从研究APD光电检测接收机被检测光场的量子统计模型出发，构造了一种适用于强度调制/直接检测的自由空间光通信系统中LDPC译码的精确实现方法。该方法在APD接收机输出统计模型(Webb-Gaussian模型)的基础上，推导了M-PPM的LDPC译码算法中初始似然比值的精确计算方法，并给出了详细的译码过程。理论分析和仿真结果表明，与目前普遍采用的APD近似非对称Gaussian统计模型下的译码性能相比较，在相同的信道条件下，该方法可以获得大约1~2 dB的性能增益。这为进一步提高自由空间光通信系统的性能提供了一定的理论依据。

Proceed from the detected optical field′s quantum statistics model of the Avalanche Photo Diode (APD) photoelectric detection receivers, an accurate realize method of Intensity-Modulation/Direct-Detection (IM/DD) Free Space Optical Communication (FSOC) fit Low Density Parity Check(LDPC)Codes decoding algorithm was investigated. This scheme, which was based on APD receivers′ output statistical model(Webb-Gaussian Model), derivated the accurate calculating method of the original value of the likelihood ratio in LDPC decoding arithmetic over M-PPM demodulation. A detailed decoding process was given at the same time. By taking theoretical analysis and simulations, a result has been proved that：comparing with the most frequent used asymmetric Gaussian approximation model of APD receivers, the scheme introduced in this paper could get a gain of 1-2 dB in the decoding performance while in the same SNR. This gives the theoretical support in order to improve the performance of FSOC system further more.