Research on an atmospheric turbulent channel equalization algorithm using the spatiotemporal feature fusion method.

Area of Science:

• Optical communication systems
• Atmospheric turbulence modeling
• Signal processing

Background:

• Atmospheric turbulence causes signal degradation (scintillation, multipath interference) in wireless optical systems, increasing bit error rate (BER).
• Existing channel equalization methods face performance limitations in mitigating these effects.

Purpose of the Study:

• To propose a novel deep learning channel equalization algorithm for atmospheric turbulence channels.
• To address the performance bottlenecks of traditional equalization techniques.
• To effectively mitigate the fading effects caused by atmospheric turbulence.

Main Methods:

• Developed an atmospheric turbulence channel model using measured light intensity data, incorporating scintillation and multipath effects.
• Proposed a spatiotemporal feature fusion deep learning algorithm for channel equalization.
• Evaluated the algorithm's performance using 16QAM and DCO-OFDM 16QAM modulations under various turbulence distributions (log-normal, Gamma-Gamma, exponential Weibull).

Main Results:

• The spatiotemporal fusion algorithm significantly reduced BER compared to convolutional neural network methods.
• Under log-normal distribution, BER decreased from 10^-2 to 10^-5.
• Under Gamma-Gamma and exponential Weibull distributions, BER improved from 10^-2 to 10^-6 and 10^-5, respectively.
• The algorithm achieved BERs of 10^-5 to 10^-6 under low signal-to-noise ratio conditions.

Conclusions:

• The proposed deep learning algorithm effectively eliminates the fading effect of atmospheric turbulence channels.
• The spatiotemporal fusion approach demonstrates significant efficiency and potential for practical application in wireless optical communication systems.
• The algorithm offers substantial performance improvements over traditional methods, particularly in challenging channel conditions.