arXiv (Cornell University) · 2010 · 10 citations · 22 references
ComputingImmunologyComputer SecurityNetwork SecurityWireless CommunicationsAd Hoc Mobile NetworksMobile Ad Hoc NetworksWireless Network SecurityWireless NetworksWireless Sensor NetworksAd Hoc NetworksCo-stimulatory ModulationNegative Co-stimulatory Signaling
We propose and evaluate an immuno-inspired approach to misbehavior detection in ad hoc wireless networks. Node misbehavior can be the result of an intrusion, or a software or hardware failure. Our approach is motivated by co-stimulatory signals present in the Biological immune system. The results show that co-stimulation in ad hoc wireless networks can both substantially improve energy efficiency of detection and, at the same time, help achieve low false positives rates. The energy efficiency improvement is almost two orders of magnitude, if compared to misbehavior detection based on watchdogs. We provide a characterization of the trade-offs between detection approaches executed by a single node and by several nodes in cooperation. Additionally, we investigate several feature sets for misbehavior detection. These feature sets impose different requirements on the detection system, most notably from the energy efficiency point of view.
22
Ad-hoc on-demand distance vector routing
Charles E. Perkins, E.M. Royer · 1999 · 10.3K citations
Wrappers for feature subset selection
Ron Kohavi, George H. John · Artificial Intelligence · 1997 · 8.9K citations
Mitigating routing misbehavior in mobile ad hoc networks
Sergio Marti, TJ Giuli, Kevin Lai et al. · 2000 · 3.4K citations · Full text
Ingo Mierswa, Michael Wurst, Ralf Klinkenberg et al. · 2006 · 1K citations