Abstract
An important problem in a multiple radar multiple target tracking system (i.e., multiple radar fusion) is to reduce radar biases in order to determine whether the observations detected by several different radars represent the same aircraft or different aircraft. The purpose of this paper is to treat the problem of estimating three dimensional (i.e., range, elevation and azimuth) radar biases in a two-radar system using several aircraft that are detected by two radars simultaneously. A batch method using a linear least squares approach was considered for aligning two dimensional radars (i.e., range and azimuth). This method requires a large number of aircraft (100 or more). We present a recursive method using a linear least squares approach that can be considered as a six-order Kalman filter. The elements of state vector are the radar bias components in two radars. We also present that a condition for unique solution and an index of accurate radar bias estimation. This index is that the minimum singular value of an observation matrix H. Our computer simulation results show that the radar bias solutions are sensitive to radar and aircraft geometries. But our computer simulation results show the effectiveness of the proposed method, because we can select aircraft geometries using the minimum singular value analysis in order to obtain a good estimation. Our simulation results also show that our new approach can achieve satisfactory estimation accuracy with 2 aircraft aligning the three dimensional radars.
