Relative Phase Equivariant Deep Neural Systems for Physical Layer Communications

2025-01-06

AI Summary

Deep neural receivers for physical-layer communications suffer from parameter redundancy, high energy consumption, and poor generalization to variations in angle of arrival (AoA). To address these issues, this work introduces **relative phase equivariance** as an inductive bias—first such application in receiver design. Leveraging group-equivariant deep learning, complex-valued neural networks, and explicit modeling of physical-layer signal symmetries, we propose a lightweight neural receiver that is equivariant to relative phase shifts induced by AoA changes. The resulting architecture achieves significantly improved parameter efficiency and inference speed; under comparable model size, it attains substantially lower bit error rates. Experimental results validate that embedding equivariant priors yields dual gains in both performance and energy efficiency, establishing a new paradigm for low-power, robust intelligent receivers in wireless communication systems.

Technology Category

• Machine Learning: Hardware-aware ML
• Intelligent Robots: State Estimation
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Security and Privacy: Privacy-enhancing technologies
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments

Abstract

In the era of telecommunications, the increasing demand for complex and specialized communication systems has led to a focus on improving physical layer communications. Artificial intelligence (AI) has emerged as a promising solution avenue for doing so. Deep neural receivers have already shown significant promise in improving the performance of communications systems. However, a major challenge lies in developing deep neural receivers that match the energy efficiency and speed of traditional receivers. This work investigates the incorporation of inductive biases in the physical layer using group-equivariant deep learning to improve the parameter efficiency of deep neural receivers. We do so by constructing a deep neural receiver that is equivariant with respect to the phase of arrival. We show that the inclusion of relative phase equivariance significantly reduces the error rate of deep neural receivers at similar model sizes. Thus, we show the potential of group-equivariant deep learning in the domain of physical layer communications.

Problem

• Deep Neural Receiver
• Angle of Arrival
• Error Rate Reduction

Innovation

• Group Equivariant Deep Learning
• Depth Neural Receiver
• Communication Performance Enhancement