Authors
Nate Golding, Zoya Popovic, and Laila Marzall
Abstract
Adaptive beamforming techniques are gaining popularity due to their superior ability to provide precise interference cancellation. As such, the phased array system seeks to give a demonstration of the capabilities of such interference cancellation at 7-8.4GHz. This work presents the digital component of the phased array demonstrator system, primarily handling the Analog Devices 16-channel Rx/TX Quad-MxFE data converter and associated Xilinx VCU118 FPGA board, shown below. The work includes programming adaptive beamforming as well as manual progressive phase shift beamforming. Additionally, this involves creating waveforms with modulations, including analog (FM, AM), PSKs, QAMs, and more complex protocols, including IEEE 802.11 and 3GPP LTE. A graphical user interface will be developed for easy access to beamforming and modulation controls. Using MATLAB tools, an initial beamforming simulation is presented that demonstrates the effectiveness of adaptive beamforming-based interference cancellation.
The phased array consists of an 8×2 uniform rectangular array for which the antennas have yet to be fully designed; therefore, cosine antenna elements are used. The radiation patterns of the array are shown in Figures 3 and 4. A series of 802.11a packets are incident upon the array at 65 degrees azimuth, 0 degrees elevation, with a small carrier frequency offset and flat fading channel with 5dB SNR. At the same time, a loud interfering source consisting of amplified white noise is incident upon the array at 10 degrees azimuth, 2 degrees elevation, with a power level 60dB above that of the 802.11a packet. A linearly constrained minimum variance (LCMV) adaptive beamforming algorithm is applied to the sequence, assuming that the angles of incidence of both the signal and interference are known. With the computed complex weights of the array applied, the resulting azimuthal radiation pattern is shown in Figure 5, demonstrating a null at 10 degrees and a distortionless (0 dB) response at 65 degrees. The resulting 802.11a snapshot was detected, demodulated, and decoded, with the data being recovered with a 0.00% bit error rate. The null in this case approaches infinite depth at exactly the frequency and angle of arrival of the simulated signal, however for a more realistic simulation, HFSS will be used to simulate the phased array system as well as the incident sources to obtain a signal which takes into account coupling between elements as well as the inexact nature of the angle of arrival, such that null depth and interference cancellation is more realistic.