Intrusion detection in wireless sensor network using genetic K-means algorithm

Security of communication systems has become a crucial issue. A harder problem to crack in the field of Network Security is the identification and prevention of attacks. An effective Intrusion Detection System (IDS) is essential for ensuring network security. Intrusion detection systems include pattern analysis techniques to discover useful patterns of system features. These patterns describe user behavior. Anomalies are computed using the set of relevant system features. The derived patterns comprise inputs of classification systems, which are based on statistical and machine learning pattern recognition techniques. Clustering methods are useful in detection of unknown attack patterns. Elimination of insignificant features is essential for a simplified, faster and more accurate detection of attacks. Genetic algorithm based clustering offers identification of significant reduced input features. We present a conceptual framework for identifying attacks for intrusion detection by applying genetic K-means algorithm.