Energy trust system for detecting sybil attacks in clustered wireless sensor network

Recently, much more attention has been attracted in the wireless sensor networks.It consists of a large number of small sensing self-powered nodes with resource constraints such as limited computing capability, memory, and energy. Hencesensor nodes generally are deployed in the remote and hostile environments; it is challenging to provide security to the sensor nodes. Furthermore, the specific constraints of the wireless sensor network make the problems even more critical. These networks are susceptible to different kinds of attacks like denial of service attacks, sybil attacks, jamming attacks, black/sink hole attacks (dropping and absorbing of the packets), and slandering attacks. However, the sybil attacks pose as a serious threat because it can be a gateway to other attacks like data aggregation, distributed storage, voting, resource allocation, and misbehavior detection. The attack occurs when a malicious node, called sybil node, illegitimately claims multiple fake identities by either fabricating new identities or impersonating existing ones. Therefore, the detection of a sybil attack in the network is very important. However, the existing sybil detection approaches have shortage to sufface for the constraints of WSNs and the nature of sybil attacks. To address these limitations, this thesis utilizes the concept of trust systems to protect the network from sybil attacks by providing multi-level detection which could work in a hierarchical wireless sensor network. For each level of detection, energy trust system was applied. Specifically, 1st level detection in each cluster head and 2nd level detection in the base station. Furthermore, a centralized management scheme was employed. Aggregation was also applied to avoid communication overhead and save energy. The proposed system was evaluated in terms of memory overhead, communication overhead, and consumed energy. Furthermore, the performance overhead and the detection accuracy were carried out through intensive simulations, as well as extensive comparison with other trust approaches (LDTS and GTMS). The results from the evaluation indicate that the proposed energy trust system for the detection of sybil attacks can provide fast and efective detection as shown by the true and false positive rates. It showed more than 70% true positive rate and less than 30% false positive rate at the 1st level of detection. For the 2nd level of detection it showed better performance reach to 100% true positive. Moreover, the proposed system was introduced light overhead and storage.

Preview Text FK 2016 69 - IR.pdf Download (1MB) | Preview

Actions (login required)

View Item