Failure-tolerant Distributed Learning for Anomaly Detection in Wireless Networks

no code implementations • 23 Mar 2023 • Marc Katzef, Andrew C. Cullen, Tansu Alpcan, Christopher Leckie, Justin Kopacz

When such failures arise in wireless communications networks, important services that they use/provide (like anomaly detection) can be left inoperable and can result in a cascade of security problems.

Anomaly Detection Federated Learning

Designing Neural Networks for Real-Time Systems

no code implementations • 26 Aug 2020 • Hammond Pearce, Xin Yang, Partha S. Roop, Marc Katzef, Tórur Biskopstø Strøm

This issue stems largely from the implementation strategies used within common neural network frameworks -- their underlying source code is often simply unsuitable for formal techniques such as static timing analysis.

C++ code