Distortion Analysis for Real-Time Reconstruction of Correlated Data Field in Heterogeneous Sensor Networks

This paper deals with the analysis of distortion introduced in the real time reconstruction of a spatio-temporally correlated field by a sink node. The field is measured by a randomly deployed network of heterogeneous sensors and the data is sent to the sink node. For the sake of analysis, we assume that the network consists of multiple clusters consisting of two types of nodes: cluster head and cluster member nodes. Given the initial energy capacities for both types of nodes, we determine the number of nodes needed for each type such that balanced energy consumption is achieved. Given the number of nodes of each type available in the network, we set up a mathematical model and analyze the average distortion in the reconstructed field. We study the relationship of this distortion with varying number of nodes, and determine the minimum number of nodes of each type needed to realize the reconstruction within a given distortion constraint. The work presented in this paper can be extended to consider different clustering methods, intra-cluster transmission schemes, and signal models.