Abstract: In wireless sensor networks, the traditional event detection algorithm has the problem of high network transmission and energy consumption. In order to solve such problem, this paper presents a new complex event detection algorithm based on pattern matching and correlation analysis. Complex events not only change the data of a single node, but also be the impact of the entire sensor network. So it is often manifested as a particular pattern at the sensor node and the correlation between nodes satisfy some relationship. According to the characteristics of complex event showed, event detection is divided into two parts: the feature detection at single node and the correlation analysis between multiple nodes. Calculation pressure is allocated to each node by the feature extraction and pattern matching. So the algorithm can control the frequency of the calculation of the correlation analysis, and it will extend the network lifetime. In order to reduce the amount of network traffic, the compressed data is used for correlation analysis. The experiments show that the proposed algorithm can effectively reduce energy consumption.

Key words: wireless sensor networks, complex event detection, discrete cosine transform, correlation coefficient, pattern matching