Prediction Based Compression Algorithm for Univariate and Multivariate Data in Wireless Communication Networks

Abstract

In wireless sensor network, energy is primary constraint in designing and installation. WSN network possess many sensor nodes and each are equipped with batteries as its power bank. Due to the usage of many sensor nodes and batteries, there comes need to reduce the power consumption and increase the efficiency. The data sensed can be compressed before transmission to improve efficiency and reduce power ingestion. The data sensed through WSN network are time series data. Data prediction plays a crucial role in forecasting data with past values, which reduce the usage of sensor nodes in sensing information. This paper proposes the differencing based prediction algorithm and leveling based compression algorithm for both univariate and multivariate data. This system reduces the number of sources and power consumption. The algorithms were simulated using a simulation tool MATLAB 8.2- R2013a. The algorithms are quantified by calculating approximation error, compression ratio, and computation complexity on the compressed data. From the results it is evident that the proposed algorithm is better suitable for compression of univariate and multivariate signals.