Channel Processing in MIMO System for Mobile Communication

Date of Award

2023

Degree Name

Ph.D. in Electrical Engineering

Department

Department of Electrical and Computer Engineering

Advisor/Chair

Eric Balster

Abstract

In the next generation mobile network, communication in high-mobility environment is an important topic with the development of many new scenarios, e.g., autonomous vehicles, high speed trains, etc. The orthogonal time frequency space (OTFS) has been considered as a promising way to study high-mobility uses, especially for massive multiple input multiple output (MIMO) communication systems. However, there is a huge performance gap between an existing MIMO-OTFS system and the expectation of the next generation mobile network, of which the data transmission of high-mobility users is supposed to be more reliable and efficient. To achieve this goal, the channel information needs to be accurately acquired and the data needs to be properly precoded before transmission. In this dissertation, an uplink aided downlink channel estimation scheme is first proposed. It is numerically justified that the latency of the channel estimations can cause high estimation error. Possible path change caused by processing latency is theoretically derived. Taking in to account path change and power leakage phenomenon, a deep learning supported path prediction and channel estimation are proposed to accurately acquire the incises and values of the significant elements. Compared with the existing channel prediction schemes, the proposed scheme can achieve higher accuracy. After combining with non-orthogonal multiple access (NOMA), the spectrum efficiency of the MIMO-OTFS system can be much improved. A spatial division assisted precoding scheme is then introduced in this dissertation. The interference is eliminated by using the spatial information of the users. To further improve the spectrum efficiency, a cross-entropy based power allocation scheme is proposed for the studied multi-user MIMO-OTFS-NOMA system. Extensive simulation are conducted. The results demonstrate that our proposed schemes outperform the existing channel estimation schemes with a much better accuracy in channel reconstructions and lower bit error rate in communications. Meanwhile, our proposed schemes also provide higher spectrum efficiency than the existing MIMO-OTFS systems.

Keywords

MIMO, high-mobility, OTFS, 6G, Deep Learning, mobile communication

Rights Statement

Copyright © 2023, Author

Link to Full Text

Share

COinS