1、 Rec. ITU-R S.1591 1 RECOMMENDATION ITU-R S.1591 Sharing of inter-satellite link bands around 23, 32.5 and 64.5 GHz between non-geostationary/geostationary inter-satellite links and geostationary/geostationary inter-satellite links (Question ITU-R 265/4) (2002) The ITU Radiocommunication Assembly, c
2、onsidering a) that inter-satellite links (ISLs) within geostationary (GSO) satellite systems use, or are planned to use, the inter-satellite service (ISS) frequency allocations 22.55-23.55 GHz, 24.45-24.75 GHz, 32.0-33.0 GHz and 59.3-71.0 GHz; b) that ISLs between GSO and non-GSO satellites use, or
3、are planned to use, the ISS frequency allocations 22.55-23.55 GHz, 24.45-24.75 GHz, 32.0-33.0 GHz and 59.3-71.0 GHz; c) that the ITU-R needs criteria and methods of calculation in order to assess the potential for ISLs of the type mentioned in considering b) to share frequencies with ISLs of the typ
4、e mentioned in considering a); d) that the criteria and calculation methods mentioned in considering c) could possibly enable the Radiocommunication Bureau (BR) to process notices submitted in accordance with Appendix 4 of the Radio Regulations (RR) for spectrum for ISLs of the type mentioned in con
5、sidering b); e) that the simulations described in Annex 1 show that instances of significant interference between co-frequency ISLs of the types mentioned in considerings a) and b) occur in only a small proportion of cases, and in those cases for only small percentages of the time; f) that Annex 1 a
6、lso verifies that the maximum levels of interference occur when the non-GSO/GSO ISL is instantaneously in the Equatorial plane, and that those levels may therefore be calculated manually; g) that considerings e) and f) make it convenient for frequency sharing between the two types of ISL to be facil
7、itated by coordination, and that coordination would not be needed in many cases; h) that the need to coordinate may be checked by calculating the minimum carrier-to-interference, C/I, ratios and comparing them with a simple threshold C/I, recommends 1 that frequency sharing between a non-GSO/GSO ISL
8、 and a GSO/GSO ISL is practical and in exceptional cases may require coordination; 2 that the need to coordinate should be determined by the method described in Annex 2; 2 Rec. ITU-R S.1591 3 that coordination should take place if the minimum C/I ratio in a bandwidth of 1 MHz at any of the four ISL
9、receivers, calculated by the method in Annex 2, is less than either the C/N identified for that receiver in item C.8 e) of RR Appendix 4 + 3 dB or, in the absence of a C/N, 30 dB; 4 that, in cases where coordination is found to be necessary from recommends 2 and 3, Annex 1 may be used by the coordin
10、ating parties for guidance. ANNEX 1 Investigation of interference between a non-GSO/GSO ISL and a GSO/GSO ISL 1 The feasibility of frequency sharing by ISLs In general, systems with ISLs in microwave bands have or are expected to have antenna diameters between 0.5 and 2 m. The lowest frequency ISS a
11、llocation is 22.55 to 23.55 GHz, and in this frequency range a 0.5 m antenna has a gain of about 39 dBi and a half-power beamwidth of about 2. For larger diameters and higher frequencies, the beamwidths can be a fraction of a degree. The angular range over which an ISL antenna may be pointed is larg
12、e, thus reducing the probability of conjunctions or near conjunctions with interfering ISLs. The number of systems using the ISS allocations in the year 2001 was small. It is anticipated that this number will grow, but will probably be limited by the introduction of optical communications in space,
