Energy-efficient landmark tracing in WSNs using random walks on network graphs

Detection of landmarks in mobile sink environment which is generally used to improve the network lifetime is an important problem in WSN. In this paper, a method is proposed to trace the landmarks in a network leading to energy-efficient networks. The algorithm for clustering and landmarks tracing proposed herein uses the concept of random walks on graphs. The proposed method comprises of two phases namely, flow escalation and flow control and it does not require location information of the nodes. Additionally, prior information about number of optimal clusters is also not required. Experimental simulations illustrate the process of landmark detection along with clustering. Results obtained from analysis of WSN lifetime and number of data packets transmitted are motivating.