Spectrum Sharing between High Altitude Platform and Fixed Satellite Networks in the 50/40 GHz band

Spectrum Sharing between High Altitude Platform and Fixed Satellite Networks in the 50/40 GHz band Vasilis F. Milas, Demosthenes Vouyioukas and Prof. Philip Constantinou Mobile Radiocommunications Laboratory, National Technical University of Athens Heroon Politechniou 9 str., Zografou, 15773, Greece, Tel: +30-210-7724196 Email: vmilas@mobile.ntua.gr Abstract This paper addresses an in-depth analysis of the Earth-to-stratosphere co-channel interference produced by high altitude platform stations (HAPS) to the fixed satellite receivers and proposes a methodology for the evaluation of the current high altitude platform ground stations (HAPGS) power characteristics set by TU-R. Results of interference-to-noise ratio for various systems in realistic scenarios indicate that efficient use of the spectrum shared between HAPS and Fixed Satellite Networks is feasible, with a more careful consideration for specific latitudes.. NTRODUCTON High altitude platform systems have been recently proposed for the provision of fixed broadband services in the millimeterwave frequency bands [1], [2]. Flying in the stratosphere at altitudes between 15.5 and 30 Km [3] HAPS could operate as a standalone or complementary network to satellite based and terrestrial communication systems, combining advantages of both such as large coverage area, low propagation delay, broadband capability and clear line-of-sight signal paths offered by high elevation angles. However there are still some critical issues for this new technology, mainly related with spectrum sharing conditions with other co-primary services. The nternational Telecommunication Union has allocated a pair of 300 MHz spectrum for HAPS systems in the 48/47 GHz band shared on a non-harmful, non-protection basis with geostationary satellite and terrestrial services [2]. TU- R Resolution 122 highlights the need for studies for power limitations applicable for HAPGS to facilitate sharing with space stations receivers. This paper proposes a methodology for the evaluation of the co-channel interference-to-noise ratio from HAPS Earth-tostratosphere emissions to geostationary satellite receivers and evaluates the feasibility of harmonic spectrum sharing under the current allocated transmitting power characteristics of high altitude platform ground stations. The structure of the paper is as follows. Section studies the intereference propagation paths generated between the two networks in the 50/40 GHz band. Section contains the technical characteristics of the wireless communication network based on HAPS and the Fixed Satellite network operating in the V-band. n order to assess the feasibility of frequency sharing in the practical operational environments, a /N calculation model and two basic cases of study are proposed (section V). Theoretical and simulation results concerning co-channel interference from HAPS uplink emissions to satellite receivers are presented in Section V. Finally our conclusions and issues for further study are drawn in Section V.. NTERFERENCE PROPAGATON PATHS The interference propagation paths generated between ground, stratospheric and space stations are depicted in fig.1 and can be categorized as: A. nterference from HAPS to satellite receivers, B. nterference from HAPGS to satellite receivers, C. nterference from earth stations to HAPGS, D. nterference from earth stations to HAPS. Fig. 1. nterference Propagation Paths nterference propagation paths A, C and D are thoroughly