Sensor array calibration for uniform rectangular array in presence of mutual coupling and sensor gain-and-phase errors
来源期刊:中南大学学报(英文版)2014年第6期
论文作者:WANG Ding(王鼎) YAO Hui(姚晖) 吴瑛
文章页码:2228 - 2239
Key words:array calibration; uniform rectangular array (URA); mutual coupling; sensor gain-and-phase errors; closed-form solution; Newton-type iteration
Abstract: The sensor array calibration methods tailored to uniform rectangular array (URA) in the presence of mutual coupling and sensor gain-and-phase errors were addressed. First, the mutual coupling model of the URA was studied, and then a set of steering vectors corresponding to distinct locations were numerically computed with the help of several time-disjoint auxiliary sources with known directions. Then, the optimization modeling with respect to the array error matrix (defined by the product of mutual coupling matrix and sensor gain-and-phase errors matrix) was constructed. Two preferable algorithms (called algorithm I and algorithm II) were developed to minimize the cost function. In algorithm I, the array error matrix was regarded as a whole parameter to be estimated, and the exact solution was available. Compared to some existing algorithms with the similar computation framework, algorithm I can make full use of the potentially linear characteristics of URA’s error matrix, thus, the calibration precision was obviously enhanced. In algorithm II, the array error matrix was decomposed into two matrix parameters to be optimized. Compared to algorithm I, it can further decrease the number of unknowns and, thereby, yield better estimation accuracy. However, algorithm II was incapable of producing the closed-form solution and the iteration operation was unavoidable. Simulation results validate the excellent performances of the two novel algorithms compared to some existing calibration algorithms.
WANG Ding(王鼎), YAO Hui(姚晖), WU Ying(吴瑛)
(Institute of Information System Engineering, PLA Information Engineering University, Zhengzhou 450002, China)
Abstract:The sensor array calibration methods tailored to uniform rectangular array (URA) in the presence of mutual coupling and sensor gain-and-phase errors were addressed. First, the mutual coupling model of the URA was studied, and then a set of steering vectors corresponding to distinct locations were numerically computed with the help of several time-disjoint auxiliary sources with known directions. Then, the optimization modeling with respect to the array error matrix (defined by the product of mutual coupling matrix and sensor gain-and-phase errors matrix) was constructed. Two preferable algorithms (called algorithm I and algorithm II) were developed to minimize the cost function. In algorithm I, the array error matrix was regarded as a whole parameter to be estimated, and the exact solution was available. Compared to some existing algorithms with the similar computation framework, algorithm I can make full use of the potentially linear characteristics of URA’s error matrix, thus, the calibration precision was obviously enhanced. In algorithm II, the array error matrix was decomposed into two matrix parameters to be optimized. Compared to algorithm I, it can further decrease the number of unknowns and, thereby, yield better estimation accuracy. However, algorithm II was incapable of producing the closed-form solution and the iteration operation was unavoidable. Simulation results validate the excellent performances of the two novel algorithms compared to some existing calibration algorithms.
Key words:array calibration; uniform rectangular array (URA); mutual coupling; sensor gain-and-phase errors; closed-form solution; Newton-type iteration