Abstract: Wireless Sensor Networks (WSNs) is temporarily formed, operated and managed by the nodes themselves. In a complex WSN, malicious nodes are well disguised, and they can attack the entire network on specific purpose by utilizing the natural cooperation of nodes. The error-prone characteristic of sensor nodes can also cause instability in WSN. Therefore, how to choose one or more suitable sensor nodes to collaborate towards a better system performance is a critical issue. Trust model can help the elements of a network to decide whether a neighbor node is being uncooperative or malicious. Trust is more important in self-configurable and autonomous systems, such as Wireless Sensor Networks (WSN). As these networks are often deployed in open environment, so they are susceptible to be attacked by some adversaries. Traditionally cryptographic mechanisms such as authentication and encryption can avoid some of these threats. In recent years utilizing of trust management is being an important security mechanism for WSNs. In this Thesis, we propose a scalable and distributed trust model for WSNs and study the resistance of that against oscillating behavior and collusion attacks. In the proposed trust model, a node maintains direct trust value for its direct neighbors, besides with indirect trust values which is obtained by other nodes. These two direct and indirect trust values are composed to find the total trust of neighbors. Trust value is a number in interval [0,1] and it has got trust specifications. The simulation results show that our model is resistant against collusion and oscillating attacks. Also it is shown that the percentage of selection trustworthy servers is not changed in networks by different number nodes. Keywords: Wireless Sensor Networks, Trust Model, Direct Trust, Indirect Trust, Radio Range |
