1、STDOCEPT ERC REPORT 30-ENGL 1994 232b414 0035383 233 ERC REPORT 30 7 European Radiocommunications Committee (ERC) .- -. d within the European Conference of Postai and Telecommunications Administrations (CEF) FREQUENCY SHARING IMPLICATIONS OF FEEDER-LINKS FOR NON-GSOMSS NETWORKS IN FSS BANDS Brussels
2、, June 1994 _ STD.CEPT ERC REPORT 30-ENGL 1994 = 232b4L4 0015384 L?T D Copyright 1994 the European Conference of Postai and Telecommunications Administrations (CEW STD.CEPT ERC REPORT 30-ENGL 3994 232b424 O035385 OUb 9 ERC REPORT 30 Page 1 FREQUENCY SHARING IMPLICATIONS OF FEEDER-LINKS FOR NON-GSOMS
3、S NETWORKS IN FSS BANDS 1. INTRODUCTION 1.1 Relevant Radio redations A phrase in RR 22, which allows the Feeder-links of MSS networks to be implemented in FSS bands, reads as follows:- “the fixed satellite service may also include feeder-links for other space radiocommunication services”. The possib
4、ility of satellite networks based on non-geostationary orbits employing FSS bands is recognised in RR Article 11, but no coordination regime is currently prescribed for such networks. WARC-92 adopted a Resolution (COM5/8) as an interim measure to provide for the introduction of non-GSO networks in c
5、ertain frequency bands between 1 and 3 GHz, but the FSS bands are not covered by that Resolution at present. Protection for FSS networks using the GSO against interference from CO-frequency networks using non-geostationary orbits is currently provided by RR 2613, reproduced here for convenience:- ”N
6、on-geostationary space stations shall cease or reduce to a negligible level their emissions, and their associated earth stations shall not transmit to them, whenever there is insufficient angular separation between non-geostationary satellites and geostationary satellites resulting in unacceptable i
7、nterference to geostationary satellite space systems in the fixed satellite service operating in accordance with these Regulations.” 1.2 Bandwidth Reuuked A review of six non-GSOMSS system proposals revealed that the requirement for feeder-link bandwidth varies between 50 MHz and 200 MHz for the up-
8、paths, and the same (separately) for the down-paths. 2. TYPES OF NON-GSO NETWORK Table 1 summarises the types under consideration, in comparison with a geostationary network. Although no proposals for MSS networks using satellites in high apogee elliptical orbits (HEOs) were found, preliminary calcu
9、lations were made which satisfied the Team that, if HEOs were employed, interference to and from GSO networks wouid be unlikely to be a significant problem, because in-line situations (see below) would not occur at times when the HE0 lis were active. 3. THE PROBLEM 3.1 Iaterference to (and frm GSO/F
10、SS networks Figure 1 illustrates the case of an MSS satellite in a Polar (90 inched) Low Earth Orbit (LEO), and an FSS satellite in the GSO instantaneously passing through the LEO plane. Both earth stations are assumed to be within the coverage areas of both satellites. The top half of Figure 1 illu
11、strates an FSS earth station receiving from the geostationary satellite. If the LEO satellite is at position L(1) or L(3), interference from it enters via a sidelobe of the earth stations antenna pattern, but if it is in position L(2) the interference enters via the main beam of that antenna. The lo
12、wer half of Figure 1 shows an MSS feeder earth station tracking a LEO satellite. If that satellite is at position L(4) or L(6) when the GSO satellite crosses the LEO plane, then the feeder stations sidelobe emissions interfere with the GSO satellite, whereas if the position is L(5) the earh stations
13、 main beam will illuminate the GSO satellite. STDoCEPT ERC REPORT 30-ENGL 3994 2326434 001538b T42 9 ERC REPORT 30 Page 2 A typical FSS/GSO earth station antenna might have an on-axis gain of 5-60 dBi, and sidelobes conforming to the 29 - 25Log(Q) envelope defined in ITU-R Recommendation 580. Simila
14、r gains might apply to a typical MSS feeder station antenna. Thus the difference in gain, and hence interference level, between the circumstances of L(2) and those of Yi) and (L3) if the off-axis angles of the latter are 3“, for example, is 55 - 29 - 25Log(3) or about 38 dB. Similar reasoning applie
15、s to the difference in interference level between the L(5) situation and the L(4) or L(6) situations. It is therefore evident that, on both the uppath and the down-path, interference from non-GSOMSS feeder-links to GSOPSS networks will be charactensed by short bursts of very high interference (herea
16、fter tmed events for the sake of brevity), interspersed by much longer periods of moderate (very probably acceptable) interference. For convenience circumstances such as those identified by L(2) and L(5) are described as in-line instances. It is important to recognise that in-lie interference will O
17、CCUT in both directions, i.e. both MSS and FSS networks will suffer it (unless steps are taken to avoid it). This can be seen in Figure 1 by considering the GSO earth station to be transmitting, and the LEO earth station to be receiving. Calculations indicate that for the FSS/GSO networks the down-
