Atmospheric Radar Research Center

Investigation of Adaptive Array Processing

The concept of a multi-function phased array radar, which would be capable of accomplishing several disparate tasks, is not only interesting from a technological viewpoint but may prove to have significant economic advantages [Weber et al., 2005]. A multi-function radar would, by necessity, use phased array technology [Mailloux, 1994; Hansen, 1998] owing to the requirement of near-simultaneous aircraft surveillance and weather observations. Phased arrays, such as the SPY-1A used at the NWRT, perform beamforming on both transmit and receive, maximizing the antenna gain. Flexible beam steering would be accomplished through a relatively complicated beam control system. In contrast to the SPY-1A, preliminary designs for the MPAR system, as envisioned by MIT’s Lincoln Laboratory, would consist of numerous overlapping subarrays, each of which would be capable of real-time hardware-based beamforming. The set of time-series signals from all subarrays would then be processed in software in order to produce focused beams within the spoiled transmit beam of the radar. This software beamforming process is termed digital beamforming and has several advantages, such as resolution enhancement and clutter mitigation.

Research in Progress

For this phase of the project, we propose to study the MPAR design using realistic numerical simulations. Our focus will be on general validation of the design, beam sharpening using adaptive techniques, and clutter mitigation algorithms.

MPAR Design Validation Using Radar Simulator

The digital beamforming design of MPAR is innovative and would certainly be considered unconventional by the general meteorological community. Therefore, it is important to validate the concept using numerical simulations which are convincing to this community. Using the radar simulator, it will be possible to closely emulate the MPAR design and generate raw time-series data for each subarray. These data would be processed to estimate the standard spectrum moments, which would be statistically compared to the source meteorological fields from the ARPS data set. Leveraging on the aircraft tracking phase of the project, multi-function capabilities will also be simulated and tested.

Resolution Enhancement Via Adaptive Digital Beamforming

Although more computationally expensive, data-dependent (adaptive) beamforming methods have several ad- vantages. In particular, it is well-known that the Capon beamforming method [Capon, 1969] has superior angular resolution with moderate computational burden. This type of adaptive beamforming has been demonstrated with vertically pointing phased array radars [Palmer et al., 1998] and is applicable to the MPAR design given the availability of subarray time-series data. It is proposed to investigate adaptive beamforming from the perspective of resolution enhancement and to detail the possibility of reduced aperture size while maintaining the angular resolution design criterion.

Clutter Mitigation

For conventional weather radars, ground clutter is mitigated by the application of temporal digital filters. Phased array radar, in contrast, can use spatial filtering to selectively null clutter targets [Cheong et al., 2005a]. Normally, these spatial filters are implemented by a set of pre-generated weights for each of the array elements. Using these pre-generated weights, however, it is not possible to attenuate signals from non-stationary clutter targets (aircraft, biological) or from unknown ground clutter sources. Fortunately, the subarray design of MPAR would allow the application of adaptive clutter cancellation [Kamio et al., 2004]. For this case, a clutter signal is defined as any signal from an undesired direction, independent of whether it is stationary or non-stationary. This type of spatial filtering scheme will be investigated in the context of the MPAR configuration.
Back to Top of This Page
Next:  System Optimization - Waveform Design Using Pulse Compression