32nd Conference on Broadcast Meteorology/31st Conference on Radar Meteorology/Fifth Conference on Coastal Atmospheric and Oceanic Prediction and Processes

Sunday, 10 August 2003
Development of a digital IF radar transceiver system
Robert J. Watson, University of Bath, Bath, United Kingdom; and S. Talebi and P. S. Cannon
Poster PDF (352.2 kB)
Atmospheric radars (wind-profilers, weather and cloud radars) have traditionally employed conventional super-heterodyne analogue receivers. The receiver architecture typically consists of pre-amplification and filter stages, followed by a mixer stage to provide down-conversion to a lower intermediate frequency (IF) and finally detection to derive a baseband signal. The baseband signal can then be quantized and subsequently processed to provide estimates of meteorologically useful quantities such as rainfall rate, cloud liquid water and wind velocity.

The recent rapid growth in the telecommunications industry has lead to the development of high performance analogue-to-digital converters and digital signal processors such that it is possible to quantize the down-converted signal directly at IF. Using such techniques it is now possible to replace substantial portions of the analogue receiver system with digital signal processing (DSP). In doing so, many of the inherent alignment, maintenance and temperature sensitivity problems can be drastically reduced. In addition to solving many of the traditionally difficult problems associated with analogue receiver systems, a digital system provides an improved linearity, increased linear dynamic range and reduced phase-jitter through reduced AM/PM conversion inherent to IF limiting often used in linear receiver systems.

This paper describes the development of a digital IF transceiver system using two approaches; the first based upon a digital transceiver system originally developed by QinetiQ and the second based upon turn-key off-the-shelf components. Both systems utilize the same high bandwidth (400MHz) analogue to digital converter stages capable sampling rates of up to 105M samples-per-second with 14-bit resolution. In both cases the ADC and associated signal processing hardware (either 2 Analog Devices SHARC or 4 Analog Devices TigerSHARC processors) fits on a single PCI card for installation within a standard PC. The system developed to be used on the S-band Chilbolton Advanced Meteorological Radar (CAMRa) research radar operated by the Rutherford Appleton Laboratory in the UK. The paper will describe the differences in the two approaches, the design methodology and initial results. Finally conclusions based on the experiences and approaches will be presented.

Supplementary URL: