessly connect small low power devices to mobile terminals. This technology is ideal to be integrated with implants for healthcare applications as it supports ultralow power consumption [102] . It contains tiny Bluetooth radio to actively send and receive messages from nearby smartphone devices. The transmission range of Bluetooth LE is up to 10 m, the data rate is 1 Mbps and the frequency band is 2.4 GHz ISM [103] . Pairing time with other devices is in milliseconds, which is appropriate for alarm and emergency conditions in healthcare applications [104] . VOLUME 4, 2021 b: ZIGBEE ZigBee has an active transmitter able to communicate with mobile devices over a distance of 10-100 m. Zigbee is considered the most cost-effective technology due to its low power and low data rates. ZigBee can operate on three ISM bands with data rates from 20 Kbps to 250 Kbps [104] . Although ZigBee provides a large transmission range, it is not a good candidate for continuous health monitoring applications due to its low data rates [101] . Medical Implant Communications Service (MICS) band is a licensed band used for implant communication and has the same frequency range (402-405 MHz) in most of the countries [105] . MICS band is suitable to be integrated for bio-cyber interface as it defines protocols that are compatible with system requirements of biocyber interface. IEEE.802.15.6 can operate within a transmission range of 3 m with data rates up to 10 Mbps [108] .
IEEE.802.15.6 has also been recognized to be used as a standard for Human Body Communication (HBC). HBC uses the human body as signal propagation medium, therefore certain concerns must be taken into account such as increased mobility of the patient, lower power consumption, small battery size yet life to span around a time several months, and management to aggregate burst data in presence of continuous triggers from physiological data. Considering these HBC concerns [109] proposed a medium access method for statistical frame-based time division multiple access (S-TDMA) protocol that demonstrates lower data latency, lower power consumption, and higher transmission efficiency.
NFC is a passive communication designed to operate in the ultra-low transmission range of up to 20 cm. NFC technology contains NFC tag similar to the RFID tag, which is powered by a readout device. Communication in NFC requires devices to touch each other or be in close vicinity to each other. Smartphone transmits power wirelessly to resonant circuits through inductive coupling which turns "ON" the NFC tag. Once the transmission session is complete, the tag is marked as "OFF" and becomes unreadable [110] .
This technology is ideal for implantable devices with a low power source as the communication is powered by the reader device (in our case smartphone). Ultra-Low transmission range in NFC is not a drawback, rather it is a strength as it provides rapid connection [101] .
RFID is a passive wireless technology that consists of a transponder (tag) to be read by an RFID reader. RFID based implants are feasible in situations where battery supply is a major issue [101] . Entire transmission in passive RFID systems is powered by the reader side [111] , which in our case can be RFID reader enabled smartphone devices. The read range of RFID is 1-100 m with the data rate of 10-100Kbps [104] . Batterypowered RFIDs with active radio transmitters are available which comes with a high cost yet low data rates.
Bio cyber interfaces need a continuous power supply for data collection, processing, and transmission. The battery is mass vise the largest unit in the bio-cyber interface and other body implants. Although considerable progress has been done in lithium rechargeable batteries, yet they cannot keep pace with the novel technology requirements and size constraints. Miniaturization of a battery source, energy harvesting methods based on human body movements, and wireless battery recharge techniques have been explored by researchers of MIT for implants in Wireless Body Area Networks (WBANs).
Evanescent waves have been considered in [112] to wirelessly power the electronic devices over a short distance. Other efforts include solar rechargeable batteries [113] , piezoelectric nanogenerators [114] , microsupercapacitors [115] and endocochlear potential-based bio batteries [116] . Implantable devices in IoBNT need a continuous power supply and prolong battery life, for this purpose mechanism has been investigated to scavenge power supply from the epidermal layer of the wearer [116] . These energy scavenging batteries are called biofuel cells (BFC), which convert chemical energy into electrical energy through biocatalytic reactions [117] - [119] . Researchers from the University of California San Diego have demonstrated an epidermal BFC [120] to scavenge continuous energy from human preparation. Lactate is used as biofuel as it is present in human sweat in an abundant amount [120] . Similarly, energy scavenging techniques from the human body can [104] , [110] be adapted for power supply in the bio-cyber interface. A detailed energy model for nanoscale devices has been demonstrated in [121] which discusses the power consumption of each component in the nanonetwork.
