Data Compression Based On Temporal And Spatial Correlation In Wireless Sensor Networks

The main goal of wireless sensor networks is to obtain accurate information about the environment or event by deploying a large number of sensor nodes which sample continuously, detect reliably and estimate characteristics of events in the monitoring area. Thus, the similarity between data collected by the same node at adjacent time cause the time redundancy and the similarity between data collected by the neighboring nodes in the same or adjacent time cause the spatial redundancy. If the sensor nodes transmit these original data which carry large amounts of redundancy directly, it will lead to a waste of bandwidth and a big network delay,which will seriously affect the stability and service life of the whole sensor network system. Data compression techniques consider how to effectively deal with a lot of redundant data from a global, retain the useful information transmitted, then transmit the processed data from the node to the destination and efficiently use of the energy of the entire network infrastructure from a data processing point of view. For wireless sensor networks data redundancy problem, this paper does a research on the data compression algorithm based on the temporal and spatial correlation in wireless sensor networks.The energy consumption of wireless sensor networks node is mainly reflected in the three stage of data acquisition, data transmission and data processing. In the data acquisition stage, the time linear model adjustment algorithm adjusts the sampling frequency in real-time to avoid " under sampling" and " over sampling "problems effectively. The mobile agent routing algorithm based on distortion function minimizes the number of node in the collection and the spatial correlation coefficient between two nodes in a certain degree of distortion function threshold range. In a word, under the premise of ensuring the user accuracy requirements, combined with the temporal and spatial data correlation of sensor nodes, the data compression algorithm can minimize the amount of data acquisition, reduce the energy consumption caused by the spatial and temporal redundancy. Also in the data transmission and data processing stage, the lifting wavelet transform algorithm represents the sampled data with several coefficients which are far less than the original data. Furthermore, the measured values transmit according to the optimal path, therefor reducing the energy consumption due to data compressing and data transmission.According to the simulation and algorithm performance analysis, this paper verifies that the data compression algorithm considering the temporal and spatial correlation in wireless sensor networks can effectively reduce the energy consumption of sensor nodes and prolong the lifetime of wireless sensor networks.