The Committee on Radio Astronomy Frequencies (CRAF) is a committee of
the European Science Foundation (ESF).
WARC-92 allocated the band 1610-1626.5 MHz on a primary basis to the Mobile Satellite Service (MSS) in the earth-to-space direction (uplink) and the band 1613.8-1626.5 MHz on a secondary basis to the MSS in the space-to-earth direction (downlink). The band 1610.6-1613.8 MHz is allocated to the Radio Astronomy Service (RAS) on a primary basis. The Radio Astronomy Service is protected by footnote 5.372 stating that "harmful interference shall not be caused to stations of the radioastronomy service using the band 1610.6-1613.8 MHz by stations of the radiodetermination service and mobile-satellite services (2904/S29.13 applies)". Currently, radio astronomy stations in 9 CRAF member countries use this band in Europe and, consequently, require protection from MSS.
The CEPT ERC considered that by adequate separation distances for the affected radio astronomy stations, this protection could be established. To this aim ERC SE project team PT17 developed a deterministic method to calculate separation distances for MSS Mobile Earth Stations (MESs) in the 1.6 GHz frequency range and after it became known that the MSS operators were not satisfied, the ERC SE project team PT28 studied the issue again and developed a statistical method to this aim using the Monte Carlo methodology.
This document gives guidance to representatives of European radio astronomy stations within the CRAF community in the discussions with their national administrations in the coordination process. This guidance is the CRAF position on the coordination with GLOBALSTAR MESs in the band 1610.6-1613.8 MHz. However, it must be noted that in all coordination issues the final responsibility is with the national regulatory administration and that CRAF can only make its opinion clear.
This document does not provide a detailed discussion of the different reports and recommendations used but the relevant conclusions and some additional evaluation.
The separation distances given in this document have been determined using the deterministic methodology of ERC Report 26. Also preliminary results are given for the Monte Carlo methodology. The calculations have been based on the current status of the GLOBALSTAR system and may be revisited when more traffic information comes available.
2.1. Deterministic method
The results of the compatibility study between the MSS MESs in the 1610-1626.5 MHz band and RAS in the 1610.6-1613.8 MHz band are given in ERC Report 26 (ITU-R Report 1126). The report gives results for the following propagation models:
For RAS it has been assumed that it is protected when according to ITU-R Recommendation RA769 the maximum acceptable level of interference at a radio astronomy observatory is -262 dBW/Hz in 20 kHz bandwidth. In addition, the separation distances are calculated assuming interference for 10% of the time. This percentage is stemming from ITU-R Report 696 and is also recommended in the draft recommendation of Study Group 7 about "the protection of the Radio Astronomy service in the frequency bands shared with other services".
With this method, the separation distance of a MES to a radio astronomy station is directly calculated.
When using this method, one must bear in mind that the concept of an acceptable data loss for radio astronomy of 10% is one of the calculation principles. This concept is at this moment not supported by ITU-R WP7D and its conclusions await finalization. It results partly from an erroneous interpretation of SE17 of ITU-R Recommendation RA 769 which mentions a data loss due to harmful interference but which is not a value of accepted data loss but an unavoidable loss due to interference because of variable tropospheric propagation conditions (see for further details the text of the recommendation).
Until now, usually the determination of separation distances is done by using deterministic methods and the application of such an approach requires some experience.
2.2. Statistical method
CEPT ERC SE project team PT28 completed a compatibility study between the MSS MESs in the 1610-1626.5 MHz band and RAS in the 1610.6-1613.8 MHz band using the Monte Carlo methodology. The results have been published in ITU-R Recommendation M 1316. The calculation method as developed by e.g. CNET in Paris uses ITU-R Recommendations P.452 and P.526 as propagation models. Other input parameters which also include a level of detrimental interference and a value of acceptable data loss for radio astronomy on the one hand and an MSS traffic model are specified by means of a separate input file.
It must be noted that the SE28 method has been developed for MES on land only and it does, therefore, not address the separation zone for a MES on a ship.
