On intrusion detection in opportunistic networks

As the interest of infrastructureless wireless sensor networks (WSN) grows, intrusion has been recognised to cause chronic problems on security and has become the major factor for the increasing volume of hacking incidents. For that reason, research that relates to intrusion detection becomes of paramount importance. In this paper, we concentrate on the intrusion detection for the opportunistic networks (oppnets), a new paradigm of WSN. More specifically, we deal with a multilayer anomaly intrusion detection technique applied to oppnets. Even though the method is well established by other research, it is proved to conjecture attacks which are yet unknown without prior knowledge of the network topology and the hardware idiosyncrasies. Furthermore, it offers increased situational awareness and understanding of the occurred security events due to the usage of corroborating evidence/information from different layers. Extended simulations on a variety of node and multilayer parameters show a good performance of the technique under the constraints of energy efficiency, low resource availability and protocol independency. Furthermore, comparative simulations verify that the multilayer approach can be a valid security mechanism for the future of oppnets.