A hybrid approach for efficient anomaly detection using metaheuristic methods
The networks are becoming an essential part of society life and anomalies may represent a loss in network performance. Modeling the traffic behavior pattern is possible to predict the behavior expected and characterize an anomaly. We proposed a hybrid clustering algorithm, Firefly Harmonic Clustering Algorithm (FHCA), for network volume anomaly detection by the combined forces of the algorithms K-Harmonic means (KHM) and Firefly Algorithm (FA). Processing the Digital Signature of Network Segment (DSNS) data and real traffic data, it is possible to detect and point intervals considered anomalous with a trade-off between the 80% true-positive rate and 20% false-positive rate.