This section briefly discusses the patch material, as biocyber interface is envisioned to be a wearable/stickable bio-electronic device. The adhesive material for the patch is under high interdisciplinary research investigation as human skin is extraordinarily stretchable ( > 100%where is the strain), highly rough (superlative height 40 µm), and generally covered with sweat and hairs [122] . Hence, delivering adequate adhesion of skin patches against human skin persists to be a challenging task. Adhesion materials should be carefully chosen to prevent issues like cytotoxicity (i.e., condition of being toxic), skin contamination, damages, risks of infection, and loss of wet adhesion make them less effective. Adhesion materials using electronic materials are extensively reviewed in [35] and bio-inspired approaches for adhesion like gecko-/beetle-inspired mushroom-shaped architectures, endoparasite-like microneedles, octopus inspired suction cups and slug-like adhesive with energy dissipation layer, has been keenly investigated in [122] . Special measures should be taken into account while selecting the adhesive material to prevent data loss and wireless connection due to strain and physical pressure on the patch while patient movement, wrinkling of skin or other factors.
There exist a few proposals in the literature that present interface either using nano EM or MC communication.
In the nano EM domain [123] have proposed an architecture of nanonetworks-based Coronary Heart Disease (CHD) monitoring system. The proposed model consists of nanomacro interface (NM) and nanodevice-embedded Drug Eluting Stents (DESs), termed as nanoDESs. The algorithm exploits the periodic change in mean distance between a nanoDES, inserted inside the affected coronary artery, and the NM, fitted in the intercostal space of the rib cage of a patient suffering from a CHD utilizing THz band. Another paper proposes a straight forward communication scheme utilizing the TeraHertz Band. The architecture consists of nano nodes and nano router, deployed in the dorsum of the human hand. Nano nodes circulate in the blood vessel for the collection of medical data and transmit the aggregated data to the nano router. Nanorouter then transmits the data to BAN (Body Area Network) device through THz band communication, which is then relayed to the Internet. Some theoretical models have adopted MC paradigm for the design of bio-cyber interface device. For example, authors in [64] have conducted a wet-lab experiment by utilizing artificially synthesized materials as an interface between the biological and electrical world. The proposal employs optical sensitive molecules(pH rodo molecules) that activate upon expression of fluorescence through fluorescent microscopy.
The above-mentioned proposals that utilize nano nodes and nanoDES in THz communication, may not be biocompatible due to the non-biological nature of these devices. Secondly, the use of external devices like a fluorescent microscope can also cause mobility issues for the patient, which negates the whole notion of IoBNT. A novel hybrid approach is presented [30] , which suggests utilizing MC based communication inside the human body for its biocompatible and noninvasive properties. The MC can communicate with graphenebased nanosensor, implanted over the human body for communication with external devices. The nano micro interface will, therefore, consist of a THz based antenna and micro/macro antenna. The idea is to limit the nonbio devices inside the human body and to read the intrabody parameters efficiently outside the body.
The design of a bio-cyber interface is a major challenge in the realization of IoBNT. In the introductory paper of IoBNT, Akyildiz et al highlighted the design challenges of bio-cyber interface and the possibility of using electronic tattoos, and RFID based sensors as biocyber interface. In order to study the possibilities of electronic tattoos and RFID-based tattoos as bio-cyber interfaces, previous proposals on electronic tattoos and RFID-based tattoos are evaluated below. Furthermore, some preliminary work on a biologically-inspired bio- VOLUME 4, 2021 cyber interface presented by the notable MC engineering research community is discussed.
Flexible and stretchable wearable electronics are getting interdisciplinary research attention as they promise to deliver continuous patient monitoring [23] , [124] while circumventing the possible discomfort caused by regular wearable electronics. There are a number of biomedical conditions that require frequent monitoring of patients at regular intervals, for example, glucose monitoring for diabetic patients, fitness monitoring of athletes, realtime detection of pathogens in biofluids for the plausible onset of disease [125] . Traditional methods to measure these chemical analytes require extracting fresh blood or other bodily samples every time. This continuous and invasive blood sampling can cause discomfort to the patient's especially elderly patients and homophobic patients.
To address some of these concerns, researchers have ventured into the development of wearable sensors which provide non-invasive ways to continuously sense the patient's vital signs and transmit the collected data to data centers for further processing. Researchers have devised non-invasive ways of drawing out samples by utilizing the epidermal layer of the human skin [126] .
Human skin contains several electrolytes that provide clinically useful information about patients' overall health such as temperature, physiological, electrophysiological and biochemical parameters [127] . Human perspiration is also an information-rich epidermal secretion that is readily available and can be used to extract information about chemical compositions like metallic ions, minerals, glucose, lactose, lactic acid, urea, volatile organic compounds, in the human body [128] . Variously, human epidermis experiences continuous bending, stretching, and deformation while performing daily life physical activities. The mechanical properties of wearable electronics might mismatch with unique mechanophysiology of the human skin and can cause degradation of device capability and discomfort to the wearer. Novel fabrication techniques for developing flexible and comfortable wearable devices are trending in academic and industrial research [35] , [122] , [129] .