18、path in-line interference will usually be more severe than its uppath counterpart, but for MSSILEO networks the uppath in-line interference will usually dominate. Figure 2 suggests that in-line events will be statistical in nature. If no preventative measures are taken, interference from LEO satelli
19、te L1 will be unacceptable if L1 passes through a cone of angle degrees subtended at GSO/FSS earth station E (note that Q was 3“ in the example given in the previous page). Since the Earth and the GSO satellite are rotating in the Equatorial plane at 0.25 dedrnin, whie the LEO satellite is orbiting
20、in an incliied plane at between 1 and 3.6 deg/min (depending on altitude) it is evident that on most of its revolutions L1 will miss the interference cone of earth station E, but every now-and-then a hit will occur. The duration of an individuai event will depend on whether L1 passes through the mid
21、die of the cone or nearer to one of the edges, and on the angular velocity of the satellite relative to the earth station. Clearly, other LEO satellites in the 58me piane will also hit the cone occasionally, as will satellites h other LEO planes. In general the hits will occur at random intervais, a
22、nd the number of down-path hits in a given period will be proportional to the number of satellites in the LEO constellation. Uppath hits (for interference to the GSO network) will also normally occur at random intervals, but the number in a period is influenced not only by the size of the constellat
23、ion but also by the strategy of handover for each feeder station from one LEO satellite to another. It k not possible to correctly compute the statistics for up-path hits unless the handover strategy is known and built into the computations. Careful consideration of Figure 2 suggests that, if the LE
24、O is one for which the ground track repeats exactly after a given amount of time, it can be arranged for a given satellite never to be in lie with particular earth stations and the GSO satellites to which they operate. In fact the Teams studies have shown that it is possible for a constellation of L
25、EO satellites to avoid in-line interference to and from Certain earth stations by careful choice of orbit configurations and gocd station-keeping. However, owing mainly to the adverse operational impact on the MSS feeder-links of having to operate with such constraints, and also to the fact that avo
26、iding hits for some earth stations in this way would considerably increase the number of hits in a period for other earth stations, the Team has concluded that avoidance of in-line events by orbit coherence is uniikely to be feasible in practice. In fact, in the Teams opinion, frequency sharing woul
27、d be assisted if non-coherence was deliberately ensured. 3.2 Interference to (and from) FS Terminals Figure 3 shows that, if both up and down-paths of the LEOMSS feeder-links share frequencies with the FS, the interference to (and from) the latter can involve both satellites and feeder stations. The
28、 following factors are evident:- * if the LEO satellite uses spot beams for its feeder-links then interference into the main beams of PS receiving terminais (example A) is likely to originate from the sidelobes of the satellite antenna. Interference from the main beam of the satellite will normally
29、enter the FS receivers via the sidelobes of the FS antennas (example B). STD-CEPT ERC REPORT 30-ENGL 1994 232b4B4 OOL5387 989 ERC REPORT 30 Page 3 0 Since the feeder station antenna will not normally operate at elevation angles below lo“, interference from it to FS receiving terminals will derive fr
30、om its sidelobes (example C). 0 The terrestrial terminals will be protected from interference from the satellite by the pfd limits recommended by the ITU-R for PS bands shared with the FSS. b The interference from the MSS feeder stations will be the subject of RR Appendix 28-type coordination. b Sin
31、ce the LEO satellites will move quite rapidly relative to the FS terminals, and the MSS feeder station antennas will track the satellites, the level of the interference will vary Widely with time y at least 14 dB in example C and 30 dB in example A). 0 Careful choice of the geographical locations of
32、 the MSS feeder stations will maximise the worst case off-axis angles ( #L) of the interference from MSS feeder stations. 4. CRITERIA FOR ACCEPTABILITY THRESHOLDS At present the interference criteria on which most PSS networks are designed are long-tem criteria (i.e. limits which apply for at least
33、80% of the time). Interference limits for digital signals are prescribed in ITIJ-R Recommendations 523 and 735, which both require single entry interference to be no greater than 6% of total noise in the receiver bandwidth, and aggregate interference not to exceed 20% of total noise, under clear sky
