On BMD Target Tracking: Data Association and Data Fusion

In this paper we consider multitarget tracking with multiple sensors for BMD. In a previous paper multitarget tracking with a single sensor was considered [8]. A ballistic missile may be in several pieces, presenting multiple targets. Besides the ground based or ship sensor there is also the missile seeker. We consider algorithms for generating and maintaining the tracks needed for BMD. A cue of a BM from a non-organic tracking system may also be received. We consider whether the cue is already in the local track file or is a new track. The cue information can improve the existing local track. INTRODUCTION BMD data association must process measurements from the Ground Based/Ship Radar (RF) and the Missile Seeker (IR). Data association algorithms for processing RF data may be adjusted to take RF and IR data, see [4] Section 9.6. BM queuing from a remote source, e.g. another ship, requires remote track to local track fusion. Data association algorithms for processing RF data may be used to associate the remote track with a local track in a multi-track environment. Once associated with a local track, filtering algorithms may be used for the fusion step. ____________________________________________________________________ 1. Data association refers the observation-to-track assignment and track maintenance functions (such as gating) in a multi-target tracking system, see [3] page 6, [4] page 8. 2. Track fusion refers to the two step process of first reconciling two tracks from two different sets of tracks as being of the same object and then combining them into one track, e.g., a remote track fused to a local track, see [3] page 442. The first step is also called track association.