Friday, August 24, 2018   |   Scientific Development

The Baron Processor Suite, Part 1: Architecture

by Mrinal S. Balaji, Chief Scientist, Baron

Introduction

Weather radar signal and data processing has always played an integral role in identifying, classifying and tracking targets of interest while serving as the communication medium between the user and the various radar system components. Over the past years, the quantitative scale of signal processing feasible within acceptable cost, size and power constraints as well as the degree of sophistication and optimality of the algorithms now becoming possible has dramatically improved.

This article is the first of a two-part series and intends to concisely introduce the Baron suite of radar processors, while the second article will illustrate their design to address the diverse needs of weather application specialists by employing novel hybrid architectural concepts and effectively blending traditional signal processing techniques with innovative technologies like CLEAN-AP™ for superior data quality (CLEAN-AP™ © 2009 Board of Regents of the University of Oklahoma).

 

Opportunities, Ideas and Concepts

The ancient Greek philosopher Plato states in the dialogue Republic that necessity is the mother of invention and it aptly describes the reason for the genesis of the Baron suite of radar processors. In recent times, the needs of the weather community that utilize radar data have been rapidly evolving and the advent of Polarimetric capability within radars (Doviak et al. 1998) has served as an accelerant. Baron played a critical role in the design, implementation and the fielding process of the Polarimetric WSR-88D network (Saxion et al. 2012, Balaji et al. 2012) and realized that a severe need existed for a processor architecture that was capable of tuning itself to the target environment. The Baron Processor suite’s internal architecture is specifically designed to fill this gap within the data content ecosystem and meet the weather community’s diverse requirements to have access to radar data with better quality while having the tools and the ability to incorporate new algorithms and ideas into existing designs.

A historic lag has always persisted between radar algorithm design by the research community and the ability to make the technology operationally feasible by the manufacturing community. A major design goal of the Baron Processor suite is to reduce this lag by creating an algorithm implementation environment with the ability to bring the developed research concepts to the operational realm much quicker than has been in the past.

A significant responsibility of the weather data processors since their inception has been to be able to separate signals of interest from unwanted data. As Polarimetric diversity takes center stage within the operational radar systems, the techniques used for separating the proverbial wheat from the chaff introduce more multifaceted challenges. The Baron processor suite accomplishes this by utilizing cutting edge ground clutter suppression techniques like CLEAN-AP™ (Warde et al. 2009, Torres et al. 2014) designed to adapt itself to the constantly varying clutter target environment.

 

Baron Processors within the System Configuration

integrated systems diagram

Figure 1. Click to enlarge.

The new generation of Baron radars, named the Gen3 series, is equipped with a transmitter (magnetron or klystron), antenna and a receiver that houses the Intermediate Frequency Digitizer (IFD) in addition to Baron’s suite of radar processors consisting of the Radar Signal Processor (RSP), Radar Control Processor (RCP) and the Radar Product Generator (RPG). Figure 1 illustrates the process flow across the various sub-components within the Baron radar systems. The user interfaces through the RCP with the radar system and commands the system to perform a defined scan strategy for data collection. The RCP interfaces with the transmitter, receiver and the antenna/pedestal sub-assemblies transmit and receive target echoes, which are then processed within the RSP and the RCP to generate information regarding target location, strength and evolution characteristics with the RPG generating operational products.

 

Baron Processor Suite

The Baron Processor Suite comprises of the IFD, RSP, RCP and the RPG. The following sections will briefly describe the functionality of each of the components. The RSP, RCP and RPG’s architectural model supports a mixture of proprietary and custom application code within a Centos operating system.

 

  • Baron Intermediate Frequency Digitizer
Baron antenna-mounted radar electronics

Figure 2.

The receiver is mounted on the antenna in Baron Gen3 Radar systems (Figure 2) and provides low noise amplification, down-conversion while suppressing external interference and A/D conversion of the received radar echo. The last receiver stage includes Baron’s IFD that performs the digitization process at the IF frequency (Figure 3). Its proximity to the antenna ensures the system has superior sensitivity and doesn’t suffer from the VSWR issues within the rotary joint which often plague legacy weather radar systems introducing significant errors in base data measurements as the antenna rotates during operation. Built using FPGAs with embedded microprocessors that shrink the size, weight and power of the electronic components within, the Baron IFD features digitizer firmware that allows digitizing signals of extremely high rates at a resolution of 16-bits in addition to having the ability to implement real-time digital down-conversion and decimation on the digitized In-phase/Quadrature (I/Q) data.

