A reduced complexity multiple symbol differential detection algorithm in MIMO-OFDM system

In view of the shortcoming of large-capacity storage requirements and high computational complexity for most multi symbol differential detection (MSDD) algorithm, we propose PMCTS (Pruning Memory Constrained Tree Search) algorithm based on Memory Constrained Tree Search (MCTS), which uses a threshold for node cutting operation. The nodes whose metrics are greater than the threshold value will be deleted directly. Simulation results show that the proposed algorithm solves the problem of relatively high computational complexity with a small PMCTS storage in the low SNR region, but its performance loss less than 0.2dB. For example, under the condition of the RA channel model and the observation window length is 6, the average visited notes of PMCTS can be reduced up to 47.8% than the MCTS algorithm, 87.8% lower than the ML, and 36% lower than PASD (Pruning Automatic Sphere Decoding).