The National Science Foundation's Division of Electrical, Communications, and Cyber Systems has awarded a $300,000 grant to Jingxian Wu, electrical engineering associate professor, to perform research on low latency anomaly detection.

Anomaly detection is the process of identifying data patterns that do not conform to their expected behaviors. It has a wide range of practical applications, such as fault-detection for critical infrastructure, intrusion detection for cyber-physical systems, and fraud detection for financial services.

Current anomaly detection techniques focus solely on detection accuracy with little or no attention given to detection delays. Detection delay, which is defined as the time difference between the occurrence and detection of an anomaly event, is critical in the prevention of catastrophic actions, such as the collapse of a bridge or the loss of power to millions of people. A lower detection delay can shorten the response time, such that remedial actions and/or counter measures can be taken to significantly reduce the damages and economic losses caused by the cyber-attacks, system failures, fraudulent activities, or natural disasters.

This grant will allow Wu to develop a new paradigm of low-latency anomaly detection methods that can minimize the detection delay while maintaining satisfactory detection accuracy. "The expected outcomes of the project include a set of theories and practical algorithms that can be used to detect anomaly events with minimum delays," stated Wu. "The algorithms can then be applied to various applications, such as the cybersecurity of power grids or the safety of critical infrastructure."

Low-latency anomaly detection methods can be applied to civil, industrial, scientific, and military applications such as power plants, communication networks, surveillance, structure health monitoring, and financial transactions.