Abstract
The rapid and enormous proliferation of expectation on future vehicular system has led to numerous challenges in deploying vehicular networks, which is also known as vehicle-to-everything (V2X) communication networks. V2X communication is a vehicular communication system that allows future connected and autonomous vehicles to share safety-related and infotainment information wirelessly to any entity. The main objective is to propose novel designs, so that they can be considered for next-generation V2X networks to improve road safety as well as transportation experience. For both industrial and academic areas, one of the most significant challenges is ensuring safety whilst providing required services. One promising technology for enhancing the performance of next-generation vehicular networks is the in-band full-duplex (FD) communication technology. It allows the transceivers to send and receive/sense signals simultaneously over the same frequency band.In this thesis, I have investigated applying FD technology to future vehicular networks from physical (PHY) layer to transport (TRANS) layer aspects. Since two different families of technologies have been standardised as the benchmark designs, which are known as the dedicated short range communications (DSRC) standard and the cellular V2X (C-V2X) communication standard, respectively, I started from investigating how FD technology can be applied to future vehicular networks based on the DSRC standard. Then, I continued to investigate its novel designs on the application to the networks based on the C-V2X standard.
Specifically, based on the DSRC standard, I firstly proposed a FD signal collision avoidance scheme. Next, I considered the medium access challenges, and proposed a PHY- and medium access control- (MAC) layers cross-layer design to address these challenges. After that, based on the C-V2X standard, I also applied the FD technology to a beyond 5G (B5G) vehicular ad hoc network (VANET) scenario, and proposed a novel FD MAC layer protocol. Then, I expanded our research further by jointly taken machine learning techniques into consideration, an adaptive FD deep reinforcement learning design has been proposed. Finally, I surveyed the related works, and proposed our understanding and novel ideas on how to apply FD technology in 6G-V2X networks.
A few tools and methods were used in this thesis. I introduced a FD energy detection scheme for collision detection and avoidance; Poisson distribution, Central Limit Theorem, etc. were used for evaluating the network performance; and Markov Chain was used to study the protocols. Besides, OMNeT++ was used for wireless communication simulation, SUMO was used for traffic simulation, and MATLAB was used for mathematical and simulation results analysis.
Consequently, I found that the proposed FD energy detection method outperforms the standardised half duplex (HD) energy detection method, and it is also feasible with the state-of-the-art FD hardware design achievements. Moreover, since reliability and latency are the two most significant evaluation parameters for next-generation vehicular networks, I proved that the reliability can be enhanced and latency can be shortened by deploying our proposed FD-based pro-tocols, in both DSRC-based and C-V2X-based standards.
| Date of Award | 1 Apr 2022 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Mohammed Shikh-Bahaei (Supervisor) |