Estimation and confidence in bearings only tracking

A novel pseudolinear estimator for bearings only tracking based on modified polar coordinates (MPC) is presented. It is approximately unbiased in a long-range scenario, contrary to the earlier well-known pseudolinear estimator in Cartesian coordinates. The description of the new estimator indicates that the log likelihood function will be approximately quadratic in MPC. Simulations show the performance of the novel estimator and its estimate of the parameter error variance. The performance of the novel estimator is close to that of the maximum likelihood estimator. It is shown that a confidence interval for the target course based on normality of errors in MPC is better than a direct Gaussian approximation of the target course estimation error