Extensions to the ONBU Algorithm for Increased Operational Relevance

We propose maturation and extension of an innovative batch algorithm for track scene history reconstruction, and track refinement conditioned on this history, given input source track data from one or more sensors. The envisioned method is designed to process both kinematic data (e.g., track reports) and feature/discrimination data (e.g., useful to distinguish objects seen by multiple sensors to support data association), utilizes additional sensor kinematic data during critical flight periods, and can perform time/space variant bias estimation to improve data association and track refinement. The proposed solution supports all phases of flights (endo- and exo-atmospheric, boost, midcourse, re-entry), extended threat dynamics (conventional and non-conventional), is robust to variable sensor biases, and provides anytime, warm-start optimization-based batch algorithms to provide efficient revision of past solutions based on new data in support of runtime constraints. Approved for Public Release | 18-MDA-9817 (23 Oct 18)