Research On Energy Saving Methods Based On LS-SVR For Data Transmission In Wireless Sensor Networks

With the development of science and technology, especially the development of electronic technology and manufacturing industry, wireless sensor network (WSN) technology is getting great development. In order to get data of some areas where people interested in we have to deploy a large amount of sensors, thus people would make the cost of sensor lower. This kind of sensor is limited in energy and only can do a small amount of calculation. Therefore, how to use the energy efficiently and to maximize the lifetime of wireless sensor network is the primary challenge in WSN. Among many energy saving methods of wireless sensor networks (WSN), the research of wireless transmission technology and MAC protocols is from the underlying basic, routing algorithms is designed to balance the energy consumption of all nodes from an overall point of view to prolong the lifetime of the network, while sleeping mechanism has the best efficiency to save energy and is widely used.Under the fact that the energy consumption of data receiving is less than that of data sending, this paper proposes an energy saving method based on least squares support vector regression by receiving more instead of sending to reduce energy consumption fundamentally during the data transmission process. As long as the prediction model is accurate enough, we can greatly reduce the times of sending to achieve the purpose of energy saving. The performance of the proposed method is tested in the same conditions with traditional methods. The results show that the energy saving effect is better than that of the traditional methods, and the effectiveness of the method is demonstrated.Also, considering the cumulative error of prediction model, the method was optimized. When it's out of error, the sensor node would send a collected-data sequence to the sink node, which can help predicting accurate data, and improve the credibility of the energy-saving method. Considering of the effectiveness of this optimized method and the characteristics of a certain type of data, we presents a simple compression algorithm to reduce the energy consumption of data transmission of which the effectiveness is proved in the later simulation experiment.