Online Reliable Anomaly Detection via Neuromorphic Sensing and Communications

2025-10-16

AI Summary

This work addresses low-power applications such as brain–computer interfaces and remote environmental monitoring, where online anomaly detection must strictly control the false discovery rate (FDR) below a user-specified threshold while ensuring high real-time performance and robustness. Method: We propose a wireless spiking sensing system integrating neuromorphic sensing, event-driven spike encoding, and impulse-radio communication. To guarantee FDR control without prior knowledge of anomalies, we employ e-value–based online hypothesis testing. Furthermore, we formulate dynamic sensor querying as an optimal arm identification problem in a multi-armed bandit framework to enable efficient scheduling. Results: Experiments demonstrate that our framework maintains high detection reliability under stringent FDR constraints, reduces communication overhead by 42%, and shortens average detection latency by 3.8× compared to baseline methods—achieving significant improvements in both efficiency and timeliness.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Online Learning & Bandits
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

This paper proposes a low-power online anomaly detection framework based on neuromorphic wireless sensor networks, encompassing possible use cases such as brain-machine interfaces and remote environmental monitoring. In the considered system, a central reader node actively queries a subset of neuromorphic sensor nodes (neuro-SNs) at each time frame. The neuromorphic sensors are event-driven, producing spikes in correspondence to relevant changes in the monitored system. The queried neuro-SNs respond to the reader with impulse radio (IR) transmissions that directly encode the sensed local events. The reader processes these event-driven signals to determine whether the monitored environment is in a normal or anomalous state, while rigorously controlling the false discovery rate (FDR) of detections below a predefined threshold. The proposed approach employs an online hypothesis testing method with e-values to maintain FDR control without requiring knowledge of the anomaly rate, and it dynamically optimizes the sensor querying strategy by casting it as a best-arm identification problem in a multi-armed bandit framework. Extensive performance evaluation demonstrates that the proposed method can reliably detect anomalies under stringent FDR requirements, while efficiently scheduling sensor communications and achieving low detection latency.

Problem

• Online anomaly detection using neuromorphic sensor networks
• Dynamic sensor querying via multi-armed bandit optimization
• False discovery rate control without prior anomaly knowledge

Innovation

• Neuromorphic sensors detect events via spike generation
• Impulse radio transmissions encode sensed events directly
• Online hypothesis testing with e-values controls false discovery