CNN-AE在超奈奎斯特无线光通信端到端系统中的性能

End-to-end performance of a CNN-AE based faster-than-Nyquist ratefree space optical communication system

符号间干扰的存在使超奈奎斯特(faster-than-Nyquist, FTN)速率无线光通信系统的性能受到严重影响,针对此问题,提出了一种基于卷积神经网络自编码器的端到端通信系统来消除符号间干扰的影响并完成信号的恢复。针对超奈奎斯特信号非正交的特性,采用交替训练算法分别训练发射机和接收机权重,解决监督训练中模型输入输出数据尺寸不匹配的问题。在此基础上,分析了该系统在Gamma-Gamma大气信道中的误码率(bit error rate, BER)性能。仿真结果证明,与采用最大似然估计的传统系统相比,该系统的误码率性能在各种条件因素影响下都有不同程度的提高。当加速因子在Mazo限内时,该系统可以消除FTN成型和湍流信道带来的复杂混合码间干扰,其误码性能几乎与正交传输系统相等。

The performance of faster-than-Nyquist(FTN)free space optical(FSO)communication system is significantly affected by inter-symbol interference(ISI).To address this issue,an end-to-end communication system leveraging a convolutional neural network autoencoder(CNN-AE)has been proposed to eliminate the influence of ISI and facilitate signal recovery.Additionally,given the non-orthogonal characteristics of FTN signals,an alternating training algorithm has been employed.This algorithm trains the transmitter and receiver weights separately,addressing the issue of data size mismatch between model input and output during supervised training.Based on this,the bit error rate(BER)performance in Gamma-Gamma atmospheric turbulence channel was analyzed.Simulation results demonstrate that compared with traditional systems using maximum likelihood sequence estimation(MLSE),the proposed system achieves superior BER performance under various conditions.Specifically,when the acceleration factor falls within the Mazo limit,the proposed system effectively eliminates the complex mixed ISI caused by FTN shaping and atmospheric turbulence channel,resulting in BER performance comparable to that of orthogonal transmission systems.