Discrete weighted centroid localization (dWCL) : performance analysis and optimization

Abstract

In this paper, a discrete version of weighted centroid localization algorithm (dWCL) is proposed for a wireless-sensor-network-based indoor positioning system. Under this architecture, a wireless device can estimate its own position based on the location information of connected wireless sensor nodes in the network. Unlike other weighted centroid localization algorithms, dWCL classifies the sensor nodes into a number of groups and assigns them weights taken from a set of discrete values. With the weighting function and the range information decoupled, the dWCL framework is flexible to adjust its performance in accordance with the requirements. Its capability to generalize some existing weighted centroid localization algorithms is also demonstrated. In addition, the performance of dWCL is characterized by the error distribution in terms of accuracy and precision. Further enhancement is achieved by optimizing the system parameters. Finally, the simulation results are presented to validate the proposed scheme and analytical framework.