Study On Secure Data Aggregation Algorithm Based On Mutual Monitoring Mechanism In WSN

After the Internet, wireless sensor network is another emerging technology, which integrated sensors, computers and wireless communications technology, has become one of the research focus at home and abroad, is being widely used in military, environmental monitoring, family medicine and other fields. Wireless sensor networks are generally densely deployed, large-scale network, there is a certain degree of data redundancy, a single sensor node energy is limited, so to extend the network life cycle, saving energy as the main objective of the wireless sensor network design. Research indicates that the most consumption of energy is during data transmission. Data aggregation technology can effectively eliminated redundant data, reduce the transmission of data packets, and thus become an effective solution to alleviate the bottleneck problem of the wireless sensor network resources approach. However, due to some of the limitations of wireless sensor networks, data aggregation may be vulnerable to energy depletion attacks, false data injection attacks and a variety of attacks. Against external attacks, we can use traditional encryption mechanisms to ensure security. But due to deploying in the dangerous area, the sensor nodes can easily be captured and lead to internal attacks. Communication key and other secret information will be exposed. Thus false data will be sent to consume the network energy and affect the accuracy of data aggregation.In order to reduce or eliminate the impact of captured node, which may modify or inject false data on the data aggregation values, a data aggregation algorithm based on mutual monitoring mechanism was proposed in this thesis. Firstly, a secure data aggregation algorithm was proposed to filter out the false or abnormal data, to get effective, high quality data, and finally get an accurate data aggregation results. Secondly, design a member node, aggregator node mutual monitoring mechanism and algorithm was designed. The aggregator node can detect the captured member nodes by comparing the sensed data of member node and the final data aggregation value, updating the normal data counter and abnormal data counter of the member nodes. Member nodes can detect the captured aggregator nodes by useing listening technology to monitor the aggregator values which are sent out by aggregator node. Finally, the impact of data aggregation value once the node is captured was analyzed. Simulation results show that the proposed algorithm can effectively detect malicious nodes and guarantee the data aggregation effective based on the premise that energy consumption is not so large.