Coverage hole detection in WSN with force-directed algorithm and transfer learning

Coverage hole detection is an important research problem in wireless sensor network community. However, distributed approaches proposed recent years for coverage have high computational complexity. In this paper, we propose a novel approach networks called FD-TL (Force-directed and Transfer-learning) which based on layout generation capability of Force-directed Algorithms image recognition power Convolutional Neural Network with transfer learning. contrast to existing approaches, the pure topology-based since can detect both triangular non-triangular holes from input topology without relying physical locations anchor nodes. FD-TL, Algorithm used generate series possible layouts given topology. Next, recognize potential generated layouts. During training phase, learning method aid process. Experimental results show that achieve 90% sensitivity 96% specificity networks.