Variance estimation for a radar tracking system

In this paper, an approach for the estimation of the variance Q and R of the process noise W(k) and the measurement noise V(k) for radar tracking systems is presented. In the proposed algorithm, the variance Q and R are expressed as a linear combination of the covariances of the third-order difference of the measurement data. The variances can then be calculated by simply solving the matrix equation. The two cases for the measurement noise being white and color are considered in this algorithm. Simulation results showed that for both cases the steady estimation is reached after 20 samples