Fade Mitigation on Broadcast Satellite Links Using by ppp10127

VIEWS: 0 PAGES: 17

									  Fade Mitigation on Broadcast Satellite Links
Using Climatological and Meteorological Forecast
                      Data




                 Gerald B. Fitzgerald
                   gbf@mitre.org

                     29APR2002


                                        Organization: G036
                                        Project: 0702V250 1J
5/7/02                                                                              2




         Overview
         •   Problem definition
             • Compute Radio Frequency (RF) attenuation for weather-sensitive
                Satellite Communications (SATCOM) links, on a global basis, with
                emphasis on ocean areas
         •   Past approach
             • Design the SATCOM system to handle expected worst-case weather
             • Use climatology to estimate weather parameters of concern
             • Use empirical models, based on climatological data and long-
                baseline observations of signal strength, to model RF attenuation
         •   Current research
             • Operate the SATCOM system based on expected daily weather
                (mitigate rain “fades”)
             • Use Numerical Weather Prediction (NWP) and satellite-derived
                observational data to estimate weather parameters
             • Adjust attenuation models by assimilating recorded signal strength
                observations
5/7/02                                                                     3




         Current Approach: The System Design
         Phase
         • Some radio frequencies, particularly Extremely High Frequency
           (EHF), are very subject to attenuation by atmospheric gases,
           clouds, and rain
         • Attenuation may be approximated by a piecewise-linear
           weighted sum of functions of (in decreasing order of
           contribution):
            • Rainfall rate
            • Surface water vapor density
            • Vertically integrated cloud water content
         • SATCOM systems (transmitter powers, antenna diameters,
           receiver sensitivities) must be sized to accommodate a
           reasonable guess as to the worst weather through which they
           are expected to operate…
         • … So we employ climatological data to estimate the worst-case
           attenuation. (Surface air temperature, surface pressure and
           freezing height are also required by attenuation models.)
5/7/02                                                                                                                             4




         Current GDM* Model Hierarchy
                                                      Link Budget

                 Linear loss sum                        Weighted loss sum                                  G/T




                   Rain loss                                Cloud loss                        Gas loss
                                                                                             (ITU 676)
                   (DAH***)                                (ITU 840**)
                                                                                       O2 Loss            H20 vapor loss
                                 Worst-month
                                loss adjustment
                                                                                                 676 H2 O loss    L.K. H2 O loss
                                   (ITU 841)




                Rainfall rate                  Cloud H20 content            Air temp                 H20 vapor density
                 (ITU 837)                        (ITU 840)                 (COADS)                         (ITU 836)

           * Global Broadcast Service (GBS) Data Mapper
           **ITU: International Telecommunications Union ( ITU 840 == “ITU-R P.840” recommendation
           *** Dissanayoke, Alnutt & Haidara, ” IEEE Trans. On Antennas and Propagation, Vol. 45. No. 10,
           October 1997
5/7/02                                                                             5




         Data Sets: ESA Rainrate Grid
         •   From European Center for Medium-Range Weather Forecasting 15-year
             reanalysis effort
         •   A data model, as is its predecessor, ITU 837 and the Rainzone model
         •   Annual
         •   1 degree cell size
         •   Computed (variable) exceedances
5/7/02                                            6




         Data Sets: Surface Water Vapor Density
          •   Units: gm/m^3
          •   From ITU836
          •   For ITU 676, gas loss model
          •   Annual and seasonal
          •   1 degree cell size
          •   Mean supplied
          •   Variance rule supplied
5/7/02                                                7




         Data Sets: Total Cloud Water Content
          •   Units: kg/m^2
          •   From ITU 840-2
          •   For cloud loss model (also ITU 840-2)
          •   Annual only (assumed uniform)
          •   1.5 degree cell size
          •   12 exceedances supplied
5/7/02                                                              8




         Data Sets: Ocean Surface Temperature
           •   From NOAA Comprehensive Ocean-Air Data Set (COADS)
           •   For ITU-676 gas loss model
           •   5 degree cell size
           •   From 1992 monthly data
           •   Seasonal, annual averages
           •   Post-processed
5/7/02                                                                                      9




          Link Budget Outputs: NASA ACTS*
          99% Exceedence, Annual



                                                                           The ACTS
                                                                           Advanced
                                                                           Mobile
                                                                           Terminal
                                                                           (AMT) is
                                                                           usable where
                                                                           the color is
                                                                           green (good)
                                                                           through yellow
                                                                           (just barely).




         *Advanced Communications Technology Satellite, a Ka-band (EHF) SATCOM relay
5/7/02                                                        10




         Link Budget Outputs: NASA ACTS
         99% Exceedence, Worst-Month



                                          AMT would
                                          have trouble in
                                          worst-month
                                          conditions. This
                                          composite colors
                                          each grid cell by
                                          its own worst
                                          month, based on
                                          humidity - so
                                          this is not a map
                                          of any one
                                          month’s
                                          performance
5/7/02                                                                                     11




         Current Research: The Operations Phase

         •   A working SATCOM system delivers a specific, usable Bit Error Rate
             (BER - 10-5 or better), at a specific data rate. Lower data rates require
             less power.
         •   Because SATCOM systems are sized to deliver a usable BER in near
             worst-case weather, they have more than enough power to “burn
             through” more typical weather.
         •   This extra power, on a typical day, can be used to deliver a lower BER,
             or a higher data rate, or both. In the case of concern to us, if BER is
             fixed, data rate on a good day could be, for some locations, four or
             more times higher than the worst-case rate. It is inefficient to waste this
             “margin”. Note that the system is steered - it does need to handle the
             globally worst weather each day
         •   However, the process of adjusting the data rate - and the whole
             broadcast schedule that goes with it - takes time. A good estimate of
             the usable data rate needs to be made 6 to 12 hours ahead of time
5/7/02                                                                           12




