Year: 2025
Type: article
Abstract: As we enter the big data and artificial intelligence (AI) era, integrating security and communication over optical fiber has become a critical challenge. This urgency is driven by the need to protect vast amounts of sensitive data, ensuring privacy security across global high-capacity optical networks. Traditional secure communication methods often struggle to maintain high-capacity transmission performance while providing robust security. Here we propose an integrated encryption and communication (IEAC) framework, designed to maximize mutual information (MI) for legal users while minimizing it for potential eavesdroppers. Enabled by end-to-end deep learning, this holistic framework trains a random number-selected geometric constellation shaping scheme to optimize encryption processes and transmission quality simultaneously. The IEAC experiment system achieves a groundbreaking single-channel transmission rate of 1 Terabit per second (Tb/s) over a 1200-km fiber link, employing a 26-channel, 3.9 THz bandwidth, full C-band wavelength division multiplexing (WDM) configuration. The MI for eavesdroppers is under 0.2 bit per symbol where the regular value is near 4.0, ensuring secure transmission. The IEAC scheme offers a scalable, promising solution to meet the escalating demand for high-throughput, secure data transmission in the face of advancing big data and AI computational technologies.
Source: National Science Review
DOI: 10.1093/nsr/nwaf112
Authors: Zekun Niu, Yunhao Xie, Guozhi Xu, Chenhao Dai, Hang Yang, Chuyan Zeng, Minghui Shi, Lyv Li, Guoqing Pu, Weisheng Hu, Lilin Yi
Institution: Shanghai Jiao Tong University
Language: English
Topic: Optical Network Technologies
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,000
Awards: G1089831306, G7394733856
OpenAlex ID: W4409099411