ISSNIP

Anomaly Detection

In the real world, isolated node failures can bring down an entire network. To ensure reliability and security in a network of thousands of nodes with distributed control, it is essential to develop robust algorithms for detecting anomalies that minimise their communication overhead and energy consumption. We have addressed this problem by developing a distributed method to identify misbehaviour or anomalies in WSNs. Analysis of sensor or traffic measurements to discriminate normal from anomalous behaviour is needed to provide a platform on which we can establish different routing topologies, and evaluate the trade-off between detection accuracy and communication overhead under different conditions. This will provide an important validation of the effectiveness of this scheme in different operating environments.

Previous | Next 1
top of page