To overcome the aforementioned issue, wearable electronics are now being fabricated by adapting the design principles of temporary tattoos. Temporary tattoos are a form of body art that firmly attaches to the skin as "secondary skin", while their presence is barely noticed by the wearer. Tattoos are comfortable to be worn and provide ease of performing daily life activities to the wearer, by bearing mechanical stress, washing, and other harsh conditions. These favorable properties of tattoos have attracted researchers of wearable electronics to develop tattoo based electronics. The electronic tattoos typically include sensors, memory, electronic circuits, and drug reservoir units [130] . The electronic tattoos analyze the sensed information to determine the dose and release timing of the drug contained in the unit.
Screen printing techniques can be used to print electronic tattoos of desirable shape capable of extracting rich chemical information from our epidermis and transmit these analytical data wirelessly to a smartphone. Moreover, dry free form cut and paste technology [131] , [132] is also been introduced for electronic tattoo fabrication. A summarization of existing proposals for electronic tattoos in healthcare monitoring is presented in Table 3 .
Traditional methods of drug delivery through oral intake of medicine and via injections are now quickly being replaced by transdermal drug delivery patches. These patches are good alternatives to traditional methods as they promise minimally invasive and painless drug delivery. The transdermal patch is usually made of flexible and stretchable material that contains microneedles on the sticking side of the patch. Microneedles are actually drug containers that release their payload upon external stimulation. Fourth-generation transdermal drug delivery systems [133] are now capable to release controlled amounts of the drug upon receiving commands wirelessly through the internet. A number of wearable electronic patches based therapeutic stimulators and drug delivery patches are available now that have proven to provide accurate results when compared with regular laboratory equipment.
Electronic tattoos and transdermal patches can be exposed to security threats as they are continuously connected with the wireless technology.The following are some of the threats that electronic tattoos and transdermal patches may encounter.
Electronic tattoos and transdermal patches need continous wireless network supply in order to receive and send information to the healthcare provider. Mainly the electronic tattoos are connected with a gateway device (smart phone) that relays the information over the internet. In the currently published literature for electronic tattoos, it is noticed that most of the proposals have used Bluetooth or BLE as communication mechanisms between electronic tattoo and gateway device. Bluetooth technology works within close proximity and is considered safe as attacker must be present within the patient's vicinity to launch an attack. However, Bluetooth version 5.0 has a transmission range of upto 400m moreover, the range can be magnified up to a mile through omni direc-tional antennas. Researchers from Singapore University of Technology and design have discovered 12 different Bluetooth bugs which can affect 450 type of IoT devices including implantable medical devices. Moreover, the security flaw named "SweynTooth" associated with BLE can crash the device, deadlock the device or bypass security to access device functionality. Another critical threat to Bluetooth technology is Blueborne attack. In this attack, hackers infect the devices with malwares and gains access to the device to execute malicious operations [134] .
The electronic tattoos currently manufactured are tested rigorously to withstand mechanical deformation caused by external strain. However, in cases of long term use and extreme physical stress conditions, the device abilities might be degraded and cause unwanted behaviour. The bio-electronic devices which are presently designed in the form of electronic tattoo and implantable patches that do not require any surgical procedures to affix on the human body. Non-medical professionals can apply these devices easily on human skin like stickers. This nonsurgical relaxation can be exploited by local attackers to replace the bio-electronic device with an illegitimate device on an unconscious patient. The illegitimate device now becomes part of the communication and can launch several passive and active attacks like eavesdropping, replay attack, and data modification, etc. Countermeasures:Some of the possible countermeasures for above mentioned attacks are Tamper-proofing and self destruction:Use of tamper proof package for sensing devices and enable self destruction mode upon countering an intruder attack. Minimize information leakage: Protective measures like generating artificial noise, shielding and adding randomized delay can be taken to minimize information leakage. [135] Run-time attestation: Generation of proof about updation of firmware by a remote entity. [135] .
