Low-complexity Signal Detection For Massive MIMO Systems

The superior performance of Massive MIMO wireless transmission technology is the expense of the increased complexity for the communication system signal processing.Excellent signal detection algorithms are often associated with high complexity.The high complexity of the detection algorithm has become a serious bottleneck of the Massive MIMO system implementation.Therefore,this paper improves the traditional near-optimal linear detection algorithm,deduces the low complexity iterative detection algorithm by using the iterative searching method and compares the detection algorithms with multidimensional trade-off analysis.This paper is divided into three parts based on the traditional near-optimal minimum mean square error detection algorithm.The first part,decomposition iterative detection on imperfect channel,proves the premise foundation of iterative method applied to signal detection,deduces the analytical expression of decomposition iteration by using decomposition iterative method,builds the imperfect channel model with inaccurate channel,analyzes the computation complexity and simulates the decomposition iterative algorithm in the imperfect channel model.The second part,successive over relaxation iterative detection,introduces relaxation factor,deduces the range of relaxation factor,analyzes the convergence and convergence rate,deduces the analytical expression of optimized relaxation factor,builds spatial correlation channel model based on channel correlation,analyzes the computation complexity and simulates the successive over relaxation algorithm.The last part,gradient iterative detection,deduces speed gradient iterative algorithm and conjugate gradient iterative algorithm by two gradient search methods,proves the convergence of speed gradient and conjugate gradient iteration,analyzes the influence of weight matrix and signal detection with pre-processing method,analyzes the computation complexity of speed gradient and conjugate gradient and simulates the two gradient search algorithm.This paper derives the low complexity near-optimal signal detection algorithm by using iterative method and analyzes algorithm complexity based on the number of complex multiplication in analytical expression to draw conclusion that an order of complex multiplication is reduced and robustness which iterative algorithm can also have.With utilizing theoretical analysis and computer simulation,it concludes that iterative algorithms perform better with trade-off analysis in bit error rate performance,algorithm complexity,algorithm robustness and signal to noise ratio.