Adaptive swarm intelligence routing algorithms for WSN in a changing environment

Swarm intelligent algorithms have been used to design distributed and fault tolerant routing protocols for Wireless Sensors Networks (WSN), able to self-adapt to environmental changes. The principle is that each sink emits a message with the highest pheromone intensity (with reference to ant colonies) and with a limited transmission range. Pheromone spreads to the sensors and at the same time is subject to evaporation, producing an intensity gradient that drives the construction of the routing tables. We have studied swarm intelligent algorithms resorting to an analytical technique based on Markovian Agents MA. In the present work, we show that the MA model can be experimentally validated through a real physical WSN. Moreover, we extend our previous research to the study of WSN in dynamically changing environments and we show how the pheromone gradient algorithm is a strong candidate for implementing WSN routing in very critical topologies.