RFID tags are utilized for the identification of objects in many IoT and Cyber-Physical Systems (CPS) due to their wireless communication capability and easy integration in the Internet cloud system [139] . RFID tag sensors have already been extensively studied for industrial applications like food safety, logistics, healthcare, public transport, fake medicine and environmental pollution [38] . The role of RFID sensors in envisioned healthcare applications IoBNT is promising due to their easy integration with implanted interfaces for communication with the Internet. RFID enabled technology is being used for several healthcare applications like real-time tracking of patients, improving their safety, and in management and medical supplies in hospitals [140] - [147] . RFID is a wireless technology that utilizes RF signals to identify objects with RFID tags. RFID system has two components: a tag which is a microchip to store electronic information and radio antenna to receive signals. There are two types of RFID systems, passive that require no battery and are powered by the reader, and an active RFID system that has a battery to power the transmission on its own [38] . Passive RFID tags are ideal for IoBNT applications due to their properties of cost-effectiveness, long battery life, and low power consumption. RFID technology can be used as sensors [148] , for sensing applications of IoBNT and can be used as a bio-electronic interface on its own. RFID antennas that are sensitive to environmental changes like gas [149] , moisture [150] and temperature [151] , have already been developed. The sensing capabilities of the RFID system can be exploited to measure the chemical and biological values of the human body. 13.56 MHz) or ultrahigh-frequency (UHF, 400 MHz) bands, has proven adequate in powering the adaptive threshold rectifier. The HF band is for patch sensors, whereas the UHF band supports' implantable sensors [30] . The sensor consumes 12 µ W to implement an electrocardiogram analog front end, and an analog-todigital converter (ADC). The envisioned Bio-electronic interface with RFID enabled technology is likely to have the following components to be operational in the IoBNT domain. RFID sensor tags that contain chemically coated thin-film resonant sensing antenna for sensing target chemical substances in the environment. Recent RFID tags are thin and flexible, allowing them to be embedded easily in the human body for health monitoring purposes [37] . RFID sensors can be integrated with sensing materials such as water-absorbing molecules to sense humidity and carbon nanostructures as gas sensors [152] . The chemical, physical, or electrical reaction of the sensing materials in the presence of the sensed parameters modify their electrical properties (permittivity, conductivity) resulting in easy-to-observe electrical metrics, such as a shift of the resonant frequency of the RFID tag antenna, verifying the simplicity and power efficiency of RFID-enabled sensors [153] . Moreover, RFID tag sensors contain IC microchips to store and process sensed data. Novel methods for the fabrication of electronic devices through inkjet printing are getting research attention in Consumer Electronics (CE). Nanomaterials like graphene, carbon nanotubes, silver, gold and copper nanoparticles, conductive polymer and their based materials are used as inkjet ink to print electronic devices.
A comprehensive review of inkjet-printed nanomaterialbased flexible radio frequency identification (RFID) tag sensors for the Internet of nano things has been presented in [38] . According to the review, inkjet printed nanomaterials are cheaper and flexible to be used as VOLUME 4, 2021 The implant is capable to deliver two kinds of drugs namely enalapril and methotrexate for hypertension and arthritis patients respectively. The diffusion of drugs across the Nanofluidic membrane is accomplished through a low-intensity electric field.
wearable electronics as long as they do not interfere with the biological signaling capability of the human body. In context of IoT applications, Kassal et al [36] demonstrates a low-power RFID tag sensor for potentiometric sensitivity. The RFID tag has the ability to measure and store the potential of electrode, which is then wirelessly transferred to a smartphone by near field communication (NFC). The RFID / NFC tagged chemical sensor is suitable for detecting pH or ion selective electrodes as part of a network of chemical sensors for IoT. The practical application of the RFID / NFC tagging sensor was tested to detect deterioration of milk by monitoring the pH value of milk over a period of 6 days. The measurements showed the fluctuation of pH value between 5.89 and 6.10 over the 5 days, averaged to a pH of 6.03. Therefore, RFID/NFC tag sensors show potential for IoT applications. Another smart health proposal using RFID technology has been presented in [62] , where smart bandage manages chronic wounds by wirelessly transmitting wound pH status to an external readout unit in the smartphone, using radiofrequency identification (RFID). This is a lightweight, noninvasive bandage proposal that saves the patient from regular hospital visits for dressing change. The pH assessment calculated in smart bandage shows high precision and accurate results when compared with regular pH meter in the laboratory. An RFID enabled adhesive skin patch is demonstrated in [154] which monitors important biomarkers in sweat and surface temperature. In this demonstration, a commercial RFID chip is adapted with minimum components to allow potentiometric sensing of solutes in sweat, and surface temperature, as read by an Android smartphone app with 96% accuracy at 50 mM N a+ (in vitro tests). [135] . Blocking: Restricting access to the tag by public readers by using privacy bit technique. Setting the privacy bit to '1' means public scanning of tag is not possible [155] . Anonymous tag: To prevent tracking of the RFID tag from hackers, an anonymous ID is asssigned to the RFID tag. Mapping between anonymous ID and genuine ID should be stored in a look up table [156] . Distance estimation:Identification of distance between tag and reader using signal to noise ratio [157] .
