Security Performance Enhancement Chaotic Optical Communication System Based on Reservoir Computing for Physical-Layer Security

作者：Pengjin Zhu, Hongxiang Wang, Yuefeng Ji · 发表于：Journal of Lightwave Technology · 年份：2024 · DOI：10.1109/jlt.2024.3435457 · 被引用次数：5 · 研究领域：Neural Networks and Reservoir Computing、Optical Network Technologies、Advanced Memory and Neural Computing

In this article, we propose and numerically demonstrate a novel security performance enhancement electro-optical (EO) phase chaotic communication (ENRCP) system based on Reservoir Computing (RC). The proposed system utilizes the RC model to learn the 10th order Nonlinear Autoregressive Moving-average (NARMA-10) task and incorporates it into the EO feedback loop. The well-trained RC has strong high-dimensional nonlinear mapping capabilities and physical unclonable function (PUF), thus significantly improving the security performance of the proposed system. Specifically, the maximum Permutation Entropy (PE) and bandwidth of the proposed system are 0.993 and 33.697 GHz, respectively. Furthermore, the time-delay signatures (TDSs) are concealed more effectively compared to the other three chaotic systems. Simultaneously, by introducing tunable connection weights ($W_{in}$,$W$and$W_{out}$) of the RC model, the key space of the chaotic system is enhanced, which is about the order of$10^{20}$. Finally, the synchronization and communication performance of the proposed system are analyzed in detail, and the optimal parameter ranges are determined. The proposed system can effectively resist eavesdropping attacks and provide high-level security for optical communications.