Signal Detection Algorithm Based on SOR Algorithm in Massive MIMO System

From MIMO respect, the Minimum Mean Square Error(MMSE) is identified as an near-optimal method in uplink signal detection algorithms area. However, MMSE requires a high level of computational complexity. Iterative algorithms do not require accurate calculation of matrix inversion, so it can significantly reduce the complexity. Compared with other various iterative algorithms, successive over relaxation iteration algorithm can do a better job at performance and convergence part. This dissertation combines the SD (Steepest Descent) algorithm and the SOR (Successive Over Relaxation) iterative algorithm together, then, following efficient searching direction provided by SD, SOR can increase speed in convergence and it also can improve performance. In the meantime, in order to apply the hybrid algorithm to soft decision successfully, approximated method is employed to calculate LLR of channel decoding. The simulation results show that the improved algorithm not only can converge quickly and closely approach to the detection performance of MMSE, it also reduces the one order of magnitude of algorithm complexity compared with MMSE.