Compressive sensing with weighted coefficient approach for indoor source localization

Abstract

In this work, indoor sensor localization is investigated based on the received-signal-strength-indicator (RSSI) using the compressive sensing (CS) framework. The fingerprints of several ZigBee base-stations are used to construct the dictionary matrix. We propose a weighted-coefficient-approach (WCA) to determine the sensor position based on the solution of the CS problem. Numerical simulations and experiments are conducted to show that the proposed WCA outperforms the conventional CS-based approach for source localization. Numerical results demonstrate that the WCA can be used to determine the sensor position with high accuracy.