13、and in any case it could not be expected to be as many as the number of systems with fixed-satellite service (FSS) links, which successfully share frequencies by means of coordination. Interference, to or from any ISL that has a non-GSO satellite, is dependent on time-varying geometry. Hence, instan
14、ces of significant interference will occur for only short durations in situations at or very close to the worst-case geometry. At all other times the antenna discrimination at one or both ends of the interference path will be substantial and this will reduce the interference power. 2 Selection of IS
15、L parameters for computation of interference statistics From a BR database compiled from filings for spectrum assignments, the transmission parameters of ISLs of both types in all three frequency bands associated with a number of MEASAT GSO and MEASAT low Earth orbit (LEO) satellites (Malaysia) were
16、 obtained. Also obtained from this source were parameters of 32 GHz ISLs associated with satellites in the LUXSAT series (Luxembourg), and of 67 GHz ISLs associated with satellites in the SMO-GEO series (France). Additionally, relevant information on GSO/GSO ISLs was extracted from Recommendations I
17、TU-R S.1151 (32.5 GHz), ITU-R S.1326 (51 GHz FSS band) and ITU-R S.1327 (62.2 GHz). Rec. ITU-R S.1591 3 At the time of writing the United States of America has several space systems with ISLs in the ISS allocations between 56 GHz and 65 GHz; they are USBL, USFD, USGAE, USGX, USLL, and MILSTAR. Some
18、of these systems are in the planning or design phase, others are under development, and others in operation. The orbits used by these systems include LEOs, high Earth orbits (HEOs) and the GSO. The orbit combinations for their ISLs include, inter alia, GSO to GSO, LEO to and from GSO, and HEO to and
19、 from GSO. Finally, reference was made to Recommendation ITU-R S.1328 for orbital and other pertinent data on non-GSO FSS systems employing on-board processing, which might include ISLs in the bands of interest. To keep the study within practicable bounds it was decided to review all of the availabl
20、e data and then compile sets of parameters for a LEO/GSO ISL, a medium Earth orbit (MEO)/GSO ISL and a GSO/GSO ISL, each of which is hypothetical but has characteristics which are typical of those in the filed systems. Four examples of the GSO/GSO ISL were included, and the parameters produced in th
21、is way are shown in Table 1. For the non-GSO end of the non-GSO/GSO link the nearest satellite handover strategy was modelled. The antenna pattern modelled in each case was as defined by Recommendation ITU-R S.672 (single feed, 25 dB first sidelobe). Each ISL was treated as an individual link (i.e.
22、no use was made of transmission gains coupling an ISL with a preceding or following link). The case of multiple satellites in an interfering system transmitting simultaneously toward the same satellite was not modelled; it is expected that this would lead to only slightly smaller minimum values of C
23、/I than those in the present exercise; the minimum C/I instances would occur more often, but in practical cases would still aggregate to small percentages of time. Since the e.i.r.p. required for a given ISL depends inter alia on its length, it was necessary to calculate the e.i.r.p. separately for
24、each link, and this was done assuming an operating C/N of 15 dB at the input to each receiver, as follows: e.i.r.p. 20 log (4 d/) + G 10 log (k T B) = 15 dB where: d : path length (m) (maximum in the non-GSO/GSO case) : wavelength (m) G : on-axis gain of the receiving antenna (dBi) k : Boltzmanns co
25、nstant (228.6 dB(W/(Hz .K) T : link noise temperature referred to receiver input (K) B : reference bandwidth (1 MHz). 4 Rec. ITU-R S.1591 TABLE 1 Parameters of typical inter-satellite links in ISS bands below 71 GHz ISL type LEO/GSO MEO/GSO GSO/GSO (A) GSO/GSO (B) GSO/GSO (C) GSO/GSO (D) Shape of or