studied in [4]. An issue requiring further study, is related with interference propagation path B and the power characteristics of HAPS ground stations. Up today [5], an off-set was considered between the footprint of the fixed satellite and the HAPS coverage area in order to share the same spectrum. n this paper we study a coordination methodology considering the cases where the satellite boresight points in the direction of the HAPS coverage area.. DESCRPTON OF THE NETWORKS A. Wireless Communication Network Based On HAPS The typical parameters of a wireless communication network using HAPS for the provision of fixed services in the

bands 47.2-47.5 GHz and 47.9-48.2 GHz are proposed in [2]. The system considered in our study comprises: 1) a stratospheric platform placed in a fixed position at heights from 15.5 Km to 30 Km, able to stay aloft for long periods of time (up to 6 months-unmanned aircraft, up to 5 years-airships) [1], [2], 2) antenna installed on the bottom of the HAPS transmitting multi-spot beams, providing broadband channels to ground stations within the HAPS visible range. This range varies for urban, suburban, and rural coverage areas, and is extended to a radius of 36 Km, 76.5 Km, and 203 Km away from the HAPS nadir point below the platform, respectively, 3) up to 2100 HAPS ground stations arranged symmetrically in HAPS urban, suburban and rural coverage areas. B. Fixed Satellite Networks n the last 7 years more than 200 satellite networks have been filed to the TU with uplink in the 47-50.2 GHz band and downlink in the 37.5-42.5 GHz band source: Space Radiocommunications Stations (SRS) database of TU. Our interference analysis is based on the following geostationary satellite networks [6] (Table ): GSOV-B1 GEO-SV GEOSAT-X TABLE TYPCAL TECHNCAL CHARACTERSTCS OF GEO SATELLTE SYSTEMS N THE V BAND PARAMETER GSOV-B1 GEO-SV GEOSAT-X Uplink Frequency (GHz) 47.2-50.2 Downlink Frequency (GHz) 37.5-40.5 39.5-42.5 39.5-42.5 Carrier Bandwidth (MHz) 123 199.85 11 Modulation QPSK DQPSK QPSK SATELLTE Radiation Pattern TU-R S.672 Receive Antenna Gain 51.5 (dbi) 58 (dbi) 53.8 (dbi) Noise Temperature (K) 650 1450 1100 Service Link Beams 24 40 48 V. ANALYSS The methodology described in this section provides a model for the evaluation of co-channel interference to noise ratio from the uplink (Earth-to-stratosphere) emissions of high altitude platform ground stations (HAPGS) to the geostationary satellite receivers. Co-channel interference arises because several users utilize the same frequency. nterference is due to the antenna radiation pattern sidelobes and the contribution from the main lobes can be meaningful in case of beam overlap. The high altitude platform system is based on multi-spot antenna beams and looks like a cellular system where each beam is a cell. Therefore, system design should be based on the consideration of interference other than thermal noise and fading effects. To characterize interference phenomena in such a complicated scenario the interference to noise ratio represents one of the most meaningful parameters. The proposed /N model is shown in (1), (2), (3). N = 10 log i {10 log (kt S BW S ) (db) (1) i=1 i = 10 P (H S)i 10 10 G (H S)i (ϕ HSi ) 10 10 G (S H)i (ϕ SHi ) 10 L (2) ( λ L = 4πR Sni ) 2 (3) where: n is the number of HAPGS, P (H S)i is the transmitting power of HAPGS (dbw), P (H S)i is the transmitting power of HAPGS (dbw), G (H S)i (ϕ HSi ) is the HAPGS antenna gain (dbi), G (S H)i (ϕ SHi ) is the satellite antenna gain (dbi), ϕ HSi is the off boresight angle of the HAPGS in the direction