Data Compression Algorithm Research Based On Wavelet For Wireless Sensor Networks

Wireless sensor networks is currently a very active area, have been widely used in many ways, it composed by many sensor nodes which have the power wireless transmitter-receiver set, the WSN collect information and transmit it to the remote base station to carry on processing. Limited resources of sensor nodes, they carry their own battery power and it is difficult to add them, therefore, the energy problems is one of the issues of sensor nodes has been troubled by researchers, how to use technical means to achieve the purpose of saving energy is the research focus in wireless sensor networks.Data compression is a practical technique to save energy and prolong the life cycle of wireless sensor networks. Existing research shows that communication energy consumption more than the computational overhead in wireless sensor networks, data compression by removing redundant information to reduce the amount of data in the transmission of information, save the limited communication bandwidth and energy. The classical data compression is based on information theory, the class of statistics codes such as RLE and dictionary-based coding have been applied in wireless sensor networks. Data compression algorithm based on wavelet has been widely researched in the present; this is because the wavelet has good time-frequency characteristics and compared with the traditional data compression algorithm has a better compression.The data collected by sensor nodes is often having relevance of some degree, using the relationship between a variety of data can reduce the redundant information, and enhance the effect of data compression. Because the traditional wavelet interval is difficult to match the sensor signals interval in general, using traditional wavelet to be compressed may encounter the problems of boundary effects, thereby increasing the cost. This paper proposes a multiple data streams compression algorithm based on interval wavelet, simulated and compared with other three compression algorithm, obtained better compression results and reduce the energy consumption.The data d and time t have some degree of relationship in the process of data collection with sensor nodes, based on the characteristics of sensor network; this paper proposed a time-linear model, use the model to reduce the redundant data. Compared with the discrete wavelet transform, the lifting wavelet has the advantages of ascended the throne in operation and fast speed etc. Using lifting wavelet to compression the data can get better results, to save the limited energy and storage space. Based on the above two points, we propose a data compression algorithm based on time-linear model and Lifting Wavelet, compared with other two compression algorithms, the simulation results show that the indicator have been optimized.