A signal detection algorithm based on chebyshev accelerated symmetrical successive over-relaxation iteration for massive MIMO system

The minimum mean square error (MMSE) algorithm can achieve the near-optimal detection performance in Massive MIMO system. However, it involves complicated matrix inversion. In this paper, a Chebyshev symmetrical successive over-relaxation iteration algorithm (CSSOR) based on the characteristic of channel hardening in massive MIMO system is proposed to avoid matrix inversion, whose the computational complexity is reduced by about one order of magnitude compared to the MMSE algorithm. And the initial iteration value is also optimized based on the property of channel hardening to further quicken the convergence of the iterative process. The simulation results further verify that the CSSOR algorithm outperforms the newly proposed Neumann series approximation algorithm. Moreover, it converges fast, just with a few iterations it can achieves the near performance with the MMSE algorithm.