Research On Secure Top-k Query Processing In Two-tier Wireless Sensor Network

With the development of the sensor technology, embedded technology andwireless communication, wireless sensor networks become prevalent for pervasivecomputing and widely deployed for various applications such as environmentsensing, intrusion detection and earthquake predication, etc. Wireless sensornetwork is an ad hoc collection of low-cost, small-size sensor nodes with limitedcomputing, communication, storage and energy resource. Various inherentlimitations of it appear especially in the efficient data storage and query process.With the introduction of the storage node, extensive research has been conducted toaddress this limitation by developing a two-tier wireless sensor network.A two-tier wireless sensor network consists of large amount ofresource-limited sensor nodes at the lower-tier which sense the environmentinformation and plenty of resource rich relatively storage nodes at the upper-tierwhich gather data from the nearby sensor nodes and answer queries from the user.While storage nodes bring several benefits such as prolonging network lifetime,saving the memory of sensor nodes and processing queries efficiently, the storagenode faces serious security concerns in hostile environment. First, when the storagenode is compromised, the sensing data from the sensor nodes, the history of queryrequests and the corresponding query results are exposed. Second, it may causeheavy loss when the compromised sensor nodes return fake, forged or incompletedata for a query especially in military and commercial application. Therefore,developing a privacy-preserving and result-verifiable mechanism is of paramountimportance such that the authenticity and completeness of the query results can beverified as well as the privacy of the sensitive data is protected.To resolve all the above problems, a secure and privacy-preserving top-k queryprocess in two-tier wireless sensor network is proposed. For the security andprivacy, we present RSTOPK scheme and PriSecTOPK scheme respectively. Wedevise the RSTOPK scheme to effectively increase the detection rate and reduce theadditional communication cost for enabling verifiable top-k queries throughhypothesis testing method combined with computing commitment while the thebasic PriSecTopk scheme by using order-preserving encryption, and then improve it step by step to achieve various privacy requirements as well as the correctnessrequirements in three levels of threat models. Theoretical analysis and simulationresults successfully validate the efficacy and efficiency of the proposed schemes.