Detection of Sybil Nodes in Wireless Sensor Networks

Abstract

Wireless sensor networks are prone to different types of attacks from malicious sources. These attacks are aimed at generating incorrect data or modifying legitimate data that is in transit in the network. A Sybil attack is a type of attack on WSNs where a malicious node either fabricates some new identities for itself or steals identities of some legitimate nodes. It can be countered in three basic methods. The first method verifies if a particular identity actually belongs to a real unique entity. The second method manages the cost and profit of acquiring identities. The third method focuses on confining the effects of the Sybil attack. However, these methods may not be used in wireless sensor networks directly. This is so because there are restrictions on the computational and storage capabilities of sensor nodes. Currently some of the principal detection schemes that are used in wireless sensor networks are radio resource testing, verification of key sets for random key predistribution, registration and position verification. The radio resource verification defense can be broken with custom radio hardware, and validation may be costly in terms of energy. Position verification can only put a limitation on how many Sybil nodes an attacker can generate unless it is able to very precisely verify node positions. Node registration needs human effort to securely add nodes to the network, and involves a way to securely maintain and query the current known topology information. The proposed method proposes a new approach to authenticate a sensor node based on a cryptographic hash function. It detects Sybil nodes so that they are isolated and any type of Sybil attack is avoided. This method performs well in terms of energy and memory usage when compared to previous methods.