ADAF'TIIVE PROCESSING IN THE PRESENCE OF NEAR PIELD SCATTERERS

SUMMARY: In practical situations, it is necessary to have a flexible methodology for adaptive processing that can deal with the effects of mutual coupling between the various antenna elements, the coupling between the antennas and the platform and the effects of near-field scatterers. To merge the signal processing and the electromagnetics methodology with the goal to deploy it in a real system containing practical antenna elements, we propose to use a direct data domain approach for adaptive space-time processing which is quite amenable for real time processing [I, 2, 10, 111. In this approach one adaptively minimizes the interference power while maintaining the gain of the antenna array along the direction of the signal. Not having to estimate a covariance matrix leads to an enormous savings in memory and computer processor time and makes it possible to carry out an adaptive process in real time. The advantage of this new method is that we can do the adaptive processing with a single snapshot of data (a snapshot is defined as the voltages induced in the antenna elements at a particular instance of time). As there is no need to form a covariance matrix, real time implementation is possible and one can deal with both noncoherent and coherent (is., the signal and the multipaths are in phase with either additive or destructive) signals whereas a conventional statistical based approach can deal with only the noncoherent case unless additional processing is carried out. It has been shown that a direct data domain method has a lower Cramer-Rao bound for the estimate of the parameter of interest than a stochastic model based methodology, i.e., the variance of the solution is smaller in the presence of noise for a direct data domain method over the conventional statistical approaches [3]. Let us illustrate the problem through an example and this will also provide a cursory overview of the technique. Consider a semi-circular array consisting of 24 half wave dipoles of length L = 0.51 and radius r = 0.0051, where h is the wavelength of operation. The radius of the semicircular array is 3.821. The dipoles are z-directed. Each element is identically loaded at the center by 50a. In addition, there is a large structure located within a distance, which is 5 times the. radius of the semicircular array and is oriented located along the direction of - 20" off the broadside. The width of the structure is 7.641 and its height is 15.281. This is shown in Figure 1. Hence, the semicircular array and the scatterer have strong electromagnetic interactions in addition to the presence of mutual coupling between the elements. Our goal is to use this array to perform adaptive processing in the angular sector 4 spanning from - 60" to 60" off the broadside of the array. The magnitude of the signal of interest (SOI) arriving from 10" off the broadside is vaned from 1 V/m to 10.0 V/m in steps of 0.01 V/m while maintaining the amplitudes of the jammers IV/m coherent with the signal. The jammers are arriving from -20°, 40' and 50". The signal-tothermal noise ratio at each antenna element is set at 20 dB. The direct path of one of the interfering signal is blocked by the near field scatterer. We now take a single temporal snapshot of the voltages induced in the 24 antenna elements comprising the semi circular array. The