Research On Data Aggregation Algorithm Based On TEEN Protocol In Wireless Sensor Networks

In recent years, with the rapid development of Internet of things technology around the world, smart city, intelligent home and other projects have been proposed and gradually promote. Chinese government is willing to advocate "Internet Plus" program in future. All these above contribute to the development of wireless sensor networks. However, the network nodes die prematurely due to the high local load and limited battery capacity. The shortage of network life is still the core problem of wireless sensor network. Data aggregation technology has been innovated continuously in the recent forty years after its birth. In wireless sensor networks, most of the energy is used in the data transmission module,while the energy consumption of perception and processing module is relatively small. With the help of data aggregation technology, we can effectively reduce the network data forwarding and the channel congestion, thus saving the node energy and prolonging the network lifetime.Based on this, this paper designs a data fusion algorithm on the basis of studying the existing TEEN routing protocol of wireless sensor network, aiming at the problem that the inherent noise interference of sensor and data aggregation algorithm is not clear. In this paper, in order to reduce the data error and improve the data accuracy, the original monitoring data is filtered by the Kalman filter algorithm at the source node. Combined with the detection rule of the soft and hard threshold of TEEN routing protocol, less redundant data will be forwarded to the sink node. At the sink node, the Grubbs criterion is used to check the forwarding data and eliminate the anomaly data. Then, by computing the single round data after arithmetic averaging, the max-min closeness principle is used. The weighting data of each source node are weighted according to their weights, which can reduce the redundant data and reduce the data transfer to the base station while guaranteeing certain aggregation precision.At last, the data aggregation algorithm proposed in this paper is verified by MATLAB simulation. The simulation results show that by means of Kalman filtering and threshold processing at the source node,the forwarding of redundant data is reduced. Combined with the data accuracy correction at the convergence node, the data forwarding is reduced for the whole network. Energy consumption is reduced in the circumstances of equal initial energy, and the node and the overall life of the network is extended.