Over-the-Air Multi-Sensor Inference with Neural Networks Using Memristor-Based Analog Computing

TL;DR

Real-time wireless inference in resource-constrained sensor networks is limited by energy and bandwidth. The work combines memristor-based analog computing with $L_p$-norm inspired feature transformations and a trainable over-the-air sensor fusion to enable energy-efficient end-to-end inference across sensors and a central processor. Key contributions include (i) energy-efficient in-memory analog computation via memristors, (ii) $L_p$-norm inspired max-approximation for transformation-invariant features, and (iii) a trainable over-the-air fusion that adapts to sensor distributions, delivering practical impact for low-power, low-bandwidth wireless sensing with minimal performance loss.

Abstract

Deep neural networks provide reliable solutions for many classification and regression tasks; however, their application in real-time wireless systems with simple sensor networks is limited due to high energy consumption and significant bandwidth needs. This study proposes a multi-sensor wireless inference system with memristor-based analog computing. Given the sensors' limited computational capabilities, the features from the network's front end are transmitted to a central device where an $L_p$-norm inspired approximation of the maximum operation is employed to achieve transformation-invariant features, enabling efficient over-the-air transmission. We also introduce a trainable over-the-air sensor fusion method based on $L_p$-norm inspired combining function that customizes sensor fusion to match the network and sensor distribution characteristics, enhancing adaptability. To address the energy constraints of sensors, we utilize memristors, known for their energy-efficient in-memory computing, enabling analog-domain computations that reduce energy use and computational overhead in edge computing. This dual approach of memristors and $L_p$-norm inspired sensor fusion fosters energy-efficient computational and transmission paradigms and serves as a practical energy-efficient solution with minimal performance loss.