Low-Complexity Near-Optimal Iterative Signal Detection Based on MSD-CG Method for Uplink Massive MIMO Systems

Abstract

Massive multiple-input multiple-output (MIMO) wireless system is increasingly becoming a vital factor in fifth-generation (5G) communication systems. It is attracting considerable interest due to improve range, spectral efficiency, and coverage as compared to the conventional MIMO systems. In massive MIMO systems, the maximum likelihood detector achieve the optimum performance but it has exponential complexity for realistic antenna configurations systems, Moreover, Linear detectors commonly suffer from a matrix inversion which is not hardware-friendly. There is an increase in the computational complexity associated with the unique benefits of the massive MIMO systems. The system might be classified as an ill-conditioned problem and hence, the signal cannot be detected. To reduce the data detection complexity, we investigate a linear detector based on the multiple search direction conjugate gradient (MSD-CG) in the massive MIMO uplink systems. Several theoretical iterative techniques that can be used to balance complexity and performance for massive MIMO detection have been proposed in the literature. These methods whose convergence rate for common applications is slow where there is a decrease in the base station to user antenna ratio. In this paper, the performance of the CG method has been advanced by a projection method that necessitates a search direction in each sub-domain instead of making all search directions conjugate to each other. In this regard, our results show that the proposed algorithm with realistic antenna configurations is superior to the existing methods in terms of computational complexity for large-scale MIMO systems.