Termite-hill: From natural to artificial termites in sensor networks

Abstract

Termites present a very good natural metaphor to evolutionary computation. While each individuals computational power is small compared to more evolved species, it is the power of their colonies that inspires communication engineers. This paper presents a study of artificial termites in sensor networks for the purpose of solving its routing problem. The behaviors of each of the termites in their colony allow their simulation in a restricted environment. The simulating behavior demonstrates how the termites make use of an auto-catalytic behavior in order to collectively find a solution for a posed problem in reasonable time. The derived algorithm termed Termite-hill demonstrates the principle of termites behavior to routing problem solving in the real applications of sensor networks. The performance of the algorithm was tested on static and dynamic sink scenarios. The results as compared with other routing algorithms and with varying network density show that Termite-hill is scalable and improved on network energy consumption with a control over best-effort-service.