Non-Common-Sight Localization Method for Moving Source with Imperfect Sensor Positions

Current localization studies require that sufficient sensors have common sight with the object and form the measurements for position estimation at the same time. However, the sensors may suffer challenges such as urgent multi-task concurrency and substantial interference, making it difficult to have enough sensors to observe the object at any time instant. To address this problem, this paper investigates the source positioning problem in non-common-sight localization environments, where only a few sensors have common sight with the unknown object and can form localization measurements at any time instant. Inspired by the idea of time-division systems, we first formulate the sequential time of arrival (TOA) measurement model using the nonlinear relationships between sensors and the unknown source at different time instants. Then, the Cramer-Rao lower bound (CRLB) is derived in detail and a computationally attractive estimator is developed to determine the object's position and velocity. Both mean square error (MSE) analysis and numerical simulations indicate the superior performance of the proposed method.