TY - JOUR
T1 - Direction-of-Arrival Estimation Using MODE with Interpolated Arrays
AU - Weiss, Anthony J.
AU - Friedlander, Benjamin
AU - Stoica, Petre
N1 - Funding Information:
I. INTRODUCTION Direction finding has been a topic of considerable interest in recent years. A large number of’ direction estimation algorithms have been proposed, having different characteristics. One of these techniques, the method of direction estimation (MODE) algorithm [2], is of special interest because of the fact that it is asymptotically efficient (it provides a large sample realization of the stochastic maximum likelihood method (MLM)), and it remains so in coherent scenes. i.e., when some or all of the sources are fully correlated. For uniform linear arrays (ULA), MODE can be implemented by means of a ximple two-step procedure. Its computational complexity in the ULA case is similar to that of root-MUSIC. However, its accuracy is superior to that of MUSIC. or MUSIC with spatial smoothing, if the sources are highly correlated or coherent. In fact, the performance of MODE is relatively insensitive to source correlation (it changes very little as the source? go from uncorrelated to coherent.) For more general types of array?, however, the implementation of MODE requires a multidimensional search in the DOA space. While this search can be organized computationally more efficiently than for the MLM, [3]. it is still not very attractive from a practical standpoint. In this paper we present a variation of the MODE algorithm which preserves the computational simplicity of the ULA-MODE, for arrays with arbitrary geometry. This is achieved by combining MODE with the interpolated array (IA) approach presented in [SI. Briefly stated, the IA approach is based on estimating the outputs of a “virtual array” from the outputs of the real array. The virtual array is often chosen to be ULA for convenience. The interpolated array outputs are obtained by a straightforward linear interpolation technique, with Manuscript received July 16. 1993; revised June 2, 1994. This work was wpported by the United States Army Research Office under Contract DAAL03-9 I -C-O022. sponsored by U.S. Army Communications Electronics Command. Center for Signals Warfare: and by the Swedish Research Council of Engineering Science\, under Contract 91-676. The associate editor coordinating thc review of this paper and approving it ior publication was Prof. Isabel Lourtie.
PY - 1995/1
Y1 - 1995/1
N2 - In this correspondence, we present a generalization of the MODE direction finding algorithm, using the interpolated array approach. The proposed algorithm is statistically efficient for uncorrelated, correlated, and even coherent sources, and requires only a modest amount of computations. The use of array interpolation makes it possible to easily incorporate calibration data into the estimation procedure. The performance of the algorithm was evaluated by computer simulation, verifying its statistical efficiency.
AB - In this correspondence, we present a generalization of the MODE direction finding algorithm, using the interpolated array approach. The proposed algorithm is statistically efficient for uncorrelated, correlated, and even coherent sources, and requires only a modest amount of computations. The use of array interpolation makes it possible to easily incorporate calibration data into the estimation procedure. The performance of the algorithm was evaluated by computer simulation, verifying its statistical efficiency.
UR - http://www.scopus.com/inward/record.url?scp=0029207448&partnerID=8YFLogxK
U2 - 10.1109/78.365310
DO - 10.1109/78.365310
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:0029207448
SN - 1053-587X
VL - 43
SP - 296
EP - 300
JO - IEEE Transactions on Signal Processing
JF - IEEE Transactions on Signal Processing
IS - 1
ER -