Deep Learning-based Codes for Wiretap Fading Channels

2024-09-13

arXiv.org

AI Summary

This work addresses the challenge of finite-blocklength secure coding over multipath fading wiretap channels without requiring channel state information (CSI). Methodologically, we propose an end-to-end deep learning framework comprising a neural encoder–decoder architecture integrated with a hash-based physical-layer security layer; the system jointly optimizes bit-error rate (BER) and information leakage under finite-length constraints. Our key contributions include: (i) the first CSI-free deep learning approach for finite-blocklength physical-layer security; (ii) novel analytical insights into how fading order, channel coefficient variance, and hash seed jointly govern security performance; and (iii) empirical validation demonstrating precise BER control, substantial reduction in information leakage, and strong robustness against channel statistical mismatch—even in multi-tap fading environments. The framework establishes a new paradigm for practical, CSI-agnostic physical-layer secure communication.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Learning to Search
• Machine Learning: Adversarial Learning & Robustness

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The wiretap channel is a well-studied problem in the physical layer security (PLS) literature. Although it is proven that the decoding error probability and information leakage can be made arbitrarily small in the asymptotic regime, further research on finite-blocklength codes is required on the path towards practical, secure communications systems. This work provides the first experimental characterization of a deep learning-based, finite-blocklength code construction for multi-tap fading wiretap channels without channel state information (CSI). In addition to the evaluation of the average probability of error and information leakage, we illustrate the influence of (i) the number of fading taps, (ii) differing variances of the fading coefficients and (iii) the seed selection for the hash function-based security layer.

Problem

• Develops finite-blocklength codes for secure communication.
• Evaluates error probability and information leakage in fading channels.
• Analyzes impact of fading taps and hash function security.

Innovation

• Deep learning-based finite-blocklength code construction
• Multi-tap fading wiretap channels without CSI
• Hash function-based security layer analysis