A new multi-sensor fusion algorithm based on the Information Filter framework

Abstract

his paper presents a new efficient track-to-track fusion (T2TF) algorithm based on the Information Filter (IF) framework which takes into account phenomena encountered in practical applications. In fact, it combines a modified version of the IF framework and a self-tuning fusion procedure based on likelihood functions to address issues such as the correlation of the estimates, the transmission shortcomings and the measurement origin uncertainty. The proposed method is evaluated through simulations and the obtained results show that the proposed algorithm performs as well the optimal centralized fusion schema in an idealized environment while it exhibits better robustness capabilities than existing decentralized algorithms when observation origin uncertainty is considered. Moreover, its reduced complexity cost is suitable for real time applications.