Antenna geometry optimization for 2D direction-of-arrival estimation for radar imaging

Abstract

This paper focuses on the optimization of the 2D geometry of sensor arrays for 2D direction-of-arrival (DOA) estimation. Such arrays can be used for radar imaging purposes. Due to the optimization, the number of array channels can be kept quite small, which reduces hardware costs, while highly accurate DOA estimation accuracy can be achieved. Therefore, we derive a very simple expression of the 2D Cramér-Rao bound (CRB). By minimizing the geometry dependent part of the CRB, we can reduce the fine error variance of DOA estimation. In addition, we define a modified beampattern for the single source case, which is valid for all physically possible DOA's. By varying its side lobe level, the probability of DOA outliers can be controlled. Furthermore, external conditions such as the range of possible DOA's and the DOA region of interest are also included in the optimization process to adjust the array to external requirements. By this means, optimum (single source) 2D DOA estimation performance can be achieved for a specific problem. As optimization algorithm, we use an evolution strategy. To show the improvement in DOA estimation accuracy of the optimized arrays, simulation results are presented and compared to standard 1D and 2D array geometries. For empirical validation, we have developed a 77 GHz prototype radar sensor with 16 RX channels. The optimized 2D array geometry is realized with microstrip patches. To estimate the distance and the relative velocity of targets, we apply frequency modulated continuous wave signal processing. In order to show the functionality of this radar imaging sensor, we present some measurement results.