With this method the separation distance of a MES to a radio astronomy station cannot be calculated directly but is determined by an iterative process by several runs of the program and varying input parameters for each run to achieve ultimately the best or optimum solution. The result is sensitive for the number of iterations in a single program run, the percentage of acceptable interference (see section 2.1), the density distribution of the MESs and the minimum distance of the MESs.
It must be noted that until today no experiments are known which provide test data to check the experimental reliability of this statistical methodology. The software may work fine, but the validity of the underlying principles have not been tested yet.
Preliminary tests performed by CRAF in October 1999 show that after carefully using this method, the resulting separation distance of an MES to a radio astronomy station does not differ more than a few percent from the results obtained by the method described in section 2.1.
Since the RAS must not be used as the victim to test coordination methodologies and the experimental reliability of the calculated results, CRAF can support the statistical approach only after careful and adequate testing and support by conclusions from studies completed by ITU-R WP7D.
2.3. Warning
It was recently brought to the attention of the CRAF secretariat that the basic ITU-R recommendation for the determination of propagation parameters in coordination work is ITU-R Recommendation P 620. The ITU-R software to be used in this work is the program SCAT.EXE. The documentation accompanying this program (also in the code itself) explains that the propagation calculations are not applicable for antenna gains below 20 dB. When one uses values below this 20 dB, the program does not give any warning and the user will trust the results that in this case are uncertain. This limitation is due to the definition of the radiation patterns used, i.e. in the calculations a cone-approximation is used. Such an approximation is not applicable for low gain antennas. This background is not given in the text of the ITU-R recommendation!
Since this recommendation forms the basis for several other recommendations, such as the ones used in the methods briefly summarized in sections 2.1 and 2.2 a warning to the user who is determining separation distances may apply.
This implies that in all coordination discussions we must insist to revisit the conclusions when a better propagation model is available.
To determine the separation distances for GLOBALSTAR MESs ERC Report 26. The user can also calculate the separation distance by himself by using the facility on the CRAF website under http://www.astron.nl/craf/tools.htm - access via the CRAF homepage under 'useful formulas'.
3.1. experience with values for the Iridium system MESs
Using ERC Report 26 one calculates for the Iridium MESs operating in the band 1621.35-1626.5 MHz for which the out-of-band emission affects radio astronomy observations in the band 1610.6-1613.8 MHz a separation distance of about 35 km. This value has been introduced in German regulation (also considering the special case of the Effelsberg as an instrument in a valley). In the Netherlands, the separation distance will be 50 km because we also want to take into account the impact of an aggregate of systems and of the uncertainty with which the Iridium system determines the locations of the MESs, i.e. about 10 km mean error. Other countries may have similar values - however, we must note again, that the final result remains the responsibility of the administration: CRAF recommends 50 km in the Iridium case according to the considerations of the Dutch administration.
3.2. GLOBALSTAR MESs
Following ITU-R Report 1126, the tropospheric scatter model is used. It must be noted that SE17 seriously considered to recommend the extrapolated Okomura-Hata model because this gives somewhat larger separation distances and is therefore safer for radio astronomy. The separation distances for a density of MESs of 0.001545 per square kilometer are as follows for the tropospheric scatter model, i.e. ITU-R Recommendation P.620 forms the background:
| Over land: | elevation 0o: | 266 km |
| elevation 1o: | 158 km (Okomura-Hata gives: 226 km) | |
| Over sea: | elevation 1o: | 203 km |
It must be noted that these results depend in the characteristics of the MES terminals and their traffic density, which where not well known at the time ERC Report 26 was developed. Therefore, a check using the tools on the CRAF website is highly recommended.
3.3. separation distances for GLOBALSTAR MESs
The estimates for the separation distances are given in Table 1 for both the deterministic and the statistical methods mentioned in section 2. The input parameters to the calculations have been chosen to represent the MES situation realistically and to provide worst-case results for the radio astronomy station involved. In some cases specific terrain conditions differing from a flat country may require the addition of terrain specific parameters. The results given below are typical for a radio astronomy station surrounded by a flat terrain.
