Knowledge-Based Analysis of Directed Diffusion

Abstract

Message communication is the most battery consuming function of Wireless Sensor Networks. Therefore, protocols used in these networks must have little message exchange. This paper investigates the use of a knowledge based approach in analysing protocols and reducing their communication overhead. Such an approach enables us to examine what the nodes come to know through message exchange, and explore other ways of obtaining that knowledge. We apply this analysis to the popular directed diffusion protocol and show that it is possible to reduce message communication without compromising the protocol's performance.