OAM-SK Optical Communication System Based on CNN in Plasma Sheath Turbulence

In this paper, we present a CNN-based adaptive demodulation strategy for OAM-SK optical communication in plasma sheath turbulence. We investigate the average bit error (ABER) performance under varying transmission distances, fluctuation variances, and anisotropy parameters. Specifically, we simulate the transmission of Laguerre-Gaussian (LG) beams in anisotropic plasma sheath turbulence using the power spectrum inversion method for generating random phase screens. The chosen demodulation approach is ShuffleNet v2. Results indicate that the CNN adaptive demodulation strategy achieves significantly lower ABER compared to the conjugate mode sorting demodulation strategy. Even in severe turbulence, the CNN approach maintains a relatively low BER. Furthermore, CNN adaptive demodulation demonstrates superior robustness, being less affected by turbulence conditions. This work serves as a valuable reference for laser communication schemes in plasma sheath turbulence.