The domain of IoBNT is biologically inspired and adapts the design of devices and their communication mechanisms from nature. Therefore researchers are investigating methods to develop skin-mounted bioelectronics that support the seamless integration of biological materials with electrical components of the IoBNT network. In this direction, some proposals have been presented by exploiting the biochemical properties of biologically engineered materials and synthetic artificial cells to be used as device components. The device called bio-cyber interface [25] , [63] is not only capable of sensing but also contains drug reservoirs to release controlled amounts of the drug upon commands received from external devices. A bioresorbable device [158] was fabricated using naturally occurring silk as the first step towards the development of remote control implantable devices. This device biodegradable device is used to eliminate Staphylococcus aureus infection from in-vitro environments by triggering thermal stimulations and targeted drug delivery operations. The device has the wireless capability to be turned on wirelessly and it disappears once it has performed the required task, thus eliminating the need to remove the device through surgical procedures.The use of biologically inspired materials for transmission processes in bio-cyber interfaces is described in the following scholarly works. A summarization of biologically inspired bio-cyber interfaces is presented in Table 4 .
Field-effect transistors (FETs) are a type of transistors that use an electric field to control the flow of current. FETs consist of three electrodes namely source, drain, and gate. In traditional FETs, voltage is applied to the gate electrode which in turn modulates the conductance between source and drain electrodes. The conductance is reflected as the voltage-current alteration in the output channel. FET based technology is now being utilized for affinity-based electrical sensing using nanomaterials (nanowires, nanotubes, and graphene) as transducer unit [30] , [159] . FET transistors can be utilized for biosensing by replacing the gate electrode with a biofunctionalized surface called Biorecognition Unit (BU), for the detection of target molecules in the environment [79] . VOLUME 4, 2021 The BU contains receptors on the surface of the FET channel which binds ligands with intrinsic charges which result in accumulation depletion of carriers in the semiconductor channel, and hence modulation of conductance and current. The addition of BU in conventional FETs for molecular recognition makes them bio-inspired and is therefore called BioFETs. The bioFETs work on the principle of ligand-receptor pairing i.e., binding of a ligand (signaling molecule) to its receptor (receiving molecule) and to produce a response e.g., signal transmission. There a number of ligand-receptor pairs that can be used in modeling BU of bioFETS e.g., antibody-antigen, aptamer-natural ligand, natural receptor/ligand [160] depending upon the target molecule. Semiconductor materials like NW [161] , single-walled carbon nanotubes(SWCNT) [162] , graphene [163] , molybdenum polymers(MoS2) [164] and organic nanomaterials like conducting polymers [165] can be used as transducer channel of bioFETs. Among NW materials, Silicon nanowire (SiNW) has been proven to be the best fit for bioFETs due to their low power consumption, high-speed sampling, high integration density, and high sensitivity [140] , [166] , [167] . In this direction, Kuscu et al [78] have proposed SiNW bioFET based molecular antenna which receives information molecules as biochemical signals and converts them into equivalent electrical signals. The proposed model employs the theory of ligand-receptor binding and considers microfluidic advection-diffusion channel for the propagation of information ligands. The receiver model consists of three functional units. The Biorecognition Unit (BU) works as the interface for sensing the concentration of ligands. In the Transducer Unit (TU), ligand-receptor regulates the gate potential of the FET through the fieldeffect resultant from their built-in current charges. The output unit shows the current flow as a result of the modulated gate potential. Moreover, an analysis and optimization framework has been presented by providing a closed-form expression for fundamental performance metrics, such as SNR and SEP. The proposed SiNW bioFET is capable of providing efficient in-device, label free and continuous processing of sensed molecules.
Grebenstein et al [168] proposed a microscale modulator to transduce optical signals into chemical signals. The modulator is realized using synthetically engineered E.Coli bacteria that express protons into the environment upon stimulation from an external light source. Light-emitting diode (LED) of the modulator uses proton pump gloeorhodopsin (GR) to express light. The E.Coli bacteria change the pH level of their surrounding environment as a chemical reaction to an external light source. The proposed testbed achieves higher data rates on the order of 1 bit/min as opposed to previous proposals with data rates of 1 bit/hour.
New research in biology has recommends the usage of redox as a global signaling modality. Authors in [169] have adopted an approach that is inspired by sonar, which access the redox information through collaborative electrochemical probing. Authors further utilize attuned electrical inputs that are coupled with diffusible redox mediators (electron shuttles) to access redox information in a local environment and generate complex but interpretable electrical output signatures. Redox (Oxidation-Reduction) reaction is also utilized for biochemical-electronic transduction mechanisms in a number of proposals [170] - [174] . A wet lab interface prototype has been proposed recently in [173] for transducing chemical signals into electrical signals by the virtue of redox modality. The nterface prototype device consists of a dual film with inner film contains hydrogelbased film entrapping E.Coli bacteria and the outer film consists of a redox capacitor to amplify electrical signals. These cells are engineered as reporters, which respond to the presence of a certain molecule (signaling molecule AI-2) by converting the redox inactive substrate 4-Aminophenyl β-D-galactopyranoside (PAPG) molecules into redox-active p-aminophenol (PAP).
