Cluster-Base Directional Rumor Routing Protocol in Wireless Sensor Network

Abstract

Predicting occurrence of events in the network has always been a notable problem in computer networks especially in wireless sensor networks. Estimating the location in which the next event occurs, can be used as an important routing factor in these networks. Events that occur randomly in the network can't be modeled with methods like curve fitting, therefore, to reach this goal, other methods like clustering are needed. This paper proposes a novel algorithm using clustering methods with the aim of estimating the future events locations. Simulation results illustrate that using methods like "Complete-Link" can be an efficient help in leading routing agents - in algorithms based on routing agents- toward new events in network. © Springer-Verlag Berlin Heidelberg 2010.