Research On Detecting And Defending Sybil Attack In Wireless Sensor Networks

With the development of wireless communication and microelectronic technique, wirelesssensor networks (WSNs) have become a hot topic in academia and industry at home and abroad. Ithas been considered to be a very promising application technique. WSNs are often deployed inharsh hostile areas or unattended wild fields. Coupled with the inherent resource constraints, thesecurity issue has aroused great concern. As a major attack appeared in WSNs, Sybil attack do agreat harm to the network in many aspects, such as distributed routing protocols, resource allocation,competitive voting mechanism and abnormal behavior detection, and so on. Some traditionalapproaches against Sybil attack are not suitable for the developing WSNs because of the poorsecurity and low detection rate. Sybil attack in WSNs is studyed in this thesis. And the main tasksare described as follows:(1) The concept of WSNs and the current security issues that it is facing are summarized andanalyzed. And the characteristics and attack modes of Sybil attack are also summarized. Besides,the existing Sybil defense and detection approaches are analyzed in detail.(2) In order to solve the problems appeared in the existing defense approaches, distributedcommunication and key predistribution scheme are took advantage to propose an approach againstSybil attack based on chain key predistribution. During the chain key predistribution phase, ourapproach uses a lightweight hash function to generate several c hain keys by hashing the uniqueidentity information of every node sequentially in the trusted base station. And these chain keysconstruct a pool of chain keys. Then during the pairwise key authentication establishment phase, anode-to-node chain key based authentication and exchange (CK-AE) protocol is proposed. In thisprotocol, each node shares the unique pairwise key with its neighboring node. Finally, the CK-AEprotocol is provably secure in the universally composable security model. The analysis indicatesthat our approach can not only improve the security of the predistributed keys in each node, but alsoenhance the ability of resilience to Sybil attack. And the results show that our approach also reducesthe communication overhead significantly during the pairwise key authentication establishmentphase.(3) To solve the problems in the existing Sybil attack detection methods which are under powercontrol condition, such as low detection rate, complicated algorithm and idealized network modelfor Sybil attack detection, a Sybil attack detection scheme based on nodes cooperation for power control performing nodes in WSNs is proposed. First based on large scale propagation model whichincludes the path loss and shadow fading, twice difference of RSSI is analyzed. Then according tothe characteristic of twice difference of RSSI which remains unaffected to the change of node’stransmission power, a Sybil node identification algorithm based on observation nodes cooperation(SIOC) is proposed. The algorithm is implemented by advanced observation nodes collaborativelyto identify Sybil nodes. The simulation results show that the approach can not only enhance thedetection accuracy, but also extend network lifecycle.