Cooperative cross layer detection for blackhole attack in VANET-OLSR

In this paper, we address the problem of detecting blackhole attack targeting the Multi Point Relays (MPRs), using Vehicular Ad hoc networks-Secure Optimized Link State Routing protocol (VANET-OLSR). In the literature, watchdog model was proposed to detect network related attacks. However, this technique relies on routing level monitoring, which suffers from high false positives due to channel collision. As a solution to improve the watchdog detection, we propose a cooperative intrusion detection system based on cross layer architecture, that correlates both MAC and network layers detections. This is achieved by (1) monitoring the number of RTS/CTS (request to send/clear to send) requests of the watchdogs and the detected nodes at the MAC layer; and (2) recalculating watchdogs detection percentage after aggregating the results with MAC monitors. Cooperative monitoring in both network and MAC levels helps recognize the presence of channel collision and hence, reduces the false alarm rate. Simulation results corroborate that the use of cooperative cross layer design improves the detection percentage and minimizes the false positive rate.