Error performance bounds for tracking a manoeuvring target

Abstract

The paper explores the subject of Cramer-Rao type lower bounds for tracking a manoeuvring target. Tar- get dynamics are modelled by the switching of multi- ple (possibly nonlinear) motion regimes. Two types of bounds are considered. The flrst is evaluated as the expected value over the regime sequences. The compu- tational complexity of this bound grows exponentially with time. The second bound is applicable only to the deterministic trajectories, and assumes that the knowl- edge of the particular regime sequence is known. Nu- merical examples are provided in support of theoretical considerations.