Airborne Radio Frequency Interference Studies at C-band Using a - - PowerPoint PPT Presentation

ElectroScience Lab

Department of Electrical Engineering/ElectroScience Laboratory The Ohio State University *NOAA Environmental Technology Laboratory +Department of Electrical and Computer Engineering Virginia Tech 23rd Sept 2004

Airborne Radio Frequency Interference Studies at C-band Using a Digital Receiver

IGARSS 2004: Frequency Allocations for Remote Sensing Joel T. Johnson, A. J. Gasiewski*, G. A. Hampson,

• S. W. Ellingson+, R. Krishnamachari, M. Klein*

ElectroScience Lab

Motivation

• As seen in this session, RFI at C-band is a major concern for

radiometry; AMSR-E, WindSAT showing data corruption

• Designing effective RFI mitigation strategies for C-band requires

detailed knowledge of RFI environment

• Some flexibility in choosing center frequency/bandwidth; can add-on

more rapid sampling and/or multiple analog sub-channels to traditional radiometer to improve RFI rejection as post-processing

• Use of digital receiver provides much more rapid sample rate and

larger number of channels; can implement real-time suppression

• GOAL: demonstration of interference suppressing radiometry at C-

band in airborne experiments using the prototype from the last talk

• GOAL: improve knowledge of RFI environment using prototype

ElectroScience Lab

Outline

• System design
• Experiment plan
• Deployment in SMEX04
• SMEX04 initial results

ElectroScience Lab

System Design

• The digital receiver prototype can sample a 100 MHz channel at IF

frequency 150 MHz

• Front end/downconversion for these experiments provided by the

PSR/CXI radiometer of NOAA/ETL – highly successful C-band radiometer deployed in many previous campaigns – excellent antenna and calibration systems – collaboration eliminates need for redundant development efforts

• “Spectrometer” modification to PSR/CX implements downconverter;

located inside PSR/CXI scanhead

• PSR/CXI includes both traditional direct detection analog sub-

channels, as well as tuned 10/100 MHz direct detection channels

• Use of both analog and digital receivers allows comparison of results
• Combined system: C-band Interference Suppressing Radiometer

(CISR)

Silver Spring, MD July 12, 2004 IPO Briefing - PSR and WB-57F

Polarimetric Scanning Radiometer (PSR) System

PSR/L:

1.4 GHz

PSR/S

V IR

18/21

50-57 37

89

118 183

340 380 424

PSR/A:

10.7, 18.7, 21.5, 37, 89 GHz Polarimetric

PSR/CX:

6-7.3, 10.6-10.8 GHz Polarimetric with Interference Mitigation

ElectroScience Lab

Basic PSR Spectrometer Schematic

• Antenna located in aircraft bomb-bay, steered in Az/El
• Spectrometer includes 10/100 MHz tuned analog channels
• Mixer rejects upper-side-band; LO tuned to allow observations from

5.5 to 7.7 GHz

ElectroScience Lab

Digital Receiver Module

• Includes additional channel separation stage: 100 MHz IF at

125 MHz to 2 50 MHz IF’s at 150 MHz

ElectroScience Lab

Digital Receiver Module

• Channel selection, PC-104 computer, digital receiver mounted in

cabin-rack box

• 120 GB internal storage -> data rate ~ 8 GB/hr to capture detailed RFI

information; USB 2.0 interface to external hard drives for archiving

ElectroScience Lab

Outline

• System design
• Experiment plan
• Deployment in SMEX04
• SMEX04 initial results

ElectroScience Lab

• PSR scanhead computer controls LO tuning, as well as internal

switches; digital receiver waits for LO settling before acquiring data

• Simple 1 bit interface: PSR “pulses” this bit when digital receiver

measurement should begin

• Digital receiver measures for ~ 20 msec, then observes noise diode for

~ 10 msec; “pulse” comes every ~38 msec

• Scanhead and digital receiver computers have internal clock systems,

synchronized through IRIG-B standard; should be within 1-2 msec

• Time synchronization allows later alignment of data in LO freq., etc.

Interface to PSR

23 msec 15 msec LO tune 20 msec 10 msec 38 msec PSR trigger pulse Digital receiver acq.

ElectroScience Lab

Experiment Plan

• Spectrometer/digital receiver observe 100 MHz channel, tuned through

22 states 5.5-7.7 GHz

• Each state = 38 msec; complete sweep of channels in ~ 836 msec
• PSR antenna rotates ~ every 3 secs, so we get 3.6 sweeps per

rotation

• Digital receiver operates in 5 modes:

– Direct capture of data: 1.31 msec, sampled every 10 nsec – Integration, blanker on : integration length is 1.31 msec – Integration, blanker off – Max hold, blanker on – Max hold, blanker off

• Duty cycle of capture is low due to large amount of data collected;

integration/max-hold operations approx 60% duty cycle neglecting tuning delays, etc.

• Digital receiver mode switched approximately every antenna rotation;

data stored in a large RAM buffer, no observation when writing out

ElectroScience Lab

Deployment in SMEX04

• The PSR/CX spectrometer system first deployment was the Soil

Moisture Experiment 2004

• SMEX04 flights occurred 13 days (54 flight hours) from August 4th-27th,

2004, including transit to/from Patuxent River Naval Base in Maryland

• SMEX04 flights were based out of Tucson, AZ, and included flights
• ver characterized soil moisture sites in AZ and northern Mexico

Navy P-3

ElectroScience Lab

Spectrometer/CISR Performance: SMEX04

• First flight of PSR/CXI + CISR: a few issues encountered

– Spectrometer LO feed through to direct detection channels

• Results in internal RFI to standard PSR/C channels
• Mostly eliminated by adding absorber inside scanhead
• Further mitigations planned post-SMEX04
• Consequence: Direct detection channels and

spectrometer/CISR could not be operated simultaneously for most of SMEX04 – Cabling issues into CISR

• IRIG-B synchronization lost
• Can still synchronize in post-processing, but more effort

required

• Stability of system impaired: failed to operate some flights
• Spectrometer/CISR data sets of interest acquired only on 8/14 (partial

flight), 8/24 (full flight), 8/26 (partial), and 8/27 transit to PAX (no CISR)

ElectroScience Lab

Sample Data: PSR H-Pol 400 MHz Channels, 8/26

Uncalibrated voltages versus time for four 400 MHz channel Each channel normalized by its mean This plot includes scan and flight location effects; no huge RFI effects seen in these wide channels Calibrated data shows low-level corruption Narrower channels should be more sensitive to in-band RFI….

Frequency (GHz) 6.0 6.5 6.92 7.32 0.5 1.5 67 min

ElectroScience Lab

Sample Data: PSR 10/100 MHz Log Channels

Frequency (GHz) Frequency (GHz) 0.98 1.02

100 MHz (swept) 10 MHz (swept)

67 min

ElectroScience Lab

Digital Receiver: H pol, Ch 12 (6.6-6.7 GHz)

24 min 2

ElectroScience Lab

Digital Receiver: H pol, Ch 18 (7.2-7.3 GHz)

24 min 2

ElectroScience Lab

Conclusions

• CISR/PSR spectrometer combination should provide useful

quantitative information on the effectiveness of alternative RFI mitigation strategies

• Detailed information on the C-band RFI environment should also be
• btained
• Further examination of SMEX04 data in progress; improvements to

combined system for future flights as well

• Next deployment: Oct 2004 in Antarctic campaign; RFI data to be
• bserved on transit flights