Online Continual Learning for Anomaly Detection in IoT under Data Distribution Shifts

2026-03-08

AI Summary

This work proposes OCLADS, a novel framework that integrates online continual learning with communication optimization to address the degradation of anomaly detection models on IoT devices operating in non-stationary environments due to data distribution drift. OCLADS dynamically updates lightweight anomaly detection models through an intelligent, drift-aware sample selection mechanism at the device level and a collaborative drift detection strategy coordinated by edge servers. By adaptively identifying and transmitting only the most informative samples for model adaptation, the framework significantly reduces the frequency of model updates while maintaining high detection accuracy. This approach enhances both the practicality and energy efficiency of resource-constrained IoT devices, enabling sustained performance under evolving environmental conditions without excessive communication overhead.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Application Domains: Internet of Things, Sensor Networks & Smart Cities

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

In this work, we present OCLADS, a novel communication framework with continual learning (CL) for Internet of Things (IoT) anomaly detection (AD) when operating in non-stationary environments. As the statistical properties of the observed data change with time, the on-device inference model becomes obsolete, which necessitates strategic model updating. OCLADS keeps track of data distribution shifts to timely update the on-device IoT AD model. To do so, OCLADS introduces two mechanisms during the interaction between the resource-constrained IoT device and an edge server (ES): i) an intelligent sample selection mechanism at the device for data transmission, and ii) a distribution-shift detection mechanism at the ES for model updating. Experimental results with TinyML demonstrate that our proposed framework achieves high inference accuracy while realizing a significantly smaller number of model updates compared to the baseline schemes.

Problem

• Online Continual Learning
• Anomaly Detection
• IoT
• Data Distribution Shifts
• Non-stationary Environments

Innovation

• continual learning
• anomaly detection
• distribution shift
• edge intelligence
• TinyML