针对多像素光子计数器（MPPC）恢复信号的过程中易受背景光干扰且同步时钟信号提取困难的问题，提出一种基于MPPC的改进型光子计数同步通信方法。首先利用MPPC完成光电转换，采用泊松分布模型中的均值（MV）、安斯库姆根（AR）变换和最大似然（ML）最优阈值检测算法对电脉冲进行预判决；然后生成相位不同的门控周期时钟，并将其作为使能信号以统计脉冲计数总和，通过查找其最大值的方式确定最佳时隙相关同步时钟；最后比较每比特脉冲计数与计数阈值的相对大小以确定比特信息。为了验证方法的有效性，搭建基于MPPC的无线光通信系统。实验结果表明，在波长为520 nm、通信速率为2 Mbit/s、采样频率为200 MHz和比特误码率为3.8×10^-3的条件下，改进型PCPB-MV、PCPB-AR和PCPB-ML检测算法的灵敏度较传统检测算法可分别提升3.0 dB、3.2 dB和3.8 dB。

Aiming at the problem that the multi-pixel photon counter (MPPC) is susceptible to background light interference and the synchronization clock signal is difficult to extract in the process of signal recovery, an improved synchronous communication method of photon counting based on MPPC is proposed. First, the MPPC is use to complete the photoelectric conversion, the mean value (MV), Anscombe root (AR) transformation and maximum likelihood (ML) optimal threshold detection algorithm of the Poisson distribution model are used to pre-judge the electrical pulse. Second, the gated periodic clock with different phases is generated, which is used as the enabling signal to count the sum of pulse counts, and the best slot-dependent synchronous clock is determined by searching its maximum value. Finally, the relative size of pulse count per bit and count threshold is compared to determine bit information. In order to verify the effectiveness of the method, a wireless optical communication system based on MPPC is built. The experimental results show that under the conditions of a wavelength of 520 nm, a communication rate of 2 Mbit/s, a sampling frequency of 200 MHz and a bit error rate of 3.8×10^-3, the sensitivity of the improved PCPB-MV, PCPB-AR, and PCPB-ML detection algorithms is 3.0 dB, 3.2 dB, and 3.8 dB higher than that of the traditional detection algorithms, respectively.