DHARMA: Digital Hardware Architectures for RF Multi-Dimensional Arrays
Introduction
An aperture array has antennas that can be deployed in particular geometric patterns to detect radio signals at a given range of frequencies. Using an array is beneficial for electrical engineering because it can magnify a radio signal in the direction of that signal while suppressing noise and interference through a process known as beamforming (see Fig. 1). Lower levels of noise and interference leads to better detectability in radar applications, more reliable and robust communications, and clearer pictures, for imaging. In addition, an antenna array can be used to detect the directionand distance of the signal’s source. This is known as source localization. Localization is used for applications in position, mapping and surveillance. Aperture arrays are a crucial component of scientific instruments that measure the spatial distribution of radio sources.
At the Advanced Signal Processing Circuits (ASPC) Lab my students and I investigate aperture array signal processing algorithms, circuits and electronics based on multi-dimensional (MD) systems and signal processing for applications in microwave and millimeter-wave radio-frequency (RF) antenna array processing.
Fig. 1. Mathematically computed far-field beam patterns from 3-D cone-filterbank NRPA superimposed from artist’s impression of a Square Kilometer Array (SKA) aperture array with station-level beamforming.
The main idea of our projects funded by the Office of Naval Research (ONR) is the directional enhancement of propagating electromagnetic planar waves based on their directions of arrival. A particularly relevant example for such electromagnetic “beamforming” applications is in active electronically-scanned array receivers, which are part of advanced radar sensors used by the Department of Defense (DoD). Our proposed MD theory allows new beamformers to be realized using novel classes of array processing algorithms, circuits and systems using both analog and digital implementation platforms, that may outperform traditional phased-array beamformers that are extensively deployed in DoD applications, such as radar and wireless communications.
Imaging Applications for Radio Astronomy
Our proposed beamforming techniques show improved steerability, directivity and bandwidth compared to traditional phased-array based systems. Our work has led to theoretical developments in the field of adaptive algorithms that outperform available beamformers using the properties of MD algorithms. We have applied such MD beamforming developments for defense applications. Our work will lead to improved beamformers in the presence of excessive interference, noise and jamming. Furthermore, we investigate reconfigurable array signal processors, circuits, and systems for use in array processor-based antenna apertures for wireless communications, localization, and sensor systems. An example is the use of apertures for detection of airborne threats having low radar visibility. The research includes the investigation of both MD signal processing theory and circuit theory as well as explores practical aspects pertaining to analog, digital and mixed-signal VLSI implementation at microwave and millimeter wave frequencies.
Fig. 2 Closely-packed hexagonal beams for multiple sky visibilities. These hexagonal beams provide optimally sampled radio pixels that have inter penetration curves that occur at -3 dB level in the magnitude response of the array patterns.
Fig. 1 shows how our research has a wide impact on several classes of beamforming applications, starting with tile and statio-level RF multiple visibilities in advanced radio telescope systems such as the square kilometer array (SKA), as shown in the illustration in Fig. 1. Here, the mathematically accurate beamshape from our MD beamformer is superimposed on the artist impression of an SKA beamformer station, to show how the beamformers can be used in emerging high-performance instrumentation. Fig. 2 shows an optimal sky sampling pattern using a filterbank of RF beams having hexagonal cross sections, such that optimal pixel packing is achieved for space imaging applications. Radio telescopes, such as the SKA instrument, rely on aperture arrays to generate precise radio images of electromagnetic sources for experimental cosmology and space science. In order to build aperture beamformers to be used for a specific purpose, such as airborne sensing and signal intelligence, efficient schemes for processing the antenna array signals must be developed to reduce the computing time, energy consumption, and costs for the hardware required in the sensing system. MD beamformers can be deployed in aerial vehicles and other cyberphysical systems for advanced sensing, imaging, communications and signal intelligence.
Network Resonant Phased-Array (NRPA) Beamformers
Our research on microwave and millimeter-wave analog/digital MD systems for antenna array processing including aspects of electronic implementation is currently supported by the National Science Foundation (NSF). The main idea of this NSF project is to investigate MD filter theory from the standpoint of beamforming, together with sparse sampling theory, in order to realize aperture arrays for a wide variety of important applications. The proposed research creates new algorithms and digital computing architectures that will produce highly-focused hexagonal radio pixels for the most demanding of microwave imaging applications. The same aperture arrays are used in radar and wireless communication systems for signature detection and signal intelligence. Our research tackles the problem of highly directional sparse aperture arrays using the mathematical properties of multi-dimensional recursive digital filters.
Fig. 3 The proposed network-resonant phased-arrays (NRPAs) can be used as a pre-filter to improve the directivity of traditional phased-array beamformers.
The NSF-sponsored effort will develop hardware systems for aperture arrays based on the proposed concept of network-resonant phased-arrays (NRPAs). MD circuit theory and digital hardware form an enabling technology for imaging algorithms that can greatly improve performance over traditional technologies. This research proposes groundbreaking techniques based on array signal processing, circuits and systems. It will result in a significant improvement in the directional sensitivity while using a lower number of array elements compared to traditional phased array receivers of the same sensitivity. The proposed NRPAs combine the concept of network resonance with phased array technology to gain significant improvement in both directionality and sensitivity. The MD circuit theoretical concept of network resonance allows digital beamformers to have complex pole manifolds. These properties are shown to have advantages in terms of ultra-wideband frequency response, exceptional directionality, multi-beams with shape control, rapid steerability, and low computational complexity (see Fig. 3). This project investigates radio beams with a hexagonal sky-print for optimal sensing and microwave imaging over wide fields-of-view and bandwidths. The proposed NRPAs will be extended to both sparse and random arrays via theoretical formulations for decreasing hardware cost, reducing energy expended in computers and increasing design flexibility. The research is expected to impact high-performance wideband steerable antenna aperture array applications having multiple end uses, including but not limited to wireless communications, cognitive radio, radar, microwave imaging, localization, space science, radio astronomy and signal intelligence.