Link-Based Wormhole Detection In Wireless Sensor Networks

Wireless Sensor Network (WSN) is a kind of multifunctional low-power wireless ad hoc network. It is widely applied in many fields, such as military monitoring and detection, atmosphere and waterlines monitoring, forest fire prevention, landslide prediction, greenhouse irrigation, bird watching, tracking and positioning, etc. The new technology proposed and studied in the new concept of the smart home and smart city will greatly change people’s lives. Some characteristics, which are the limit of node energy and processing power in WSN, result in the WSN born with frill. Network security has become a key issue constraining the further application of new technologies.There are a lot of security threats in wireless sensor networks, such as information spoofing, tampering, repeat, flood attacks, Sybil attack, sinkhole attack, wormhole attack. A wormhole is a channel with high-quality and low-latency between two distant nodes. When a packet needs to be routed, the channel will have a great advantage since it is able to save a lot of time from packets’ routing. But the existence of the channel is not only harmful to the network routing mechanisms, but will also cause the acceleration of nodes’ energy consumption. Because of the attacker can discard or tamper with a lot of packets, it will seriously affect the parts of the state of environmental monitoring. Even though a network has some authentication and encryption mechanisms, the attacker would still be able to prevent normal network from information collection easily. So wormhole attacks become one of the major security threats, which are difficult to predict in WSN.This paper introduces the characteristics of the wormhole attacks and the security problems of the presence of wormholes in WSN. And then, a link-based wormhole intrusion detection mechanism (CBW) in WSN is proposed. It is different from some wormhole detection protocols, which don’t need any strong constraints like hardware or clock synchronization. CBW not only includes detecting whether the existence of wormhole attacks or not, but also distinguish wormhole from explicitly and hidden attacks. The wormhole channel nodes are found out in the explicit wormhole environment. Some nodes in the infected area are detected in the hidden wormhole environment.The detection system consists of four parts:ordinary nodes, anchor nodes, the root node, admin node. The OMNeT++ is used to simulate large-scale WSN with higher simulation speed and flexibility. The simulation system includes some key modules, such as the construction of the network, the settings of the network packets’ structure, information processing in nodes, links’ intersection algorithm in explicit wormhole attack, communication plan in the hidden wormhole attack.Finally, CBW is valid by experiments. Being compared with the existing method, the link-based detection method has higher detection success rate in the environment of low redundancy. By the experiment of anchor nodes’ intensity, CBW will obtain that the detection success rate is nearly 90% through the total of anchor nodes is only 0.4% in the network.