 

Baron radar processor

Figure 3.

The digitized I/Q data from the IFD is used by the RSP to generate base data which are subsequently used to create radar products in the RPG. Multiple channels containing received echoes from targets of interest along different polarizations serve as the input for the RSP’s processing chain. A sample of the transmitted pulse is also provided to the RSP through a separate IFD output channel for power and phase correction to be performed on the received echoes’ I/Q data on a pulse by pulse basis. The RSP software runs on a COTS machine within a Centos operating system (Figure 4).

 

 

 

Baron signal processor

Figure 4.

The RSP software controls and configures the IFD appropriately for data collection, commands it to send out applicable signals to the transmitter in addition to performing Automatic Frequency Control (AFC) of the oscillator’s frequency when  a magnetron transmitter is used within the radar system. Additional abilities of the RSP include the ability to mitigate contamination due to spurts of radio frequency interference, ground clutter (using a variety of ground clutter filtering processes including CLEAN-AP™) and point clutter. The signal processor also bears the responsibility to determine the validity of return echoes using a combination of data thresholds for each base data parameter after extracting measurements of range, azimuth/elevation location, radial velocity, spectrum width and Polarimetric characteristics of the received echoes corresponding to the targets of interest using time and frequency domain processing.

 

 

  • Baron Radar Control Processor

The RCP serves as the primary machine through which the user interfaces with the radar. The RCP software runs on a COTS machine within a Centos operating system (Figure 5) and controls and monitors the status of all the radar system components including the transmitter, antenna/pedestal, the antenna mounted electronics (AME) that houses the receiver and the calibration modules, the RSP and the RPG while maintaining all the software licenses within the radar system. The software raises alarms in conditions where the monitored conditions cross specified limits in addition to being able to disseminate this information to the users through email.

Baron radar control processor

Figure 5.

The RCP manages when and what system calibrations need to be performed. Both online and offline calibrations are supported within the software. It also allows technicians and engineers to perform system maintenance by providing direct control utilities to the various system components like the transmitter, antenna/pedestal, AME, the RSP and the RPG.

To aid with weather radar data collection the RCP includes utilities to define various types of scan strategies including constant elevation based scans, constant azimuth based scans and point scans. The defined scan strategies can then be used to create weather specific Volume Coverage Patterns (VCPs). The user is provided with visualization tools to view the base data coming into and leaving the RCP.

A special processing algorithm commonly referred to as “split cut processing” within the weather radar community is supported within the RCP which works in conjunction with the RSP in performing range unfolding by placing valid dominant trip echoes in their appropriate location.

 

  • Baron Radar Product Generator
Baron radar product generator

Figure 6.

The RPG serves the role of generating operational products based on the user choices. The RPG software like the RSP and the RCP runs on a COTS machine within a Centos operating system (Figure 6). The software is designed to create radial by radial, scan and VCP based data products in addition to creating them every scan. Several derived products like rain rate, rain accumulation, de-aliased velocity, etc. are also supported.

 

 

 

Read Part 2 here.

 

References

  • Balaji M. S., Ellis J. R., Cartwright R. D., Romines J. H., Lee J. H., and Walker W., 2012: An Engineering Illustration of the Dual Polarization upgrade for the WSR-88D, Proceedings 7th European Conference on Radar in Meteorology and Hydrology: Advances in Radar Technology, Toulouse, France.
  • Doviak R. J., Zrnić D. S., 1998: Polarimetric Upgrades to Improve Rainfall Measurements. NOAA/NSSL Report, 120 pp.
  • Saxion D. S., Ice R. L., 2012: New Science for the WSR-88D: Status of the Dual Polarization Upgrade, 28thIIPS AMS, 14 pp.
  • Torres S.M. and Warde D. A., 2014: Ground clutter mitigation for weather radars using the autocorrelation spectral density, Atmos. Oceanic Technol., 31, 2049-2066.
  • Warde D. A. and Torres S. M., Automatic detection and removal of ground clutter contamination on weather radars, AMS 34th Conf. Radar Meteorology, Williamsburg, VA, USA, Oct. 5–9, 2009, paper P10.11.