Research And Application Of Key Technology Of Secure Data Fusion In Wireless Sensor Network

Wireless sensor networks are composed of sensor nodes, which are self-organized and utilize the multi-hop mode of communication to real-time monitoring, processing and transmission of the perceived object data information of the coverage area, a process which has been applied in a wide range of fields, including military defense, environment monitoring, logistics tracking, ecological protection and more. Through the effective processing of the original data, data fusion technology not only reduces the redundant information of the network, saves network energy and improves the accuracy of the data, but it also plays an important role in picture processing and decision making. However, for fusion data eavesdropping, tampering and forgery, the captured aggregation node, node intrusion and other attacks have brought many new security risks and challenges to the network. Work based on data fusion in wireless sensor networks for key security issues, such as confidentiality and privacy protection of the data, focuses on the security data fusion schemes based on the trust management mechanism, credible data fusion and preserving reliable data privacy, analyzes network performance and the security performance of the scheme.(1) In wireless sensor networks, the traditional encryption technology cannot effectively solve the security risks caused by captured nodes to attack within the network. Through calculations and evaluation of the trust value of sensor nodes, the trust management mechanism can timely recognize the interference of captured nodes and malicious nodes, identify incorrect data and, therefore, effectively improve the security performance of data fusion from inside the network. In view of the shortage of the higher trust value’s nodes, excessive use caused by these nodes’energy consumption rapidly in the RDAT algorithm and the shortage of the reliability decline of the data link affected the credibility of data in the iRTEDA algorithm. Work propose an energy-efficient, reliable trust-based data aggregation protocol for WSN called the ERTDA protocol, which can effectively detect the captured nodes, guarantee the availability and reliability of the data transmission link and extend the life of the network effectively while safeguarding network security.(2) In the process of data fusion, reputation systems can cover the shortages of cryptology. It can estimate the reliabilities of data generated by the nodes in wireless sensor networks and further delete the malicious data, which can improve security. However, the compromised nodes can improve their reputation values by hiding their network attack behaviors and their fault data can mislead the fusion results significantly; additionally, affected by the surrounding environments, the normal sensors may also have a large measurement error and can mislead the fusion results. To solve the problems above, work propose a secure data fusion algorithm named RDIF, which can be used in networks with layered routing algorithms. In RDIF, both the reputation values of the nodes and the data values are considered to delete the fault data based on the modified DBSCAN clustering algorithm. Then, the remaining data are fused on the cluster heads, effectively improving the security and accuracy of the data fusion.(3) The purpose of data privacy protection is to ensure that the private data in the network is not accessed by unauthorized users, even if the nodes’monitoring data in the network is intercepted, also preventing the attacker from cracking the important sensitive data information of the nodes. In wireless sensor networks, when illegal users are using eavesdropping, tampering and other attack means to access the data, it is a serious threat to data privacy. Work propose a secure data fusion algorithm named FTLEDA, based on the data cutting idea, in the network of tree structure, where the decision method of different data cutting times is respectively provided for the leaf nodes, the child node and the cluster head. The FTLEDA algorithm can effectively protect the data privacy, significantly reduce the energy consumption and improve the accuracy of data fusion.(4) Most of the early forest fire monitoring approaches are based on satellite remote sensing technology, and the application scenarios of these approaches are constrained because of quality of remote sensing images. As a result, the approaches based on wireless sensor networks, which are more robust and flexible, are widely researched. However, these approaches may also misjudge the fires because they lack interactivity with the users. In consideration of the severity of the forest fire, such misjudgments cause big losses. Work proposes a novel forest fire monitoring approach named FFMP, based on mixed wireless sensor networks. This approach integrates energy-efficient coverage, nodes motion strategies, data fusion and data visualization. In FFMP, static nodes are used to monitor the environment and mobile nodes compose a layer of data transmission. Before being transmitted to the users, the data generated by the static nodes are pre-processed by the mobile nodes to reduce the transmission cost. The users visualize the data by Delaunay triangulation and locate the fire, which can be used to predict the development tendency. There are four goals when designing the approach:early forest fire monitoring, establishing temperature distribution, fire location and saving energy to prolong the lifetime of the networks.