False Data Filtering In Wireless Sensor Networks

Wireless sensor network (WSN) is one of the hot-spots in current research, and it holds a broad application foreground. In WSN, the nodes exhibit characteristics such as low power, weak communication and computational ability, limited storage space, etc. WSN is often deployed in unattended or even hostile environments, thus the nodes are easy to be compromised by the adversary. As nodes are often not equipped with tamper-resistant devices, and thus the keys stored in them will be disclosed when being compromised.The adversary may abuse the compromised nodes to inject false data reports, tamper the transmitted legitimate reports, or send replayed reports into the network, which may cause not only false alarms, but also the depletion of the limited energy. In this thesis we focus on the problem that how to detect and filter out false reports in wireless sensor networks. The main research aspects are as follows.(1) Aiming at the problems of existing schemes' low filtering probability and unbalanced overhead among nodes, we propose a cluster-based false data filtering scheme CFFS. First, nodes are grouped into clusters and a Sink-rooted tree of all cluster heads is constructed. We further propose a distributed key assignment method based on the overhead of nodes, which guarantees that the clusters closer to a source cluster hold more authentication keys for the source cluster than those clusters farther from it do. As a result, false reports generated by the source cluster can be filtered out within little hops. Analysis and simulation results show that CFFS outperforms existing schemes in filtering efficiency, and achieves more balanced key distribution.(2) Aiming at the problem that existing schemes cannot resist the collaborative false data injection attacks, we propose a Geographical information based False reports Filtering Scheme (GFFS) in sensor networks. In GFFS, each node distributes its location information to some downstream nodes after being deployed. Each data report must carry the MACs and locations of t detecting nodes that sensed the event simultaneously. All forwarding nodes then validate not only the correctness of the MACs and locations, but also the legitimacy of the locations. As a result, false reports injected collaboratively by compromised nodes from different geographical areas can be detected and filtered out. For exampe, with ten compromised nodes, the probability for the attacker to break down SEF and GFFS is93.2%,3%, respectively.(3) Considering that GFFS requires the support of expensive positioning devices, we further propose a NFFS scheme. In NFFS, each node distributes its neighbor information to some other nodes after deployment. When a report is generated for an observed event, it must carry the IDs and MACs from t detecting nodes. Each forwarding node checks not only the correctness of the MACs carried in the report, but also the legitimacy of the relative position of these detecting nodes. As a result, collaborative false data injection attacks can be resisted efficiently. For example, with ten compromised nodes, there is a probability of only0.7%for the attacker to break down NFFS.(4) Aiming at the problem that existing schemes can detect and filter out only false reports but the replayed reports during forwarding, or can only detect replayed reports by the Sink but not the forwarding nodes, we propose a Neighbor information and one-way Hash chain based Filtering Scheme (NHFS). In NHFS, each node distributes its initial hash value to some other nodes after deployment. When a report is generated for an observed event, it must carry the MACs and fresh hash values from t detecting nodes. Each forwarding node checks the correctness of the MACs and hash values, and the freshness of these hash values. Analysis and simulation results show that NHFS can detect and filter out false reports and replayed reports efficiently.In a word, the dissertation performs an in-depth study on the filtering of false data reports in wireless sensor networks. Moreover, several filtering protocols with higher performance are proposed. Therefore, the research has strong theoretical and practical significances.