of the satellite receiver (degrees), ϕ SHi is the off boresight angle of the satellite in the direction of the HAPGS (degrees), R Sni is the length of the interference propagation path (m), k is the Boltzmann s constant (1.38 10 23 W/(Hz k)), T S is the satellite receiver s temperature (K) BW S is the satellite receiver s bandwidth (Hz). We consider the following two cases for our analysis: Case A, where the HAPS network is deployed near the equator, Case B, where the HAPS network is deployed near the poles. A. Case A - Deployment near the Equator The first case covers the intense interference received by the geostationary satellite from the uplink emissions of ground stations of a HAPS network which is deployed near the equator (coordinates: latitude 0, longitude 0 ). We consider that the satellite boresight points in the direction of the platform s nadir (fig. 2). For this particular interference geometry we assume that the angle ϕ HS of the satellite receiver in the direction

Fig. 2. nterference Geometry - Case A of the HAPGS is significantly small ( 0 ). Therefore G (S H)i (ϕ SHi ) = G (S H) where G (S H) is the maximum satellite antenna gain (Table ). The length of the interference propagation path between the geostationary satellite and the ground stations, R Sni, is equal to R geo = 35786Km. Assuming the same transmitting power, P (H S)i = P (H S), for all the ground stations, eq.(1) for Case A is simplified as: N = C A+10 log 10 G (H S)i(ϕ HSi ) log {T S BW S where C A is a constant. B. Case B - Deployment near the Poles i The second case covers the relatively weak interference received by the geostationary satellite from the uplink emissions of ground stations of a HAPS network which is deployed in high latitudes near the poles. The interference geometry of Case B is depicted in fig. 3. We consider that the satellite boresight points in the direction of the platform s nadir. The length of the interference propagation path between the geostationary satellite and the HAPGS, R Sni, is equal to R Sn = (R e + R geo ) 2 + R 2 e = 42634Km, where R e = 6370Km, and R geo = 35786Km. This covers the ground stations in the extremes of HAP s coverage area (points A, B) as well, since W is relatively small ( =1 Km, for min. elevation angle=10, 3.2 Km, for min. elevation angle=5 ).Assuming the same transmitting power, P (H S)i = P (H S), for all the ground stations, eq.(1) for Case B is simplified as: N = C B+10 log 10 G (H S)i(ϕ HSi ) log {T S BW S where C B is a constant. i (4) (5) Fig. 3. A. Total G (H S)i (ϕ HSi ) nterference Geometry - Case B V. ANALYSS RESULTS The radiation pattern of the HAPS ground station antenna is based on the following formula [7]: G max 2.5 10 ( 3 D λ ϕ) 2 for 0 < ϕ < ϕ m G G(ϕ) = 1 for ϕ m ϕ < 100 λ D 52 10 log D λ 25 log ϕ for 100 λ D ϕ < 48 10 10 log D λ for 48 ϕ 180 (6) where: G 1 = 2 + 15 log D λ (7) ϕ m = 20λ Gmax G 1 (8) D 20 log D λ G max 7.7 (9) ϕ = ϕ HSi, D is the antenna diameter (m), and G max is the antenna gain found in the second column of Table. Considering a 7:1 frequency reuse scheme, a fully loaded HAPS would be able to support 100 co-channel ground stations arranged symmetrically in each of the three coverage areas. Working on the specific geometry shown in fig. 4, the off boresight angle can be obtained by (10), (11): ϕ HSi = 90 ϑ i (CaseA) (10) ϕ HSi = cos 1 (cos(ϑ i ) cos(180 δ i )) (CaseB) (11) The total transmitting gain G (H S)i (ϕ HSi ) for the multiple ground stations of a HAPS at 21 Km, was calculated based on (6), (10), (11). The results are presented in Table.