26、bits Circular Height of orbits (km) 1 400 10 360 35 786 Number of orbit planes 7 (+GSO) 4 (+GSO) 1 Number of satellites per plane 9 (+1 in GSO) 6 (+1 in GSO) 2 Non-GSO inclination (degrees) 48 82.5 0 Ascending node displacement (degrees) 25.714 45 0 True anomaly displacement (degrees) 0 Not availabl
27、e Longitude(s) of GSO satellite(s) (E) 3 0 and 159.3 0 and 158.74 0 and 133.23 0 and 9.99 Adjacent plane phasing (degrees) 28.57 45 Not available Hand-over strategy Nearest LEO satellite Nearest MEO satellite Not available Satellite beam pointing method Tracking Fixed ISL antenna diameter (m) 1 ISL
28、antenna peak gain (dBi) 23 GHz 32.5 GHz 62.2 GHz 45.4 48.4 54.0 ISL antenna pattern Recommendation ITU-R S.672 LN= 25 dB ISL beamwidth (degrees) 23 GHz 32.5 GHz 62.2 GHz 0.91 0.65 0.34 ISL length (km) Variable, maximum 43 Variable, maximum 45 83 128 82 877 77 396 7 339 C/N at receiver input (dB) 15
29、e.i.r.p./MHz (dBW) 41.8 42.4 47.6 47.5 46.9 26.5 ISL system noise temperature (K) 700 Rec. ITU-R S.1591 5 3 Simulation model The geometrical details of the four GSO/GSO ISLs are given in Fig. 1. Although manual calculations of the maximum interference levels in both forward and return directions of
30、transmission were made, in order to check the results of those calculations by simulation it was only necessary to model one direction in the present case. The model was set up to simulate transmissions from satellite SG1to SG2and from SN1to SN2, and hence interference from SN1to SG2and from SG1to S
31、N2. Satellite SN1, not shown explicitly in Fig. 1, is in the non-GSO constellation and changes from time to time as determined by the tracking strategy. 1591-01SG1CSG1BSG1ASG1DSG2SN2DABCOFIGURE 1Example GSO/GSO ISLGSOradius42 162 kmMEO shellradius16 736 kmLEO shellradius7 776 kmi.e. GSO end ofnon-GS
32、O/GSOGSO/GSOISL(B)GSO/GSOISL(C)GSO/GSOISL(A)GSO/GSOISL(D)OA = 7 576 kmOB = 7 776 kmOC = 16 736 kmOD = 42 002 kmSG2OSG1A= 159.3SG2OSG1B= 158.7SG2OSG1C= 133.2SG2OSG1D= 10SN2OSG2= 36 Rec. ITU-R S.1591 For the main runs the geocentric separation between one end of each GSO/GSO link and the GSO end of ea
33、ch non-GSO/GSO link was 3, since this is the minimum spacing for most co-frequency, co-coverage GSO/FSS systems. Thus SG2was located at 0 and SN2at 3 W. For GSO/GSO ISL(A) the longitude of the satellite at the other end (SG1(A) was set at 159.3 E, which makes it the longest ISL which could possibly
34、create interference from SG1(A)to SN2; for a longer ISL the interference path would be blocked by the Earth. GSO/GSO ISL(B) was set up to be tangential to the orbit shell of the LEO constellation, and GSO/GSO ISL(C) to be tangential to the orbit shell of the MEO constellation. GSO/GSO ISL(D), a shor
35、t ISL spanning a longitude range of only 10 was included for the sake of completeness. This model enabled 48 sets of C/I statistics to be obtained in a single run, i.e.: a) LEO/GSO ISL interference to GSO/GSO ISL(A) in each of the three ISS bands, b) LEO/GSO ISL interference to GSO/GSO ISL(B) in eac
36、h of the three ISS bands, c) LEO/GSO ISL interference to GSO/GSO ISL(C) in each of the three ISS bands, d) LEO/GSO ISL interference to GSO/GSO ISL(D) in each of the three ISS bands, e) MEO/GSO ISL interference to GSO/GSO ISL(A) in each of the three ISS bands, f) MEO/GSO ISL interference to GSO/GSO I
37、SL(B) in each of the three ISS bands, g) MEO/GSO ISL interference to GSO/GSO ISL(C) in each of the three ISS bands, h) MEO/GSO ISL interference to GSO/GSO ISL(D) in each of the three ISS bands, i) GSO/GSO ISL(A) interference to LEO/GSO ISL in each of the three ISS bands, j) GSO/GSO ISL(B) interferen
