A Research On Wavelet Data Compression Algorithms For Wireless Sensor Networks

Wireless sensor networks have been the targets of active research in the recent years due to its military and civil applications. Wireless sensor networks usually have limited energy and transmission capacity, and they can't match the transmission of a large number of data. So, it is necessary to perform in-network compression of raw data sampled by sensors. In this paper, we introduce the problem of data compression for wireless sensor networks on the basis of wavelet techniques. Our main works include the foliowings:(1) The related works of data compression for wireless sensor networks are classified from three points of view: in-networking data compression, wavelet data compression and route protocol.(2) A wavelet compression algorithm based on logical mapping is proposed. For the sensory data received from sensor nodes, cluster heads don't store them according to the order of time, but design a mapping to produce a new data set. The mapping data set is formed on the basis of the data correlation and wavelet function. Aiming at the mapping data set, a wavelet compression algorithm is presented to obtain high compression efficiency.(3) A distributed wavelet algorithm is proposed based on chain topology. First, we construct a chain topology and design an algorithm to decide the optimal positions storing the low-frequency wavelet coefficients. Then the chain, which is used to perform distributed wavelet transform, is decided. A distributed compression algorithm using wavelet function with arbitrary supports is then presented. Our algorithm can eliminate the redundant data before their transmission to the cluster head and thus the energy consumption is decreased.(4) A wavelet compression algorithm with compression rate adaptivity is proposed. According to the existing communication bandwidth of sensor nodes, the quantizer and the encoder used to wavelet compression are selected. We then design an algorithm to filter the proper high-frequency wavelet coefficients, and thus adjust the compression rate adaptively along with quantizer and encoder. So, The output of sensor nodes can match the existing communication bandwidth.