On the parameters the following comments must be made:
| Table 1 | |||
| Separation distances for GLOBALSTAR MES |
| Channel | Frequency | traffic density: 0.01 users/km2 | traffic density: 0.02 users/km2 | Monte Carlo Results |
| (ERC Report 26) | (ERC Report 26) | (ITU-R M.1316 | ||
| 1 | 1611.06 MHz | 425 km | 507 km | 420 km |
| 2 | 1612.30 MHz | 425 km | 507 km | 420 km |
| 3 | 1613.52 MHz | 425 km | 507 km | 420 km |
| 4 | 1614.76 MHz | 108 km | 139 km | 70 km |
| 5 | 1615.98 MHz | 81 km | 88 km | 30 km |
| 6 | 1617.22 MHz | 54 km | 59 km | 30 km |
| 7 | 1618.44 MHz | 54 km | 59 km | 30 km |
| 8 | 1619.68 MHz | 54 km | 59 km | 30 km |
| 9 | 1620.90 MHz | 54 km | 59 km | 30 km |
The results in the Monte Carlo column could need some little adjustment since the calculated difference between the received signal and the levels from ITU-R RA769 could be a little better adjusted, but this is at the level of a dB or a fraction of it. The calculations satisfy the condition that the received interfering signal must be lower than the ITU-R RA769 levels. The case of interference entering the main beam of a radio telescope has not been considered. But in any case, we can conclude that the results from the deterministic methodology and from the statistical approach are rather similar, especially when one considers the uncertainties of several parameters in the calculations based on the ITU-R Recommendations.
In some countries administrations ask radio astronomy stations to test whether the interference from an MES is observed considering some separation distance. At the time of writing this document, this process has been followed in one country by e.g. coordination of Iridium and GLOBALSTAR MESs.
Considering that an administration has the authority to ask for such tests, CRAF considers the test results not reliable for general coordination issues. Coordination is site specific and it is therefore wring to try to transfer the conclusions from the test results obtained for one site to another site even though there may be a natural inclination to do so. CRAF also want to point out that site specific coordination may easily develop into site specific coordination regulation with all its pitfalls of divide-and-rule arguments and erroneous comparisons between different radio astronomy stations.
CRAF's reservation on the applicability of experiments in the coordination process for one radio astronomy station is based on the following arguments:
In conclusion, CRAF recommends that for MESs operating in the band 1610.6-1613.8 MHz the separation distance to a radio astronomy station are as given in Table 1. These values have been calculated using ERC Report 26 and rounded to take into account possible effects due aggregates of systems and positional inaccuracies of the MES operating system. These number should not automatically lead to conclusions on separation distances since it must be considered that the accuracy of the position determination for an MES is about 10 km. Furthermore, when speaking of separation distances, the possibility of an aggregate of systems must not be excluded.
CRAF strongly discourages the use of experiments to produce results that could be applied in the regulatory process.
However, in the case that an administration demands so, experiments must be done at some radio astronomy station, CRAF recommends to explain to that administration its concern (section 4) and to insist to revisit the conclusions from the test results periodically to update the achieved agreement when necessary. As this experimental process has been chosen by the administration, any revisiting of the case must be accompanied by new experiments to check/confirm the previous one(s). This checking procedure must be included and explained in any unavoidable agreement. Furthermore, the conditions of the coordination agreement for a specific radio astronomy station must be made dependent on the demanded experimental checks.
Some administrations recommend or demand an agreement between a radio astronomy station and GLOBALSTAR on the issue of separation distances. The license to GLOBALSTAR will refer to that agreement. CRAF explained GLOBALSTAR that it does not accept to enter an agreement, since it confuses the coordination responsibility of the administration with private negotiations. Following Recommendation ITU-R M.1316, a radio astronomy station/CRAF and GLOBALSTAR must agree on the input parameters for the Monte Carlo methodology. At the moment of writing this document, this agreement does not exist. Therefore, the Monte Carlo results in this document must be considered as preliminary.
Given the development of the Iridium system, CRAF recommends that GLOBALSTAR opts to use the frequency band 1621.35-1626.5 MHz since CRAF expects that in the near future this band will become available for CDMA MES uplink applications. The GLOBALSTAR system has the ability to extend to the frequency bands 1621.35-1626.5 MHz.