Distributed Compressive Data Aggregation in Large-Scale Wireless Sensor Networks

Abstract

 Abstract—As wireless sensor networks are used extensively in environment and habitat monitoring, the large volume of data transmission can increase the workload of the sensor nodes and reduce their useful lifetime. The compressive sampling techniques have been proposed to reduce the volume of data transmission when the data is sparse in certain domain. While finding the optimal routing path that minimizes data traffic is an NP-complete problem, a near-optimal routing protocol in the literature requires omniscient knowledge of the entire network and thus incurs extensive message exchanges in real applications. In this paper, we propose a distributed algorithm that uses local minimization to dynamically construct a routing path to reduce the data traffic for compressive sampling based aggregation. This algorithm does not require the omniscient knowledge of the global network topology and incurs much lower overhead than the near optimal solution, and therefore, is more suitable for practical applications.