Fast beampattern evaluation by polynomial rooting

[摘要] Current automotive radar systems measure the distance, the relative velocityand the direction of objects in their environment. This information enablesthe car to support the driver.

The direction estimation capabilities of a sensor array depend on itsbeampattern. To find the array configuration leading to the best angleestimation by a global optimization algorithm, a huge amount of beampatternshave to be calculated to detect their maxima. In this paper, a novelalgorithm is proposed to find all maxima of an array's beampattern fast andreliably, leading to accelerated array optimizations. The algorithm works forarrays having the sensors on a uniformly spaced grid. We use a generalversion of the gcd (greatest common divisor) function in order to write theproblem as a polynomial. We differentiate and root the polynomial to get theextrema of the beampattern. In addition, we show a method to reduce thecomputational burden even more by decreasing the order of the polynomial.

[学科分类] 电子、光学、磁材料