Network Intrusion Detection System Based on the Combination of Multiobjective Particle Swarm Algorithm-Based Feature Selection and Fast-Learning Network

Given the growth of wireless networks and increase advantages applications communication networks, especially mobile ad hoc (MANETs), this type network has attracted attention users researchers more than before. The benefit these types in various kinds environments is that MANET does not require to hardware infrastructure communicate send receive data packets within network. It one main reasons for using fields. On other hand, increased popularity led many challenges, most important which security. In regard, a lack regulatory security MANETs caused some problems sending receiving data, where intrusion been recognized as issues. MANETs, notes act link between source destination nodes play role relays routers Therefore, malicious node penetration destruction information packages become feasible. Today, detection systems (IDSs) are used solution deal with problem through remote monitoring performance behaviors existing sensor networks. addition detecting network, IDSs can predict behavior future cases. present study introduced IDS (NIDS) entitled MOPSO-FLN by combination multiobjective particle swarm optimization algorithm- (MOPSO-) based feature subset selection (FSS) fast-learning (FLN). work, we KDD Cup99 dataset select features, train test model. According simulation results, method was able improve terms evaluation criteria, compared previous methods, creating balance objectives number representative features training errors on evolutionary power MOPSO.