An Optimal Grey Wolf Optimization with Fuzzy Support Vector Machine based Intrusion Detection System in Clustered Wireless Sensor Networks
Sibi Amaran · International Journal of Advanced Trends in Computer Science and Engineering · 2020
Wireless Sensor Network (WSN) includes inexpensive, compact and battery powered sensor nodes to observe the physical parameters exist in the deployed region.Recently, cluster based WSN becomes more popular and achieves energy efficiency.As the sensor nodes undergo deployment in the harsh open environments, it is highly prone to diverse kinds of attacks.This study devises a new Intrusion Detection System (IDS) for clustered WSN (CWSN) using optimal grey wolf optimization (OGWO) and fuzzy support vector machine (FSVM) algorithm.The proposed OGWO algorithm includes a differential evolution (DE) technique for population initialization of GWO algorithm.The OGWO-FSVM algorithm operates on two-stage processes, namely feature selection and classification.The OGWO algorithm selects the optimal subset of features and then FSVM model classifies the data instances into normal and intrusion.The performance of the OGWO-FSVM model has been tested by benchmark dataset.The simulation results indicated that the OGWO-FSVM technique has shown better results over the compared methods significantly.