Abstract
The current surge in wireless connectivity, anticipated to amplify significantly in future wireless technologies, brings a new wave of users. Given the impracticality of an endlessly expanding bandwidth, there’s a pressing need for communication techniques that efficiently serve this burgeoning user base with limited resources.Multiple Access (MA) techniques, notably Orthogonal Multiple Access (OMA), have long addressed bandwidth constraints. However, with escalating user numbers, OMA’s orthogonality becomes limiting for emerging wireless technologies. Non-Orthogonal Multiple Access (NOMA), employing superposition coding, serves more users within the same bandwidth as OMA by allocating different power levels to users whose signals can then be detected using the gap between them, thus offering superior spectral efficiency and massive connectivity.
This thesis examines the integration of NOMA techniques with cooperative relaying, EXtrinsic Information Transfer (EXIT) chart analysis, and deep learning for enhancing 6G and beyond communication systems. The adopted methodology aims to optimize the systems’ performance, spanning from bit-error rate (BER) versus signal to noise ratio (SNR) to overall system efficiency and data rates.
The primary focus of this thesis is the investigation of the integration of NOMA with cooperative relaying, EXIT chart analysis, and deep learning techniques. In the cooperative relaying context, NOMA notably improved diversity gains, thereby proving the superiority of combining NOMA with cooperative relaying over just NOMA. With EXIT chart analysis, NOMA achieved low BER at mid-range SNR as well as achieved optimal user fairness in the power allocation stage. Additionally, employing a trained neural network enhanced signal detection for NOMA in the deep learning scenario, thereby producing a simpler signal detection for NOMA which addresses NOMAs’ complex receiver problem.
Date of Award | 21 Nov 2023 |
---|---|
Original language | English |
Supervisor | Faheem Khan (Main Supervisor) & Qasim Ahmed (Co-Supervisor) |