Anomaly Event Detection for Sensor Networks on Apriori Algorithms and Subjective Logic
Jinhui Yuan, Hongwei Zhou, Zhang Lai-shun · 2019
Aiming at the challenges that anomaly events are difficult to be characterized and heterogeneous nodes can not cooperate directly in sensor network, a novel anomaly detection method for sensor networks is proposed in this paper. We use Apriori algorithm to mine the sample data and find the mapping from the events to sensing data, which provides the rule for the detection. In the detection, the node first find the suspicious event, and the relevant nodes cooperate to validate it. To fuse the detection result of heterogeneous nodes, we first map the results to subjective logic opinions, and generate reasonable and comprehensive conclusions with its fusion rules. Our analysis shows that our solution is able to refine the range of node detection events, improve detection efficiency and generate detection results reasonably.