Research On Data Aggregation Security Of Wireless Sensor Networks

Data aggregation technology is one of the important technologies for data processing in wireless sensor networks,obtaining accurate query results and reducing energy consumption.It has been widely used in many different aspects.Because of the differences in energy and computational performance between wireless sensor networks and traditional networks,it is difficult to directly use most of the security technologies that have been widely used in traditional networks.Therefore,data fusion technology also has many problems in terms of security.Attackers can attack various types of attacks on the network by trapping normal nodes.As a result,data aggregation results are greatly affected.In order to solve the above problems,this thesis researches a large number of data aggregation security solutions and decides to design a secure data aggregation solution that can effectively resolve internal network attacks based on trust management.The main research work and results are as follows:(1)For wireless sensor networks,traditional encryption mechanisms do not work well for attacks initiated by compromised nodes within the network.The thesis design the suitable for WSNs structure model of trust management mechanism,and established the corresponding on the basis of the Beta distribution model,through the network deployment of monitoring mechanism,using each other between the nodes monitor its behavior in response to a neighboring node may initiate malicious nodes forward and malicious collection attacks,and at the same time within the node constitute a node list.Based on this list,the direct trust value of the node to its adjacent node is calculated.In the calculation of the trust value,the hidden attacks that may be suffered in the network are considered,and the node direct trust value calculation model is optimized to better identify such nodes.Then,the nodes to calculate its neighbor nodes in turn indirect trust value and comprehensive trust value,comprehensive trust value and will be sent to the cluster head nodes,for the unity of all nodes in the cluster trust value computation,used as the basis for the next step on trusted node selection.(2)Since the nodes in the wireless sensor network are always performing monitoring tasks and their trust values change at any time,it is not so effective to exclude malicious nodes by presetting thresholds.In view of this situation,this thesis designs a corresponding security scheme based on the trust management model of wireless sensor networks,and fully considers the trust value of nodes within the cluster and the information obtained thro ugh monitoring.By combining the improved model with the clustering algorithm,the trust value from the previous stage is used,and the trust value is substituted into the clustering algorithm for classification,so that the captured nodes existing in the network can be dynamically detected.In the data aggregation process,the information transmitted by the malicious node is removed,and the data sent by the group of nodes with the highest trust values in the clustering result is combined with the trust value as a weighted calculation,and finally a reliable fusion result is obtained,thereby ensuring the accuracy of data aggregation results.Finally,the proposed scheme is compared with RDAT scheme through OMNET++ simulation.Analyze the change of the trust value of normal and malicious nodes under the same conditions for the two schemes.According to the proportion of malicious nodes,several corresponding simulation experiments were performed,and the detection rate of malicious nodes was compared and analyzed.At the same time,it also laterally compares the results of the aggregation with the RDAT solution,which proves the effectiveness and accuracy of the proposed protocol in ensuring the data aggregation security of wireless sensor networks.