Real-Time Narrowband and Wideband Beamforming Techniques for Fully-Digital RF Arrays

Abstract

Elemental digital beamforming offers increased flexibility for multi-function radio frequency (RF) systems supporting radar and communications applications. As fully digital arrays, components, and subsystems are becoming more affordable in the military and commercial industries, analog components such as phase shifters, filters, and mixers have begun to be replaced by digital circuits which presents efficiency challenges in power constrained scenarios.

Furthermore, multi-function radar and communications systems are exploiting the multiple simultaneous beam capability provided by digital at every element beamforming. Along with further increasing data samples rates and increasing instantaneous bandwidths (IBW), real time processing in the digital domain has become a challenge due to the amount of data produced and processed in current systems. These arrays generate hundreds of gigabits per second of data throughput or more which is costly to send off-chip to an adjunct processor fundamentally limiting the overall performance of an RF array system.

In this dissertation, digital filtering techniques and architectures are described which calibrate and beamform both narrowband and wideband RF arrays on receive. The techniques are shown to optimize one or many parameters of the digital transceiver system to improve the overall system efficiency. Digitally beamforming in the beamspace is shown to further increase the processing efficiency of an adaptive system compared to state of the art frequency domain approaches by minimizing major processing bottlenecks of generating adaptive filter coefficients. The techniques discussed are compared and contrasted across different hardware processor modules including field-programmable gate arrays (FPGAs), graphical processing units (GPUs), and central processing units (CPUs).