KDTree-SOM: Self-organizing Map based Anomaly Detection for Lightweight Autonomous Embedded Systems
Ping-Xiang Chen, Dongjoo Seo, Biswadip Maity, Nikil D. Dutt · 2024
Self-Organizing Maps (SOM) promise a lightweight approach for multivariate time series anomaly detection in lightweight autonomous embedded systems. However, the enormous volume of time series data from autonomous systems testing requires huge SOMs with impractical search overhead. We present KDTree-SOM that effectively optimizes the winner node search for huge SOMs by reconstructing the SOM as a k-dimensional tree (kd-tree). KDTree-SOM achieves on average a 4 × inference time reduction for huge SOMs while achieving up to 95% anomaly detection accuracy with only KB-level memory overhead, demonstrating its potential for anomaly detection in lightweight autonomous embedded platforms.