An efficient node deployment method for indoor passive localisation

Abstract

Most existing studies on device-free indoor localisation aim to improve localisation performance by deploying a large number of sensor nodes in indoor environments, thereby resulting in high hardware cost, more energy consumption and other drawbacks. This paper proposes a wrapper, heuristic and sensor-based node deployment method, which reduces the number of deployed sensors while achieving a high localisation accuracy by selecting the fittest link feature subset in localisation datasets. The performance results reveal that the proposed method outperforms existing work in terms of the number of deployed nodes, consequently can greatly thin out the extra nodes, shorten the localisation response time and save energy. Furthermore, the method can easily adjust the number of specific nodes by flipping the coefficient ratio of transmitters and receivers in the fitness function.