38、ce to LEO/GSO ISL in each of the three ISS bands, k) GSO/GSO ISL(C) interference to LEO/GSO ISL in each of the three ISS bands, l) GSO/GSO ISL(D) interference to LEO/GSO ISL in each of the three ISS bands, m) GSO/GSO ISL(A) interference to MEO/GSO ISL in each of the three ISS bands, n) GSO/GSO ISL(B
39、) interference to MEO/GSO ISL in each of the three ISS bands, Rec. ITU-R S.1591 7 o) GSO/GSO ISL(C) interference to MEO/GSO ISL in each of the three ISS bands, and p) GSO/GSO ISL(D) interference to MEO/GSO ISL in each of the three ISS bands. Each run was continued until about six days of orbit time
40、had been simulated, using time steps of 1 s, thus computing about 6 24 60 60 = 518 400 samples of C/I for each of the 48 scenarios. This was sufficient to obtain a smooth cumulative distribution function (CDF) in each case. 4 Results The model was first run for the reference scenario, in which satel
41、lite SN2was located at 3 W i.e. 3 from satellite SG2, and the results were plotted as cumulative time distributions of C/I. As expected, the interference varied with time in every case which confirms the wisdom of basing frequency sharing on a short-term criterion. For most of the links even the sho
42、rt-term interference was negligible, and in no case did the C/I fall to a value anywhere near a likely interference threshold. The minimum C/I occurred for interference from the 23 GHz MEO/GSO ISL to GSO/GSO ISL(B) and was about 31.5 dB. As examples the results for this case and for the best case GS
43、O/GSO ISL(D) are shown in Figs. 2 and 3; the other plots (14 Figures each with three graphs) were prepared. 1591-0220 28 36 6844 9252 8460 10076101102103104110210C/I (dB)PercentageoftimeC/Iabscissavalue23 GHz32.5 GHz62.2 GHzFIGURE 2MEO/GSO ISL interference to GSO/GSO ISL(B)8 Rec. ITU-R S.1591 1591-0
44、390 92 94 10296 10898 106100 110104110210102101C/I (dB)PercentageoftimeC/Iabscissavalue23 GHz 32.5 GHz 62.2 GHzFIGURE 3GSO/GSO ISL(D) interference to LEO/GSO ISLNOTE 1 Figure 2 shows the lowest of the 48 sets of C/I statistics and Fig. 3 shows the highest. By carrying out further runs with satellite
45、 SN2in various longitudes relative to SG2it was confirmed that, as anticipated, the lowest C/I ratios occur when SN2is close to SG2. It was therefore decided to run the simulation for successively smaller SN2-SG2separations, in order to determine the circumstances under which frequency sharing might
46、 be difficult. The lowest C/I ratios thus obtained are listed in Table 2. These are all for 23 GHz; the ratios increase with increasing frequency in every case because of increasing antenna gain. TABLE 2 Lowest C/I ratios from simulations with SN2close to SG2 From Table 2 it is evident that, for co-
47、frequency operation of typical ISLs, satellite SN2could be safely located at longitudes with respect to SG2down to 3 without creating unacceptable interference peaks, and hence coordination would not be necessary if the geocentric angle between SN2and SG2was greater than 3. Angle between SN2and SG2
48、(degrees) Interfering ISL Victim ISL Lowest C/I (dB) 3 MEO/GSO GSO/GSO(B) 31.3 2 GSO/GSO(A) MEO/GSO 19.3 1.5 GSO/GSO(A) MEO/GSO 13.0 1.0 GSO/GSO(A) MEO/GSO 14.4 0.5 MEO/GSO GSO/GSO(A) 4.6 Rec. ITU-R S.1591 9 Finally, checks were made to ensure that the minimum values of C/I occur at some of the inst
49、ants when the non-GSO satellite in a non GSO/GSO ISL is passing through the Equatorial plane. This was done by carrying out a run to identify when the minimum C/I on a given link occurred, and then repeating that part of the run whilst plotting C/I against time. An example of this process is illustrated in Fig. 4, where both the upper and lower diagrams were printed from the simulation for the 23 GHz non-GSO/GSO ISLs interfering with GSO/GSO ISL(A) when SN2was separated from SG2by only 0.5. The lower diagram is a view from a po
