Multi-Dimensional Learning-from-Signals on Edge Devices - SCADA Security, Cognitive Radio and Condition-Based Maintenance

摘要

ORNL has recently developed a biologically inspired multi-dimensional Learning-from-Signals (LFS) capability that is being applied to several applications including physical layer authentication to support supervisory control and data acquisition (SCADA) systems, condition-based maintenance signal processing and white-space detection for Cognitive Radio applications such as Dynamic Spectrum Access. This paper shows how bio-inspired machine learning capabilities can be ported to edge devices.