Research On Signal Detection Algorithm In Massive MIMO

Massive multiple-input multiple-output(MIMO)technology as the research focus of the fifth generation(5G)mobile communication network,it means that hundreds or thousands of antennas are set at both ends of the transceiver to improve the channel capacity and spectral efficiency of the system.Designing a reliable detection algorithm to recover the transmitted signal can ensure communication quality,but the algorithm complexity increases with the number of antennas,which makes it difficult to apply to engineering.Therefore,designing a detection algorithm with low complexity and high performance is of great significance.In view of this,this paper mainly innovates and improves from the following two aspects:1.Aiming at the problem of the traditional conjugate gradient(CG)search generates error propagation and causes performance degradation,an improved parallel CG soft output detection algorithm is given.This algorithm uses the orthogonality of the residual vector in the CG algorithm,integrate and simplify the steps,solves the search step and conjugate vectors in parallel to avoid error propagation,thereby improving the detection performance and accelerating the convergence speed.At the same time,in order to avoid errors caused by hard decision directly through Euclidean distance,the posterior probability of the bits in the channel coding and decoding is introduced,and the information decoded by the decoder is used as a soft output decision.Simulation results show that the detection performance of the algorithm is much better than other approximate linear detection algorithms in both decision mode,when iteration is 3 and bit error rate is52 10-?,the soft decision have a performance gain about 6d B was obtained relative to hard decisions,that is the soft output decision can further improve the detection performance of the algorithm.2.Aiming at the problem that the likelihood ascend search(LAS)algorithm is prone to fall into the local optimal due to it only accepts a solution that is better than the current solution,an improved global optimal LAS detection algorithm based on simulated annealing is proposed in this paper.Firstly,the simulated annealing algorithm is combined with the traditional LAS algorithm,and uses the Metropolis criterion to accept solutions in the neighborhood that are worse than the current solution,thereby jumping out of the local optimal solution,and improving the detection performance of the algorithm;Then,the weighted symmetric successive over relaxation(WWSOR)algorithm in numerical analysis is introduced into LAS algorithm as the initial solution method to avoid high-dimensional matrix inversion,thereby reducing complexity of the algorithm;In addition,set a multi-neighborhood candidate set to accelerate the search speed of the algorithm;Finally,a double threshold is set as the iteration termination condition to avoid invalid loop iterations,so as to control the calculation amount of the algorithm.Simulation results show that the performance of improved global optimal LAS algorithm is far superior to the traditional LAS detection algorithm and its derivative algorithms.At the same time,as the signal-to-noise ratio increase,its performance gradually approaches the lower bound of ML detection.