基于对数正态(L-N)湍流信道模型, 研究了LDPC编码大气光通信多输入多输出(FSO-MIMO)系统迭代检测算法。分析了内迭代与外迭代次数对系统性能及计算复杂度的影响, 并针对PDA检测算法复杂度高的问题, 提出了一种基于反馈阈值的M-PDA迭代检测算法。改进的算法利用译码输出的比特先验信息对发送比特进行分类, 从而降低检测复杂度, 并通过选取合适的阈值, 可以获得更好的检测性能。仿真结果表明, 实验时需要综合考虑计算复杂度和性能, 选取合适的内外迭代次数, 仿真条件下, 误码率为10-5、反馈阈值为0.7时, 可以获得0.15dB的性能增益。

Based on the L-N turbulence channel, iterative detection and decoding algorithm of LDPC coded atmospheric optical communication multiple-input-multiple-output (FSO-MIMO) system was studied. The influence of the number of iterations and outer iterations on system performance and computational complexity was analyzed. To solve the problem of high complexity of PDA algorithms, an improved M-PDA detection algorithm based on the feedback threshold was proposed. Improved detection algorithm classified transmission bits uses a priori information of decoder output to reduce the complexity of the detector. Simulation results show that the number of iterations of the internal and external iterations can be selected by considering the computational complexity and performance. Under the simulation condition, when the error rate and the feedback threshold is 10-5 and 0.7 respectively, the performance gain of 0.15dB can be obtained.