Anomaly detection in geographic routing protocols

Abstract

Providing security for sensor nodes is one of the important issues in WSNs because they are deployed in the harsh environment. WSNs are distributed in nature and preparing centralized security solutions cannot be a suitable for WSNs. There are various sensor holes that may be formed in WSNs and create variations in network topologies and also threaten the upper layer applications such as sinkhole, black holes, wormholes and Sybil attack. Wormhole attack occur when two or more malicious nodes receive packets at one point of the network and transmit them to another malicious node by a wired or wireless tunnel; wormhole is carried out easily without requiring compromising sensor nodes in the network. Also, the wormhole attack may cause other type of attacks like DoS or Sybil attack. Using a cryptographic technique is not enough to prevent wormhole attack. In this study, a new scheme is proposed based on modifying the forwarding packet process, selecting the best neighbour and using hop count technique in order to prevent and detect wormhole and thus, reduce the number of untrusted packets in geographic routing protocols. Performance evaluation of the proposed scheme is done using mathematical modelling and simulation. The results are compared with some related works such as Authentication of Nodes Scheme (ANS) and Reverse Routing Scheme (RRS).