دانلود رایگان مقاله تشخیص فریبکار برای تکرار حملات در شبکه های حسگر موبایل

Abstract

In a node replication attack, an adversary creates replicas of captured sensor nodes in an attempt to control information that is reaching the base station or, more generally, compromise the functionality of the network. In this work, we develop fully distributed and completely decentralized schemes to detect and evict multiple imposters in mobile wireless sensor networks (MWSNs). The proposed schemes not only quarantines these malicious nodes but also withstand collusion against collaborating imposters trying to blacklist legitimate nodes of the network. Hence the completeness and soundness of the protocols is guaranteed. Our protocols are coupled with extensive mathematical and experimental results, proving the viability of our proposals, thus making them fit for realistic mobile sensor network deployments.

6. Conclusions

The node replication attack is one of the most insidious attacks in sensor networks. Although several countermeasures exist, almost all practical schemes assume a stationary network model where sensor nodes are fixed and immobile. In this work, we proposed solutions that can be used to detect imposters in mobile sensor networks, where nodes freely and randomly move around in the sensing region. These schemes are fully distributed and completely decentralized. Contrary to prior work, the proposed schemes can effectively detect and quarantine the presence of mul-tiple imposters faking the identity of different legitimate nodes in the network. We have proved the completeness and soundness of the proposed detection methods and came up with schemes that are adaptive in their operation; even if the number of imposters is unknown, the adaptive schemes not only find all imposters but also respond in a self-healing manner eventually bringing the number of false-positives to zero. Hence they constitute the preferred method for realistic sensor network deployments. Our findings were coupled with both analytical and experimental results, proving the viability of our proposals. Note that when an imposter is detected, the ID possessed by the imposter is considered compromised. As a result, the remaining nodes stop communicating with the imposters but also with the legitimate nodes bearing the same IDs. In terms of future work, it would be interesting to investigate the possibility of re-instating the legitimate nodes back in the network. This could be done by reprogramming those nodes [24,28], hence installing new software, new cryptographic material, and as a result, new IDs to them.