Abstract
Wireless sensor networks (WSNs) have a huge potential to be used in critical situations like military and commercial applications. However, these applications are required often to be deployed in hostile environments, where nodes and communication are attractive targets to attackers. This makes WSNs vulnerable to a variety of potential attacks. Due to their characteristics, conventional security mechanisms are not applicable. In this context, we propose an intrusion detection framework for a cluster-based WSN (CWSN) that aims to combine the advantage of anomaly and signature detection which are high detection rate and low false positive, respectively.