TABLE V /N RESULTS - CASE B Service Transmitting - Case B Area Power (dbw) GSOV-B1 GEO-SV GEOSAT-X Urban -8-45.6-44.7-35.1 Suburban -7-41.5-40.6-31.0 Rural -1.5-32.4-31.5-21.9 Fig. 4. ϕ HSi geometry TABLE TOTAL G (H S)i (ϕ HSi ) Coverage Antenna Gain Coverage Radius Total Gain (dbi) Area (dbi) (Km) Case A Case B Urban 23 0-36 26.76 10.01 Suburban 38 36-76.5 14.95 13.08 Rural 38 76.5-203 14.95 16.62 The simulation results, which are presented in figs. 5, 6, 7, demonstrate that efficient use of the spectrum shared between HAPS and Fixed Satellite Networks is feasible but certain consideration is needed prior to the HAPS network deployment in suburban areas for latitudes from 55 to 65, and in rural areas for latitudes from 60 to 75. All the simulation trials were performed considering that the satellite boresight points in the direction of the HAPS coverage area, which represents the worst possible interference scenario. By comparison of the results of figs. 5, 6, 7, we conclude that the worst case of spectrum sharing is between rural HAPGS and GEOSAT-X satellite receiver. B. Theoretical and Simulation Results of /Ns Based on eq. (1), (2), (3) and Table the interference tonoise ratios for Cases A, B were calculated. n Tables, V the results for GEOSAT-X, GSOV-B1 and GEO-SV satellite networks are presented. TABLE /N RESULTS - CASE A Service Transmitting - Case A Area Power (dbw) GSOV-B1 GEO-SV GEOSAT-X Urban -8-27.2-26.3-16.7 Suburban -7-38.0-37.1-27.5 Rural -1.5-32.5-31.6-22.0.. Fig. 5. /N simulation results for urban HAPGS The results of Tables, V demonstrate that there are sufficient margins (from 6.7 to 28 db for Case A, from 11.9 to 35.6 db for Case B) for both cases with respect to the value of /N=10%. n order to evaluate the feasibility of harmonic spectrum sharing for the studied interference propagation path in realistic situations, we have performed various simulations. The study framework was the same with the theoretical analysis. The variable in our simulation was the deployment area (latitude of the nadir point) of the HAPS network. V. CONCLUSONS AND OPEN SSUES This paper has described a model for the evaluation of cochannel interference-to-noise ratio from the uplink (Earth-tostratosphere) emissions of high altitude platform ground stations to geostationary satellite receivers. The effect of different latitudinal deployment of HAPS network in /N values was studied, and the current transmitting power characteristics of HAPS ground stations proposed by TU-R, were evaluated in terms of co-channel interference produced towards satellite receivers. t was found that harmonic spectrum sharing

.. 40 Fig. 6. /N simulation results for suburban HAPGS [2] Recommendation TU-R F.1500-1, "Preferred characteristics of systems in the fixed service using high-altitude platforms operating in the bands 47.2-47.5 GHz and 47.9-48.2 GHz". [3] R. Miura, M.Suzuki, "Preliminary Flight Test Program on Telecom and Broadcasting Using High Altitude Platform Stations," Wireless Personal Communications, An nternational Journal, Kluwer Academic Publishers,vol.24, no.2, Jan. 2003, pp. 341-361. [4] V. F. Milas, P. Constantinou, "nterference Environment between High Altitude Platform Networks (HAPN) Geostationary (GEO) Satellite and Wireless Terrestrial Systems," Wireless Personal Communications. An nternational Journal, Kluwer Academic Publishers, Special ssue "High Altitude Platforms - Technologies and Applications," Vol. 32, pp. 257-274, February 2005. [5] Recommendation TU-R SF.1481-1 "Frequency sharing between systems in the fixed service using high-altitude platform stations and satellite systems in the geostationary orbit in the fixed-satellite service in the bands 47.2-47.5 and 47.9-48.2 GHz" [6] TU-R Recommendation, S.1328-6, "Satellite system characteristics to be considered in frequency sharing analyses within the fixed-satellite service" [7] Recommendation TU-R F.699-6 "Reference radiation patterns for lineof-sight radio-relay system antennas for use in coordination studies and interference assessment in the frequency range from 1 GHz to about 70 GHz".. 40 Fig. 7. /N simulation results for rural HAPGS conditions are guaranteed between a single HAPS network and a Fixed Satellite network. Further study is needed for the situations where multiple HAPS networks interfere with satellite receivers. ACKNOWLEDGMENTS This work was carried out with the support from the Operational Program for Educational and Vocational Training (EPEAEK ) and particularly the Program PYTHAGORAS. The project is co-funded by the European Social Fund (75%) and National Resources (25%). REFERENCES [1] J. Thornton, D. Grace, C. Spillard, T. Konefal, and T. C. Tozer, "Broadband communications from a high-altitude platform-the European Helinet Programme," EE Electron. Commun. Eng. J., vol. 13, no. 3, pp.138-144, June 2001.