Introducing an ACO Based Paradigm for Detecting Wildfires using Wireless Sensor Networks

In this paper, we introduce an ant colony optimization (ACO)-based paradigm for detecting wildfires using wireless sensor networks. By considering a data-centric approach and trying to monitor and report events from specific locations according to the requests of the sinks, energy is saved in terms of precious battery-power and additional communication overhead is reduced. The ant colony heuristic provides a robust mechanism to choose the most stable routing paths. Based on the probability of occurrence of the data sites according to the user queries, the movements of the ants are biased towards those sites. Zones of reachability are assigned in the network to better estimate a fire and also to reduce the overhead involved. Sensing of multiple phenomena is carried out by an odor-analysis approach. Local interactions are made use of here to globally optimize the tree generated. Ant-aggregation, of particular use here, reduces the data to be forwarded while sending various kinds of reports back to the sinks. Simulation is carried out in NS-2 and the results obtained are promising when a suitable zone radius is chosen.