A Research On Data Compression Algorithm In Wireless Sensor Networks

The main task of Wireless Sensor Network(WSN) is to collect monitoring data.It’s decided that the sensor network had to use quantity huge sensor nodes to meet theneeds of practical application. Therefore, researchers need to solve both problems ofdisseminating data and reducing sensor nodes’ energy consumption.How to pre-process the collected raw data and data compression under thelimited network node bandwidth and energy, which is not only the key to reduce datatransmission delay, but also the focus of this thesis.The thesis studies the way of compression and filtering data by sensor nodes inthe wireless sensor network, including based on the standard deviation unusual dataauto-adapted filtration, based on differential coding’s data compression, as well as thespace and time two-dimensional wavelet data compression which optimizes based onthe difference. The prime task induction is as follows:The thesis firstly introduces the emergence, development, characteristics,applications and key technologies of the sensor network. Then, it introduces the maindata compression methods classified and highlights some of the sensor networkapplications and sensor networks， some common methods of wavelet datacompression. It also introduces the data processing methods in the sensor networkwith difference code. Finally, it introduces the wavelet data compression methodsbased on filter and compares then by simulating experiments.This paper presents a data compression method based on filters for the researchof data compression in sensor network. The experimental data of the verificationalgorithm came from Intel-UC Berkeley joint research experiments, includingtemperature, humidity and the corresponding voltage signals. It also compared thecompression ratio of algorithm in different conditions with the experiments.This thesis also proposed a difference optimized two-dimensional wavelet datacompression algorithm. It analysis the principle and process of the wavelet datacompression optimized by difference optimized in detail. Finally, it simulated theperformance of compression method and verify the validity.