Characterization of hello message exchange for estimating distribution of network residual energy
Shudong Fang, Stevan Mirko Berber, Akshya Kumar Swain · 2009
This paper investigates the practicability that a sensor node develops the probability density function (pdf) of its local network energy via exchanging hello messages with its neighboring nodes in the context of dense node deployment. The pdf is proven to approach Gaussian and can be used to decentralize a recent clustering algorithm. To alleviate the broadcast storm problem, a node is considered broadcasting hello messages to its neighboring nodes without immediate feedback from the receiving ones. Thus the broadcasting node cannot be guaranteed that its neighboring nodes have received its messages which are at high risk of channel collision. Characterizing hello message exchange becomes nontrivial, as the discovery ratio, which measures the effectiveness and the sufficiency of the message exchange, is identified having decisive effect on the precision of the developed pdf. A set of time asynchronous and slot-based channel access rules is presented for the sufficient and fast exchange of hello messages.