Fixed Lag Smoothing Technique for Track Maintenance in Clutter
Rajib Chakravorty, Subhash Challa · 2005
A fixed lag smoothing algorithm for the multi-target scenario is proposed, taking the algorithm of the integrated probabilistic data association (IPDA) filter as the basis. Both target state and its existence are estimated in IPDA. An expression for the smoothing of track existence is derived. The proposed algorithm can be implemented as an extension to PDA smoothing. PDA smoothing computes the state estimate while the derived equation can be used to smooth the existence of the tracks. Thus, a recursive form for fixed lag smoothing of track existence is incorporated into PDA smoothing. This should improve the estimate of both the track states and existence probability which should give better track maintenance. In a practical application like "situation awareness", this improvement can have a great significance.