KEWLS and KFF 2D Comparative Model

With growing numbers of wirelessly connected devices and location based services, accurate and energy efficient localisation of dynamic targets is critical. However, sequential acquisition of range-based data degrades location estimation performance. Existing localisation schemes for sequentially asynchronous localisation commonly employ computationally exhaustive sequential nonlinear filters. This paper presents a novel solution for location estimation using sequentially acquired nonlinear data in multisensor systems. The proposed system considers a hybrid Kalman Filter (KF) and weighted multilateration approach in which several low-dimensional Linear KFs' (LKF) predict and synchronise range measurements to a single estimation instant. The Proposed Kalman Extrapolated Weighted Least Squares (KEWLS) solution is coupled with an additional KF (KKF) to improve tracking stability. Simulations compare the proposed solutions against existing sequential nonlinear filters under variable measurement noise, sampling latencies and nonlinear trajectories to highlight their suitability. Results indicate that the proposed KEWLS and KKF solutions are 193% and 84% faster, respectively, while exhibiting satisfactory results close to or exceeding the sequential nonlinear filters in various scenarios.