Simulation study of wideband interference rejection using STAP

Array antenna systems are often used to enhance the received signal to interference and noise ratio (SNIR) when the signal operates in a heavily jammed environment. An array antenna consists of many sensor elements; the array antenna output is the weighted sum of the sensor element outputs. This configuration is very effective in suppressing the tone interferences. However, if the interference waveform is wideband, the conventional array antenna cannot effectively suppress this wideband interference. To suppress wideband interferences, the single sensor element weight must be replaced by an adaptive FIR (finite impulse response) filter. This processing device is referred to as space-time adaptive processor (STAP). In addition to suppressing interference waveforms, another important design criterion of a reliable communication system is to preserve the desired signal by not introducing signal distortion. This goal can be achieved by imposing the constraint that the desired signal will be undistorted at the output of the STAP. This paper uses computer simulations to evaluate the performance of the STAP under wideband interference. Proper modeling of the received data at different antenna elements is important in evaluating the performance, especially when both the signal and interference have wide frequency bands. The received data can be regarded as a space-time sampled waveform with different sensor elements generating in the waveform in the space dimension. The time sequence data is referred to as the received waveform of a given antenna element. The antenna output is the linear combination of all antenna elements. In conventional narrowband beamforming, time sequences at different antenna elements are related by some fixed phase shift. The amount of phase shift is determined by the waveform's direction of arrival (DOA). For the wideband waveform, time sequences at different antenna elements are related by some time delay and not merely by phase shift. If a wideband waveform arrived in an arbitrary angle from the broadside, the time delays may not be the integer multiple of sampling time (fractional delay). Thus, to properly model the received waveform at each antenna element, interpolation is used to derive the waveform from the sample data. The interpolation function used in this simulation study is sin(x)/x, where the power spectral density within the signal frequency band is assumed to be uniform. Approximating the received waveform by using a finite number of time samples in the interpolation formula is equivalent to approximating in the least square error sense. If the power spectral density function is not uniform, then a different interpolation function is needed. The array antenna in this simulation study is assumed to have seven omni-directional elements placed on a plane arranged in a honeycomb manner. The spacing between the elements is assumed to be half a wavelength associated with the center frequency. Simulation results based on different processing algorithms (Applebaum, LMS, Sample Matrix Inversion [SMI]) are compared with the theoretical calculations. Different interference and noise scenarios are presented in this paper.