基于阶梯矩阵高效初始化的信号检测算法OA

In massive multiple-input multiple-output(MIMO)systems,as the number of users and antenna scales continue to expand,the traditional minimum mean square error(MMSE)detection algorithm faces the challenge of sharply increasing compu-tational complexity.Meanwhile,its inherent serial processing characteristics also limit the potential for parallel implementation.To address these issues,this paper proposes a stair matrix-improved iterative algorithm(SM-IJA).The algorithm first introduces a stair matrix to optimize the initial solution of iteration,and then applies linear acceleration to the traditional Jacobi algorithm.This effectively improves the convergence speed of the algorithm with only a small increase in complexity.In addition,a new hardware ar-chitecture is proposed for the SM-IJA algorithm,defining the required physical components and their interrelationships.Simulation results show that the proposed algorithm achieves a low bit error rate(BER)and high parallelism,with complexity reduced by an or-der of magnitude compared to MMSE.