Virtual cluster tree based distributed data classification strategy using locally linear embedding in wireless sensor network

Abstract

With recent advances in wireless communication and low cost, low power sensors are enabling the deployment of large-scale and collaborative wireless sensor network (WSN), which performs different tasks using the subsets formation of sensor on specific requirement of event monitoring. For implementing the differentiated monitoring task of the heterogeneous sensor network in the same geographical region, efficient classification of the various sensed data becomes a critical task due to stringent constraint on network resources, frequent link and indeterminate variations in sensor readings. In this paper, we present a virtual cluster tree based distributed data classification strategy using locally linear embedding (LLE). The strategy can realize the structure representation of original sensed data by LLE to meet the classification for forming the virtual cluster tree of the various monitoring task. The theoretical analysis and experimental results show that the proposed strategy can effectively reduce the energy consumption using LLE based classifier. © 2012 Springer-Verlag.