Journal of Computer Applications (计算机应用) 2013/33:12 PP.3453-3456
Concerning to the problem that Wireless Sensor Network (WSN) data collection has large data redundancy, large cumulative error and low data accuracy, according to the temporal correlation between collection data, a data compression and optimization algorithm for WSN was proposed. It established segmented one-dimensional linear regression model by analyzing linear relationship of collection data in temporal series. According to the error between collection data and predicted value of regression model, it adaptively adjusted next collection time, and dynamically adjusted the regression model. The simulation results show that the proposed algorithm can reduce the data redundancy and network traffic, and improve the reconfiguration precision of the collection data under different conditions of data changes. The test results in a real scenario show the feasibility of the proposed algorithm.