Course and miss-distance estimation for constant-heading targets with variable speed

Techniques for tracking targets moving with variable speed in a fixed direction fall into two categories: those that require prior knowledge of the target speed profile and those that do not. Both are maximum likelihood estimators, but they have differing sets of assumptions. Monte Carlo simulations are used to compare the accuracy of each method for a realistic tracking scenario. It is shown that, for miss-distance and heading estimation, it is often more accurate to make an approximate choice of speed profile than to make no choice at all.<
>