FRET (Fluorescence Resonance Energy Transfer) mechanism has been adopted by El-atty et al [175] to model uplink/downlink bio-cyber interface for the Internet of bio nano things. In FRET-based optical sensing, biocyber interface is designed for targeted drug delivery applications of IoBNT. The downlink of bio-cyber interface is designed by adopting spreading principals of SIR (Susceptible, Infected, Recovered) epidemic scheme and decode forward (DF) basis. Three types of nanomachines are used in the downlink model to realize the targeted drug delivery namely nanoreciever, nano transmitter (Infected) and nanorelay(Susceptible) according to SIR epidemic scheme. The recovered nanomachines are ones that have transmitted their exciton to the nano receiver. The uplink is designed by considering two types of nanomachines namely nanosensors and nanoactuators. Uplink signal notifies the medical server about successful drug delivery through bioluminescence reaction.
Currently the research bio cyber interface security is immature and there is a minimum published literature in this field. [176] have presented the possibility of ML based adversarial attacks for biologically inspired bio cyber interface which are presented below. Details of possible generic attacks on bio cyber interfaces can be found in Section V-B.
In the case of bio-luminescent or thermal signaling based bio cyber interface adversaries might launch an attack by manipulating the internet-enabled parameters. By manipulating the parameters, attackers can cause inappropriate amount of drug release, initiate selfannihilation of drug molecules and modify monitoring information provide by bio chemical processes. In redox based bio cyber interfaces, changing the input electrical signal can lead the capacitor charging and unwanted redox activity such as activating/deactivating the redox substrate to affect enzyme production. Bio FETs work on the principals of ligand binding through charging of electrodes. The attacker can launch sentry attack to repel required ligands or black hole attack to attract unwanted ligands to bind to the receptor to affect the current control. The changes in external current control can cause the Bio FET based bio cyber interface to exhibit unwanted behavior. Countermeasures The attacks related to biologically inspired bio cyber interfaces can be categorized as MLadversarial attacks. Possible countermeasures to ML adversarial attacks are [42] Data sanitization: This process refers to pre-processing, validating all the input data, and rejecting the harmful samples. Adversarial training: Inclusion of adversary information in the training samples to recognize attack vector. Defence distillation: Creating secondary ML model with less sensitivity and more general results. Differentail privacy: A cryptographic mechanism of adding noise to susceptible features of data. Hormonic encryption. A cryptographic mechanism to perform computations over ciphered data to generate encrypted result.
It is obvious from the above-mentioned literature that there are a lot of possibilities for designing a bio-cyber interface. The design model of the bio-cyber interface primarily depends on the application scenario. For example, a bio-cyber interface design and components for sensing applications will be different from drug delivery applications. Our take on the choice of bio-cyber interface option goes in the favor of bio-inspired interfaces.
The reason being biocompatible in nature and the research in these interfaces in line with the direction of IoBNT. Extensive in-vitro and wet-lab experiments are required to validate the theoretical proposals for the biocyber interface.
The advent of skin implanted bio-electronics and the IoBNT paradigm will not only open up a plethora of novel biomedical applications but also its wireless connection capability will enable the adversaries to utilize it malevolently. Connecting the intra-body biological environment with the cyber domain through bio-electronic devices will provide the attackers with an apparent opportunity to devise new terrorist mechanisms to harm the patient remotely. Maliciously accessing the human body through the internet to steal personal information or to create new types of diseases by malevolent programming of bio-electronic devices and intra-body nanonetworks is termed as bio-cyber terrorism [25] . Biocyber terrorism can take advantage of wirelessly accessing the human body to launch fatal and life-threatening attacks from a remote site. Therefore security features have to be embedded either in a separate component of the bio-electronic device, which may enlarge the size of the device or might be infeasible in some applications. Another possibility is to delegate security services to external devices in close proximity with sophisticated resources as compared to bioelectronics devices. For example, in a similar field of IMD (Implantable Medical Device), some researchers propose to assign security functionality to an external device like Cloaker or Med-Mon [177] . Nonetheless, the bio-electronic device must execute a lightweight authentication mechanism at least once to establish a secure connection with external devices. Bio-electronic devices will also be linked to a gateway device (such as a smartphone) to send and receive information from the healthcare practitioner. Because of the technological differences, this section discusses the security needs for nanonetworks and biocyber interfaces separately. The security goals, regardless of the underlying technological variations, remain the same. The STRIDE threat approach can be used to model security threats against IoBNT. STRIDE is the acronym for Spoofing, Tampering, Repudiation, and Information Disclosure, Denial of Service and Elevation of privilege. These six categories present a broad classification of threats and can be further divided into other related threats. Each threat category is related to a security goal: Spoofing-Authentication, Tampering-Integrity, Repudiation, Non-Repudiation, Information Disclosur, Confidentiality, Denial of Service-Availability, and Elevation of privileges-Authorization [178] , which is presented in Table. 5. Authentication (Au): Authentication ensures that the identity of each communicating party is established be-VOLUME 4, 2021 -A theoretical model for bio cyber interface has been proposed as an interface between biological and electrical domains. The model comprises of transduction units that convert an electrical signal into biochemical signals through the bioluminescence process and biochemical signals into electrical signals through photoresponsive and thermal responsive biomolecules.
