학술논문

Minimum Symbol Error Probability Discrete Symbol Level Precoding for MU-MIMO Systems With PSK Modulation
Document Type
Periodical
Source
IEEE Transactions on Communications IEEE Trans. Commun. Communications, IEEE Transactions on. 71(10):5935-5949 Oct, 2023
Subject
Communication, Networking and Broadcast Technologies
Precoding
Symbols
Modulation
Phase shift keying
Quality of service
Signal to noise ratio
Error probability
Symbol level precoding
MIMO systems
minimum symbol error probability
minimum union bound symbol error probability
low-resolution quantization
Language
ISSN
0090-6778
1558-0857
Abstract
This study focuses on the development of low-resolution symbol-level precoding techniques for multiuser MIMO downlink systems with PSK modulation. While for QPSK the established minimum symbol error probability criterion is used, a criterion for PSK modulation, in general, is proposed based on the minimum union-bound symbol-error probability. Based on these criteria different low-resolution precoding approaches are proposed. First, suboptimal solutions are computed via a partial greedy search method. Then the suboptimal solutions are utilized as initialization for a novel branch-and-bound algorithm that can exploit knowledge of the system’s quality-of-service demands. Different than existing branch-and-bound approaches the proposed quality-of-service branch-and-bound method searches for a solution that attains a target symbol-error probability while going in the direction of the global optimal solution. In this sense, the proposed branch-and-bound method allows for tunable complexity performance trade-offs. Numerical results confirm that the proposed quality-of-service branch-and-bound algorithm yields reduced symbol-error probability with significantly smaller computational complexity than other state-of-the-art branch-and-bound designs.