The Information Theoretic Approach to Signal Anomaly Detection for Cognitive Radio
Figure 4
Flowchart of information content analysis algorithm with clustered anomaly detection. First, clean data is used to initialise the reference event histogram and event probabilities. Then, events in the signal under test are used to update the reference histogram and event probabilities. The updated values are used to estimate the information content of the events. If the information content of contiguous events exceeds a predefined threshold, , an anomaly may be present in the signal under test.
