Research On Compressive Sensing Based User Access Management In MIMO Communication System

With the rapid development of wireless communication technologies and the increasing popularity of multi-media mobile communication services, the spectrum resources become increasingly scarce. In order to meet the ever-growing wireless communication requirements and provide the necessary capacity, wireless communications develop from single-input single-output (SISO) system to multiple-input multiple-output (MIMO) system. Due to the importance of downlink management in MIMO system, there exists a substantial body of related works such as user detection and channel estimation. we try to solve this problem by exploiting the newly emerging compressive sensing (CS) techniques.Firstly, we cover the fundamentals of CS. Besides, we also discuss the current research progress on applying CS in wireless communications, by summarizing the techniques of other CS techniques like Bayesian compressive sensing (BCS) and presenting the principles.This paper put an emphasis on the application of CS into the downlink management. Specifically, we proposed a CS-based scheme, which fully takes advantage of the sparsity of accessed users and wireless channels and could achieve more accurate user detection and channel estimation performance, with the aid of the same identity sequence. At the same time, it can improve the utilization rate of the channel resources.The paper designs an adaptive method to adjust the length of identity sequence, based on the number of network devices. Compared to the method using constant length of identity sequence, when the number of network devices decreases, the proposed method could be timely reduce the sequence length and degrade the channel resource usage and device energy consumption.Meanwhile, we also proposed a Hadamard matrix-based method to construct the identity sequence, which could further enhance the detection performance and improve the utilization rate. As a result, the simulation results validate that the proposed method could obtain the same channel detection performance with shorter length of required identity sequence.Finally, we also discuss the future directions on applying CS.