SiNW bioFET based Molecular receiver [78] Biological to electrical signal conversion through bioFET technology.
-The Silicon Nanowire (SiNW) is used as the conductive nanomaterial for a molecular antenna. Microfluidic channel has been considered as a propagation medium for information molecules to flow from transmitter to receiver in unidirectionally through diffusion.
Biological Optical-to-Chemical Signal Conversion Interface [168] Optical to chemical signal through illumination effects.
-A biological signal conversion interface is designed based on E Coli bacteria that can change the pH of the surroundings by pumping protons in response to external light stimuli.
Interface [173] Chemical to electrical signal conversion through redox modality.
-The interface prototype device consists of a dual film with inner film contains hydrogel-based film entrapping E.Coli bacteria and the outer film consists of a redox capacitor to amplify electrical signals.
FRET-based biocyber interface [175] Optical to electrical and vice versa signal conversion through FRET Technology -FRET technology has been utilized to model the uplink/downlink biocyber interface for targeted drug delivery applications of IoBNT. SIR epidemic model is adopted to model downlink and the bioluminescent reaction has been utilized to generate an uplink signal. Denial of Service Availability Inability to transmit irregular vital signs of an elderly or unconscious patient who is solely dependant on the bio-electronic device for communication with the medical server.
An internal attacker can misuse access privileges to steal or tamper information.
fore executing any operation. Moreover, authentication also ensures that the data is coming from the authorized source and unauthorized users cannot access or modify the data. Authentication needs to be done on users as well as a message [29] , [179] .
Integrity ensures that the message exchanged between legitimate entities is not tampered or modified by unauthorized entities. Non-Repudiation (NR): In the traditional networking paradigm, all the communication transactions are logged to track the network anomalies and gain the attacker's profile in case the attacker tries to misuse his/her privileges. Nonrepudiation can be violated if the attacker gets access to the logs and delete the records to remove traces.
Confidentiality ensures that the attacker should not learn the content of the message exchanged between the sender and receiver. The data must only be accessible to authorized personnel upon authentication through some mechanism priory. Availability (A): This goal ensures that the services and communication of the device are always available on request. Authorization (Auth): Authorization property ensures that only those entities can execute a specific operation that has privileges to order it. Authorization requires that an entity must have been authenticated previously through the regular login (ID, password) procedure to establish the identification.
This section presents the possible attacks and existing proposals related to the nanonetworks. A summarization of security proposals for nanonetworks is presented in Table 6 .
Eavesdropping refers to passively listening to the transmission between two nodes. The listened information can be stored and later used maliciously to launch attacks. Eavesdropping in nanonetworks can take place when two legitimate nanomachines are exchanging messenger molecules and a nearby malicious nanomachine intercepts the messenger molecules silently. The passive eavesdropper can be detected in nanonetworks by a mechanism such as stochastic geometry, distance estimation techniques. The active eavesdropper might absorb the messenger molecules in the case where MC is used as a communication medium, this attack might be prevented through secrecy capacity. Other anticipated eavesdropping prevention mechanisms include beamforming, game theory (collation formation games), and artificial noise generation. Several proposals have been demonstrated to detect eavesdropper location and secure the nano communication paradigm from eavesdropping attacks. Islam et al [180] proposed a secure channel for Molecular communication. Firstly, a Diffie-Hellman algorithmbased secure key is exchanged between sender nanomachine and receiver nanomachine. Then hardware cipher- VOLUME 4, 2021 ing is performed using the secret key. As MC is a resource-constrained paradigm, therefore, Exclusive OR (XOR) cipher is used in this work due to its simple implementation and inexpensive computation. Moreover, hardware ciphering used in this work further reduces the associated time, instead of its software counterpart. The results are presented through simulation. In our opinion, the proposed method is simple and computationally less expensive but the overall security is compromised as MC needs more mathematically resilient models for security. Guo et al [181] have proposed a mathematical model for eavesdropper detection and localization in a random walk channel. This is the only work in MC that considers the detection of an absorbing malicious receiver in a random walk channel. [182] . The channel impulse response of 1-D DbMC has been exploited to detect transmitting eavesdropper in the transmission region.
