ABSTRACT
Ever-growing high-speed networks should be able to provide quality Internet services. Improving the security of the Internet is a crucial challenge for researchers. An intrusion detection system (IDS) is an important mechanism to protect these widespread networks from malicious activities. Nowadays, the network is constantly growing and becoming heterogeneous. Thus, IDSs should be able to handle a high volume of data traffic and be capable of operating with different types of networks. The research community is trying to address these through different ways, for example, hardware-based IDS [1], flow-based intrusion detection [2], and distributed intrusion detection [3]. An intelligent system can be trained to adapt to environmental changes. Different methods of artificial intelligence (AI) have been deployed in intrusion detection, for example, artificial neural networks (ANNs), fuzzy logic, and genetic algorithms (GA). In addition, hybrid intelligent IDSs, such as evolutionary fuzzy neural networks (EFuNN) and evolutionary neural network (ENN)–based IDSs, are also used [4-6].
