Non-Orthogonal Multiple Access with Wireless Caching

Abstract

With an increasing demand for rich content driving up the need for increased system capacity, novel transmission techniques are required for future mobile networks. This thesis is motivated by the fact that non-orthogonal multiple access (NOMA), device-to-device (D2D) communications, and wireless caching are promising technologies that aid in attaining high data rates with low latencies. The complementary nature of the three technologies is used to develop system models which enhance performance metrics such as sum rate and delivery times. Four novel system models employing the combination of NOMA, D2D communications and wireless caching have been developed and evaluated. In the first model, two users utilise the uplink channel to exchange previously cached content with each other. Results indicate that this system model significantly outperforms conventional cellular communications and this gain is further emphasised by the proposed power allocation solution. The second model has two strong users transmit cached content to a third weak user and the time slot and power allocation problems are solved to maximise the sum rates based on minimum rate constraints. Again, numerical simulation results obtained help to illustrate the performance gains afforded through using the proposed power and time slot allocation as compared with a conventional delivery approach. The third model focuses on the transmission of cached content during the downlink phase where D2D transmissions underlay the BS NOMA downlink transmission. Full-duplex transmissions introduced self-interference (SI) into the system, and the sum rate maximisation problem was solved subject to minimum rate constraints. Lower complexity sub-optimal solutions which assume a negligible residual SI have also been developed to simplify the power allocation process. Simulation results outline the significant performance gains present when the strong user acts as the D2D transmitter. The final model extends from the underlay D2D case to solve the delivery time minimisation problem. The total delivery time is a useful metric to assess the quality of experience and the derivations to obtain a solution to the optimization problem indicate that the total delivery time is minimized when the delivery for all files is complete at the same time.

Date of Award	1 Aug 2023
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Daniel Ka Chun So (Supervisor) & Zhiguo Ding (Supervisor)