Deforestation detection algorithm for wireless sensor networks

One of the contributing factors to global warming problem is the deforestation. Tools mostly used in this vast destruction of forest are chainsaw. Chainsaw has a quite unique acoustic characteristic that can be detected with high accuracy. To prevent and deter the deforestation, wireless sensor nodes can be deployed to detect the activity of chainsawing. However, to the best knowledge of author, there is no paper addressing this issue. In this paper, we present the algorithm suitable for limited energy sensor node to detect chainsaw. Aiming for low complexity and workable in real world, the algorithm combines three simple techniques: (1) adaptive energy threshold, (2) delta pitch detection and (3) energy band ratio. Performance results show that this algorithm is able to achieve detection accuracy about 90%.