         Research Problems
          • We would like to develop, 6 to 12 hours ahead of time, global
            maps of achievable data rates
          • Because we are concerned with broadcast (one-way) systems,
            we cannot know in real-time if the data rates selected turn out to
            be too high (and the resulting BER too high), so we need
            information on the variability in the weather fields used, to
            compute the variability in the resulting attenuation forecasts,
            and account for it by increasing the forecast attenuation beyond
            the expected value
          • But the field of most impact on us, rainfall rate, varies most
            rapidly in space and time
          • Rainrate fields available to us include both NWP data, and new
            data which uses long revisit-time microwave imagers to
            calibrate short-revisit time infrared imagery, giving derived (and
            delayed) observational data
5/7/02                                                                           13




         Research Problems (2)
          • The ITU attenuation models used, are tuned to produce good
            results with climatological data. What risks do we run feeding
            them with forecast data?
          • Weather radars offer insight into rain rate and cloud water
            content in real time. We have beacon receivers collocated with
            some radars, so we can begin to model attenuation at some grid
            points. How do we combine these real-time, tuned models, with
            the older models?
          • Bottom line question: Is there so much variability in the forecast
            fields we need that we’d be better off sticking with climatology
            for all of them? How about some of them (like rainrate)? Or can
            we combine NWP forecast rainrate data, quasi-observational
            rainrate data, and climatology, to get a better answer?
5/7/02                                                                                                14




         Research Hierarchy: Forecast Data
                                             Link Budget


                                           Weighted loss sum                          G/T




                    Rain loss                 Cloud loss                  Gas loss
                                                                         (ITU 676)
                     (DAH)                     (ITU 840)
                                                                   O2 Loss           H20 vapor loss


                                                                             676 H2O loss




                                                                           Air temp
                                                                       Surface Pressure
                Rainfall rate:
                                                                      Specific Humidity
                -NOGAPS*                   Cloud H20 content        Freezing Layer Height
           - Calibrated IR Imagery             NOGAPS                        NOGAPS

                      *Navy Operational Global Atmospheric Prediction System
5/7/02                                                                                                    15




         Research Data Source: NOGAPS
         Navy Operational Global Atmospheric Prediction System


         •Basic equation: Primitive equations with hydrostatic approximation

         •Integration domain: Global, surface to 1 mb

         •Horizontal resolution: T159 (~0.75 degree on the Gaussian grid). Output at 1 degree (81 km).

         •Vertical levels: 5 sigma levels below 850 mb, depending on terrain elevation

         •Forecast time: 144 hrs

         •Initial fields: Machenhauer initialization from the +/- 3 hour cut-off Optimum Interpolation
         Analysis

         •First-guess fields: Previous NOGAPS 6-h or 12-h forecast

         •Moisture physics: Convective precipitation (relaxed Arakawa-Schubert), shallow cumulus mixing
         (Tiedtke) and large-scale convection.

         •Ocean surface: Sea surface temperature from U. S. Navy's Ocean Thermal Interpolation System
         (OTIS 4.0) and ice coverage percentage from the NAVICE Center weekly analysis.
5/7/02                                                                                16




         References
          •   International Telecommunications Union (ITU), Radiocommunication
              Bureau, “Handbook of Radiometeorology”, Geneva, 1996
          •   ITU, “Handbook: Radiowave Propagation Information for Predictions for
              Earth-to-Space Path Communications,” Geneva, 1996
          •   ITU Report 563-4, “Radiometeorological Data”, Geneva, 1990, p. 123
          •   ITU Recommendation ITU-R P.618-5, “Propagation Data and Prediction
              Methods Required for the Design of Earth-Space Telecommunication
              Systems”, Geneva, 1997
          •   ITU Recommendation ITU-R P.676-3, “Attenuation by Atmospheric
              Gases”, Geneva, 1997
          •   ITU Recommendation ITU-R P.836-1, “Water Vapour: Surface Density and
              Total Columnar Content”, Geneva, 1997
          •   ITU Recommendation ITU-R P.840-2, “Attenuation Due To Clouds and
              Fog”, Geneva, 1997
5/7/02                                                                                   17




         References (2)
         •   ITU Recommendation ITU-R P.837-1, “Characteristics of Precipitation for
             Propagation Modeling”, Geneva, 1994
         •   ITU Recommendation ITU-R P.841, “Conversion of Annual Statistics to
             Worst-Month Statistics”, Geneva, 1992
         •   Dissanayake, A., Allnutt, J., and Haidara, H., “A Prediction Model that
             Combines Rain Attenuation and Other Propagation Impairments Along
             Earth-satellite Paths”, IEEE Trans. On Antennas and Propagation, Vol. 45.
             No. 10, October 1997
         •   Rice., P. and Holmberg, N., “Cumulative Time Statistics of Surface-Point
             Rainfall Rates”, IEEE Trans. On Communications, Vol. COM-21, No. 10,
             October 1973
         •   Fitzgerald, G. and Bostrom, G, “GBS Data Mapper: Modeling Worldwide
             Availability of Ka-Band Links Using ITU Weather Data”, 6th Ka-Band
             Utilization Conference, Cleveland OH, June 2000

								
To top