Page 23 - ITU Journal Future and evolving technologies Volume 2 (2021), Issue 7 – Terahertz communications

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 7

4.3 MAC and higher network layers quencies ensure link secrecy, while the razor‑sharp THz
beams favors covert communication, at the cost of rapid
Similar to the physical layer‑related issues, various coordination of the communicating Tx and Rx beams. In
challenges also emerge across the higher protocol stack regards to the privacy aspects, THz communications are
layers. To start with the link layer, new medium access hard to eavesdrop from a large distance due to the huge
control (MAC) protocols are necessary to address the attenuation at THz frequencies near sea level. Meanwhile,
unique characteristics of the THz band as well as a study shows that the THz signals can be intercepted by
DSNs. Here, availability of the massive bandwidth placing an object within the razor‑sharp LOS THz beam
annihilates the basic requirement for the commu- for scattering the beam towards the eavesdropper [83].
nicating nodes to contest for the channel. Furthermore, For DSNs and drone networks at lower atmospheric al‑
THz signals with a miniature transmit duration also titudes (typically within a few 100s of meters above sea
diminishes the chances of collision. Also, with the level), secure THz ultra‑broadband communications can
be established with the razor‑sharp THz Tx‑Rx beams
razor‑sharp beams used in THz band transmission
pointing towards each other, leveraging the drone mobi-
systems, MAC design for DSNs should facilitate
lity, while the LOS beam scattering as shown in [83]
receiver‑initiated novel transmission schemes, so that
will become dif icult to realize for the eavesdroppers.
the transmitter resources are not wasted when the
This can be highly leveraged for sensitive applications,
intended receiver is unavailable. Novel MAC proto‑
such as in military surveillance, border patrolling, etc.
cols managing the THz band communications for drone
To summa‑ rize, we have studied several open issues and
networks should also involve optimization of the packet
research directions towards realizing THz‑enabled
size and error control techniques in an adaptive
drone networks and DSNs from the physical layer to
fashion. At the network layer, novel routing tech-
the higher network layers dealing with security and
niques should be devised to provision both the privacy issues. A similar study has recently been
traditional active nodes (for relaying) together with
provided in [6], where the THz wireless systems have
new passive intelligent lecting surfaces, which can been de ined based on seven features including:
relay the incoming THz signals towards the intended 1) Quasi‑optical nature, 2) THz‑based architectures,
destination node [80]. Furthermore, novel metrics of 3) Coexistence with the lower frequency bands,
routing are needed to be devised that captures the 4) Joint communication and sensing systems,
unique channel composition on the molecular level. 5) Physical layer strategies, 6) Spectrum access
These novel routing metrics can include the effect of methods, and 7) Network optimization in real time.
the molecular composition on the distance and In the following, we overview the state‑of‑the‑artar-
altitude‑dependent THz bandwidth [12]. Across the ti icial intelligence and machine leaning‑based solutions
transport layer, with THz band promising communication to the problems related to the THz communications,
at 100s of Gbps or even up to Tbps, network congestion particularly for the THz‑enabled drone networks, in
detail.
will drastically increase. This will give rise to issues
across the transport layer pertaining to the low/
congestion control, also ensuring end‑to‑end trans‑ port 5. AI/ML BASED SOLUTIONS
with reliability. For instance, it is expected that the
conventional Transport Control Protocol (TCP) conges‑ The advent of Arti icial Intelligence (AI) in the communi‑
tion control windowing will be revised to tackle the huge cations paradigm has been recently surged. Various tu‑
torials and survey works have been published within the
traf ic demands of the THz band networks [81]. In addi‑
past few years on the AI/ML implementation for wire‑
tion to adapting MAC and higher layers to work ef iciently
less communications [84, 85, 86, 87]. Among various re‑
with respect to the characteristics of the THz channel, the
cent studies, AI has been considered as the focus of 6G
potential solutions should address the mobility of drones
and DSN scenarios. networks [88, 89, 90] for complementing the traditional
methods. With the AI/ML technologies for Beyond 5G
(B5G) systems, it will be possible to minimize/replace the
4.4 Security and privacy existing manual network con iguration management, as
well as to ensure and deliver overall higher system per‑
Apart from the inherent advantages of the massive band‑
formance with increased reliability. Moreover, commu‑
width and the huge rates THz band can offer drone net‑
nication networks will be able to adapt conveniently in
works with mobility [22]; security and privacy also comes
real time based on the behavior of the users and the
up as a byproduct. Various unique characteristics of the
communication network. All in all, AI/ML will promise
THz waves highly in luence the security and privacy [82].
adaptive iguration and management of the
With the unique THz absorption Spectra, secure wireless
communication networks by learning patterns and
communications can coexist along with THz‑based detec‑ adapting to certain communication scenarios with
tion and imaging. Massive THz bandwidth favors anti‑ lexibility, e.g., learning the communication ic and
jamming approaches. However, the performance of such planning in anticipa‑ tion [91]. A similar approach can
THz systems will be environment dependent, i.e., water be considered for the THz‑enabled drone networks,
vapor concentration levels. Huge attenuation at THz fre‑ where based on AI/ML techniques, a swarm of drones

© International Telecommunication Union, 2021 11