Track-to-Track Fusion Architectures – A Review∗

This paper presents a easy to read review of the architectures for track-to-track fusion (T2TF). Based on whether the fusion algorithm uses the track estimates from the previous fusion and the configuration of information feedback, T2TF is categorized into six configurations, namely, T2TF with no memory with no, partial and full information feedback, and T2TF with memory with no, partial and full information feedback. The exact algorithms of the above T2TF configurations and the impact of information feedback on fusion accuracy are reviewed. Although (under the Linear Gaussian assumption) the exact T2TF algorithms yield theoretically consistent fusion results, their major drawback is the need of the crosscovariances of the tracks to be fused, which drastically complicates their implementation. The information matrix fusion (IMF) is a special case of T2TF with memory. Although it is heuristic when not conducted at full rate, it was shown to have consistent and near optimal fusion performance for practical tracking scenarios. Due to its simplicity, it is a good candidate for practical tracking systems. For the problem of asynchronous T2TF (AT2TF), a generalized version of the IMF is presented. It supports information feedback for AT2TF in the presence of communication delay, and was shown to have good consistency and close to optimal fusion accuracy. Finally the fusion of heterogenous tracks is discussed. For the problem of the fusion of the track from an Interacting Multiple Model (IMM) estimator from an active sensor with the track from a passive sensor, a counterintuitive phenomenon that heterogenous T2TF may have better performance than the centralized measurement-to-track fusion approach (which is the known optimum in the linear case) is demonstrated and explained.