Research On Rapid Data Collection And Secure Data Fusion In WSN

Wireless sensor network is a data-centric network which is composed of manysensor nodes. It can monitor, collect and process the information of monitoring object,and its application prospect is very broad. However, the sensor nodes in wireless sensornetworks are faced with the constraint of resources, as well as security threats. Since theoriginal sensor data to the base station nodes usually needs to go through stages of datacollection and data fusion. Using data collection methods based on prediction to processsensor data in the data collection phase can effectively reduce the energy consumptionof communication between nodes and extend network lifetime. Meanwhile, aggregatingthe collected data and taking appropriate security policies in the transmission of datafusion stage can effectively improve network security and ensure fusion results to reachbase node safely.Collecting sensor data based on auto regression model AR(p) is an effectivemethod to reduce the frequency of data communication between cluster head and leafnodes,and lower energy consumption of nodes in wireless sensor network. However,the traditional AR (p) model ignores the different influence of historical data in differentperiods on the predictive value in the modeling process and affects the predictionaccuracy of prediction model and the frequency of data communication in WSN. In thisregard, an improved AR(p) prediction model which is known as FAR(p) is proposed.By specifying a weight value for each historical data with a an appropriate fuzzymembership function introduced in AR(p) to weaken the impact of early data on thepredictive value in data sequence and strengthen the role of recent data on the predictivevalue, then reconstruct post-processing prediction model for data collection afterprocessing data and weights by weighted average second-order weakening bufferalgorithm. Finally, the simulation results show that the improved model can effectivelyimprove prediction accuracy of the predictive model, reduce data communication anddecrease energy consumption in WSNAt the same time, taking into account the data that collected by cluster head nodeis facing serious security problems during integration process and transport process, thispaper proposes a new security algorithms called HEHMAC which can confidentialityand verify integrity during transmission and integration processes timely. HEHMACalgorithm using homomorphic encryption transmission mechanisms to protect data privacy, while avoiding encrypting and decrypting data hop-to-hop, can effectivelyreduce the energy consumption of computing and communications. At the same time,using homomorphic message authentication to verify the integrity of fusion result byhop authentication, which promptly identifying data integrity and discarding false dataduring transmission and integration process, can reduce unnecessary communicationenergy. Finally, by comparing experimental results and analyzing theoretical to confirmthe safety and efficacy of HEHMAC algorithm.