Phase retrieval of M-DPSK based on improved K-means clustering algorithm

The procedure of self-differential detection in free space optical (FSO) communication is analyzed in our research, based on which, analytic expressions of the detected signals are derived. According to the derived expressions, low-order wavefront aberration would still exist after transmitted through self-differential interference structure, meanwhile, detected phases would be easily affected by noises from the detectors under a low signal-to-noise ratio (SNR) condition. Aimed at solving above problems, improved K-means clustering algorithm is designed and applied in our research. Simulation results show that the proposed method could successfully improve the quality of phase retrieval for M-DPSK modulation under circumstances of different SNR and transmission rates. Moreover, adaptive communication could be achieved through calculating loss function based on the rule of fuzzy clustering.