Blackhole attack refers to the attack where malicious nodes spread attractant molecules to draw the network traffic towards a different location from the intended target. Blackhole attack is similar to sinkhole attack in WSNs but the sinkhole attack disrupts the routing process while the blackhole attack physically moves legitimate nodes away from the target. When it comes to blackhole attacks in nanonetworks, there are a variety of approaches that can be taken. For example in scenario of artificial immune system support application, the white blood cells that are responsible for detecting and tackling with the infection, can be attracted by malicious nodes to stop them from detecting an infection in the host system. Blackhole attack and its countermeasure using two approaches, Bayes rule and simple threshold approach for MC have been proposed in [183] . Blackhole attacks are type of DoS (Denial of Service) attack in which nanomachines can be drawn away for target e-g in case of targeted drug delivery the actuator nodes might not reach the target due to Blackhole attack.
Sentry attacks are opposite to Blackhole attacks where legitimate nanomachines are impeded away from the target, due to a large number of repellent molecules spread around the target location by malicious nodes. This kind of attack can be fatal in medical applications where instant lifesaving action needs to be taken eg nanorobots that are designated to prevent bleeding by repairing veins are attacked by a sentry to impede from reaching the target. Sentry attack and its countermeasure using two approaches, Bayes rule, and simple threshold approach have been proposed in [183] . Giaretta et al [183] have proposed a blackhole attack and sentry attack for MC. The authors have described two scenarios where (L-BNTs) Legitimate Bio-Nano Things are repelled from reaching the targeted site in targeted drug delivery application, thus keeping Bio-Nano Things from performing the normal operation.
In the second scenario called black hole attack, M-BNTs (Malicious Bio-Nano Things) are attracted to the targeted site which can lead in a delivery unwanted dosage of medication in the targeted area. Next, in the proposal, a countermeasure that enables the Bio-NanoThings to make decisions and cooperate in order to overcome blackhole and sentry attacks during target localization is proposed. The mechanisms are based on known cellular decision processes using Bayes' rule as well as artificially designed genetic circuits that evaluate chemical signal threshold (this will be known as Thresholdbased decision process), which are both lightweight enabling them to be easily implementable on resourceconstrained Bio-NanoThings. Results show that the proposed countermeasure is effective against the attack, where L-BNTs successfully move towards the target.
This attack can be launched by malicious nodes by spoofing legitimate nodes identities to become trustable and enter the network. Furthermore, the malicious nodes then send fake messages in the network and alter the data. In the case of nanonetworks, consider an exemplary communication scenario where a legtimate node 'Alice' is transmitting messages to receiver 'Bob'. Attack can be launched by an intruder 'Eave" impersonates to be Alice and tries to control Bob by sending malicious commands. A distance-dependent path loss based authentication scheme for nanonetworks using the terahertz band has been proposed in [184] for spoofing attacks. Mahboob et al [184] have proposed an authentication scheme for terahertz band EMbased nanonetworks. This work exploits physical layer attribute i.e., distance-dependent path loss for authentication at nano receiver. Moreover, an algorithmic solution has been proposed for the authentication scheme. Experimentation verification of authentication scheme via tera hertz time-domain spectroscopy setup at QMUL, UK.
To ensure an end-to-end protection of IoBNT applications, the security of these bio-electronic devices is a pre-
Communication
The channel impulse response of the physical layer has been exploited to propose an authentication scheme for the detection of transmitting eavesdropper in the broadcast region.
requisite. In order to pursue a preliminary investigation for types of attacks that are possible in bio-electronic devices, the attack vectors are explored in related fields like WBAN. (Wireless Body Area Network) [28] , IMD (Implantable Medical Devices) [178] and Wireless sensor networks [187] . This investigation has helped us to individuate attacks that are likely to occur in bio-electronic devices. The major objective of a bio-electronic device in the IoBNT healthcare application is to enable twoway communication between intra-body nanonetworks and healthcare provider. The communication mode is divided into two categories: inbound and outbound, making it easier to recognize and categorize attacks.
Eavesdropping is a passive attack that enables the attacker to covertly gain access to confidential information. The bio-electronic device might possess critical information like patient identification information, clinical history, disease detail, treatment detail, patient location, and battery status, etc. This confidential information can be exploited not only to breach a patient's privacy, but also to launch other types of active attacks. Eavesdropping can be realized during the communication of VOLUME 4, 2021 Countermeasures:Traditional networking paradigms employ encryption schemes like RSA and DES, to prevent against eavesdropping attacks. These encryption schemes are effective for the prevention of eavesdropping attacks, but these are computationally expensive and must be adopted after analyzing the resources of bioelectronic devices [188] . Lightweight encryption schemes like Elliptic Curve cryptography has been proven to be effective for resource-constrained devices [189] , also a review on other lightweight encryption schemes has been presented in [190] .