34、 conditions. These limits apply for the whole time the links are available, but Rec.579 recognises that equipment faults and severe propagation fades will occasionay mur, by setting unavailability lits of 0.2% of a year for equipment and 0.04% of any year for propagation. Rec.579 also regards outage
35、s of up to 10 seconds as available time. Owing to the severe but short-term nature of in-line interference events, there is a need for the ITU-R to establish criteria for both the maximum permissible level of in-line interference and for the level which may be regarded as an outage; maximum percenta
36、ges of thne for which they can be tolerated should be associated with these two levels. The Project Teams suggestions for digital carriers in GSOIPSS networks are:- Permissible limit - 12% of the clear-air long-term noise budget should not be exceeded for more than 0.1% (0.05%) of any year, and no i
37、ndividuai excess should last for more than 30 seconds. Outage threshold - 120% (64%) of the clear-air long-term noise budget should not be exceeded for more than 0.01% of any year, and no individual excess should last for more than 10 seconds. (The percentages in brackets relate to circuits designed
38、 to meet ITU-TS Recommendation G.826 while the un-bracketed percentages relate to circuits designed to meet IW-R Rec.614, which is compatible with ITU-TS Rec.G.821.) Except in cases where bit regeneration is performed within the satellite payload, these limits should be applied to GSOIPSS carriers f
39、rom end-to-end e. including both uppath and down-path interference contributions). It has not been practicable for the Project Team to develop in-line interference criteria for the non-GSOIMSS Feeder- links, but bearing in mind that Feeder-links are effectively trunk links it is conceivable that sim
40、ilar limits to those above (non-bracketed) would emerge from an appropriate study. ERC REPORT 30 Page 4 5. ANALYSES AND COMPUTER SIMULATIONS Using parameters of carriers in the IRIDNM, GLOBALSTAR and INMARSAT 21 (MEO) feeder-links for the non- GSO constellations described in Table 1, various members
41、 of the Project Team carried out either computer simulations or statistical analyses to deduce the ii-line event statistics for typical GSOIPSS networks sharing frequencies with the feeder-links of non-GSOMSS networks. The combinations considered covered LEO and ME0 constellations and ail three PSS
42、pairs of bands in common use. Carriers were selected from among the types most sensitive to interference, and those likely to cause the most interference, and results were obtained for both the GSOIPSS carriers and the non- GSOMSS carriers in each role. A small excerpt from the many results obtahed
43、is given in Table 2, in which 3 of the six carriers labelled “IG“ are typical INTELSAT GSOIPSS digital carriers and 3 are FM/TV carriers. The Sth to 8* columns in Table 2 show the important results; these were based on a 12% of total (long-term) noise threshold, and on the earth stations of both net
44、works being located in Equatorial regions and near to the longitude of the geostationary satellite. This excerpt is included here for illustrative purposes only; for the same carrier combinations the event durations and aggregate time percentages are greater by a factor of at least S for earth stati
45、ons in mid-European latitudes. The Teams conclusions were reached by considering the full range of results contained in the main report. 6. POSSIBLE SOLUTIONS 6.1 hlementation of RR 26 13 Rom the lower part of Figure 1 it can be seen that uppath ii-line interference could be avoided by switching off
46、 the transmission (and reception) of each MSS feeder station whenever its antenna is pointhg within degrees of the GSO, where * ii) exclusion of FSS carriers other than non-GSOIMSS Feeder-link carriers from designated sub-bmds within the FSS allocations; E) allocation of designated subbands within t
47、he FSS allocations to the non-GSOIMSS Feeder-links in reverse-band mode; the Feeder-link carriers would be permitted to operate only in reverse-band mode and other FSS carriers would, as at present, be restricted to normal mode. The Project Teams terms of reference exclude study of alternative (i).
48、A study of alternative (ii) is also outside the Teams competence, except to state the obvious that it would create major difficulties for operators of the displaced GSOIFSS networks, and that pressure to minimise the amount of spectrum involved and to avoid the heavily loaded parts of C and Ku-bands
49、 * would probably be strong. Alternative (iii) has been investigated by the Team in some depth, using both normal GSO/FSS carrier parameters and also using generalised parameters conforming to RR Appendix 30B. The modes of interference would be from non-GSO feeder-link earth station to GSO/FSS earth station (and vice versa), and from non-GSOIMSS satellite to GSOIFSS satellite (and vice versa). The study shows that the satellite-twatellite interference would be well within acceptable limits even in
