Data Compression using Distributed Source Coding in Wireless Sensor Network

In this paper distributed source coding is used to compress data in a wireless sensor network in order to reduce communication energy in sensor nodes. Power is a precious resource in wireless sensor networks due to the limited battery capacity. Distributed source coding refers to the compression of the output of multiple correlated sensor that do not communicate with each other. In this paper, we first study about distributed source coding. Then we study how estimates the number of clusters needed to efficiently utilize data correlation of sensors for a general sensor network. We then present an approximation algorithm that can find an optimal rate allocation within each cluster to minimize the intra cluster communication cost. Then we evaluate the effect of spatial correlation and network size on optimum number of cluster, overall compression ratio and intra-cluster communication cost. I. Introduction In a wireless sensor network (WSN), a number of sensor nodes are densely deployed in a field of interest with one or more data sinks located either at the center or out of the field [1]. The sensor nodes sense the phenomenon at different points in the field and process the data and then finally send the data to the sink(s). Power consumption in the sensor node is for Sensing, Data Processing and Communication. More energy is required for data communication in sensor node. Energy expenditure is less for sensing and data processing. As we know that power is a precious resource in wireless sensor networks due to the limited battery capacity. Once deployed it is often difficult to charge or replace the batteries for these nodes. The capacity of batteries is not expected to improve much in the future. This work explores energy consumption trade-offs associated with lossless data compression. The data compression techniques extend the life time of sensor network. Also by reducing data size less band width is required for sending and receiving data. The observed phenomenon is usually a spatially dependent continuous process, in which the observed data have a certain spatial correlation. In general, the degree of the spatial correlation in the data increases with the decrease of the separation between sensor nodes. Therefore, spatially proximal sensor observations are highly correlated, which leads to considerable data redundancy in the network [2].Slepian Wolf coding [4] is a Distributed source coding that can completely remove data redundancy without requiring inter-sensor communication and therefore a promising technique …