基于衍射深度神经网络的四相OAM相干复用解调实现

Diffractive deep neural network-based demodulation implementation for quad-orbital angular momentum coherent multiplexing

针对OAM通信系统中相干OAM复用光束的解调技术,提出了一种基于纯振幅型衍射深度神经网络(D^(2)NN)的OAM相干复用解调实现方法。通过数值实验研究了D^(2)NN解调器对四相OAM相干复用波束的解调性能,使用误码率(BER)对其性能进行了表征。为了降低D^(2)NN解调的误码率,提出了一种改进的OAM选择策略。并与纯相位型D^(2)NN解调器进行性能对比,仿真实验结果表明,该方法对四相OAM相干复用波束具有较高的解复用和解调精度有着明显优势,为OAM相干复用通信提供了一种灵活的实时解调方法。

Aiming at the demodulation technology of coherent OAM multiplexed beam in OAM communication system,an amplitude-only diffractive deep neural network(D^(2)NN)-based implementation method for OAM coherent multiplexing is proposed.The performance of the demodulator is investigated via numerical experiments and characterized by the bit error rate(BER).An advanced OAM selection strategy is proposed to reduce the BER of D^(2)NN.And compared with the phase-only D^(2)NN demodulator,simulation experiments verify that the amplitude-only D^(2)NN has obvious advantages in high demultiplexing and demodulation accuracy for quad-OAM coherently multiplexed beam,which provides a flexible robust real-time demodulation approach for OAM coherent multiplexing communication.