Distortion Analysis for Real-Time Data Gathering of Spatially-temporally Correlated Fields in Sensor Networks

摘要

This paper presents a framework for analyzing spatiotemporal distortion associated with data gathering algorithms in sensor networks. We assume that the underlying data field being sensed is spatially and -temporally correlated and the sink performs real-time reconstruction of the field. Unlike existing work, our analysis framework incorporates realistic network/routing parameters, such as congestion and limited transmission range. Our numerical simulations show that even for a simple collision-free transmission scheme and regular grid deployment, the distortion analyses yield starkly different results when network parameters are considered. The analytical method presented in this paper can be easily extended to consider multidimensional deployments, different transmission schemes, nodes distributions, and reconstruction methods.