Localization of Multiple RF Sources Based on Bayesian Compressive Sensing Using a Limited Number of UAVs With Airborne RSS Sensor
Xinhua Jiang, Ning Li, Yan Guo, Dongping Yu, Sixing Yang
Abstract
Locating multiple Radio Frequency (RF) sources by Unmanned Aerial Vehicles (UAVs) using Received Signal Strength (RSS) measurements attracts extensive attention for its intrinsic simplicity in hardware and low cost. However, the signals from multiple RF sources are aggregated at the airborne sensors, making it hard to achieve multi-emitter localization with a limited number of UAVs. In this paper, the Bayesian compressive sensing (BCS) approach is applied to estimating ground targets’ locations exploiting the RSS data collected by the onboard sensors. Meanwhile, to support the BCS approach, which requires sufficient measurements to achieve multi-emitter localization, a sensing architecture comprising a limited number of UAVs is presented. Besides, to optimize the localization result, we put forward a trajectory planning-based Bayesian compressive sensing (TPBCS) algorithm. This algorithm dynamically plans UAVs’ trajectories based on the adaptive simulated annealing (ASA) algorithm, which can bring down the computational cost in UAV path planning compared to the traditional simulated annealing (SA) algorithm. The effectiveness and robustness of the approach are validated by simulations in the end.