Distributed Sensor Positioning System Using Virtual Trajectories

Conventional two-step localization methods have the disadvantage of making a premature decision on an intermediate measurement, discarding useful information. By introducing the virtual trajectory, we propose a direct approach to the location discovery problem in this paper. This approach utilizes the Cascade Unscented Kalman Filter (CUKF) algorithm to discover the location-unaware sensor node's position, following this virtual trajectory. Thus, this method has the advantage of making a mature decision on an immediate measurement. In addition, this algorithm has several favorable features such as high scalability, robustness to measurement error, etc. Finally, Monte Carlo simulation experiment results show the effectiveness of this proposed method.