Study of a Satellite Multimedia Broadcasting Mobile System Mission for the Platform

Transcription

1 Study of a Satellite Multimedia Broadcasting Mobile System Mission for the Platform Piero Angeletti, Riccardo De Gaudenzi, Rita Rinaldo ESA-ESTEC, Keplerlaan 1, 2200 AG Noordwijk, The Netherlands Tel: , Fax: piero.angeletti@esa.int In this paper we present the design of a Satellite Multimedia Broadcasting Mobile System (S-MBMS) operating at S-band as a possible mission embarked on-board of the proto-flight platform in the frame of the ESA program. The mission definition includes the system targets in terms of service, user terminals and coverage. System design trade-off will be then discussed as well as the selected frequency plan for the feeder and user links (satellite and terrestrial repeaters), key payload performance parameters in terms of gain, C/I as well its mass and power estimate. Link budget analysis is then provided to show the achievable link margins over the coverage region. I. Introduction Alphabus (@BUS) is a geo-stationary platform able to accommodate a wide range of future multi-band telecommunication missions. The Alphasat (@SAT) program is aimed at define, design, develop and launch a multimission satellite using the Alphabus platform by 2010 [1]. In an attempt to define candidate missions sharing platform, the European Space Agency initiated a feasibility study whose outcomes on a possible Satellite Multimedia Broadcasting Mobile System (SMBMS) are reported in this paper. The proposed S-MBMS mission intends to complement the Terrestrial UMTS networks (T-UMTS) by offering broadcast/multicast services to various classes of users on upgraded 2,/3 G user terminals, using the portion of S- Band (2 GHz) reserved to satellite. The S-MBMS concept is derived from content delivery network architectures developed for fixed IP networks [2]. As price evolution of storage and transfer technologies is increasingly favoring storage, architectures which rely on one to many distribution mechanisms of content to be locally cached ( push and store ) will increase in the coming years their business advantage toward remote and centralized client/server architectures. The content delivery architecture for a mobile network is shown in Figure 1. Figure 1: Content delivery architecture for mobile network 1

2 Therefore the basic mission of the S-MBMS system is to provide traffic optimization mechanisms that rely on multicast content delivery to the user and are exclusively intended to increase content transfer capacity over 3G networks. The concept also relies on the ability of any kind of UMTS handset to access this multicast layer. This is based on the reuse of the W-CDMA UTRA FDD air interface. Relying on MSS frequency bands adjacent to IMT removes also the need for new frequency band support for the user terminal, thus allowing the provision of multi-mode terminals (unicast and multicast) at marginal extra-cost. This benefit is balanced by the need to deploy terrestrial gap-fillers to provide the required QOS in urban and suburban areas. The system is aiming at a European coverage by means of a single GEO satellite complemented by a number of terrestrial gap-fillers operating at the same frequency. We assume as baseline the use of the 3GPP W-CDMA physical layer to maximize commonalities with the 3G UMTS terrestrial standard. However, being the payload transparent, other schemes can be supported as well (e.g. TDM or OFDM). The required quality of service will be obtained by the combination of link margin, exploitation of Forward Error Correction (FEC) and interleaving over a large block of data and when possible carouseling techniques. As shown in the companion paper [3], this upperlayer FEC mechanism allows to efficiently counteracting time limited link obstructions, deep shadowing and fading provided that a transmission delay is acceptable for the delivered service. The proposed S-MBMS payload is characterized by a DC power consumption of KW thus consuming about half of platform DC power. The remaining power can be used for other missions sharing the same platform. The S-band payload consists of a multi-shaped-beams antenna system based on a large deployable reflector. The high power transmit section is composed by a cascade of a low-level Beamforming Network (BFN) a stack of High Power Amplifiers (HPAs) and a lossless high-power recombination section based on Butler-like microwave hybrid matrices. The selected payload architecture, whose trade-offs are described in a detailed companion paper [], allows full power re configurability among the linguistic beams and their shaping according to the required contours. II. Mission Requirements A. Service Requirements Taking into account the technical characteristics of DMB, the potential services have been split into 3 service categories: Entertainment services, Information services and Webcasting (Web site replication). For each of those categories, the applications that seem to be suitable for the DMB broadcast mission are detailed in Table 1. Entertainment services Information services Webcasting Pre-stored audio on demand services Pre-stored music on demand Pre-stored radio on demand Pre-stored video on demand services Pre-stored movie on demand Pre-stored WebTV or TV on demand Pre-stored video clips (ex : movie trailers) Software delivery: games, software for mobile personalization (ring tones, logos, animated images, ) Audio streaming : as a support service (e.g. if cache is empty) Video streaming : as a support service (e.g. if cache is empty) Basic information (text): Traffic information, tourist information, finance, news, public information (security, ), job offers, weather forecast, horoscope, classified adds, Rich information (audio, video, image & text): sport (video clip showing a goal during a football game), video document for news, Personalized information portal: based on DMB cache and filtering capabilities Replication of sites: most famous sites, shopping gallery sites, In-car dedicated web site replication Replication of Intranets or mobile professional portals Table 1 : Service classification for the broadcast mission 2

3 The service categories are described according to their main characteristics : Delivery types : Real-Time (RT), Non-Real-Time (NRT) or Quasi-Real-Time (QRT). RT delivery is not really in the scope of DMB as DMB is more suitable for caching and delivering content. Among all the selected applications, audio or video streaming are the only RT applications. Streaming should only be a support service (for example if the cache is empty or outdated). However some broadcast information is time-sensitive (e.g. sport news or traffic info) and will be delivered in QRT. Application types : Car related or Unspecific to vehicular. It is interesting to distinguish between applications that are specific to the car environment (e.g. traffic information) and those which are not related to the vehicular environment (e.g. video). Market : Consumer or Corporate markets. The main market to consider is the consumer market. B. Architectural requirements The satellite based multicast layer architecture is designed to avoid any modification to the 3GPP architecture with which it interacts. It uses terrestrial 3GPP, IETF standardized technology to access a large market. The satellite based multicast layer architecture for mobile networks is shown in Figure 2. S-MBMS satellite W-CDMA MSS band FSS band W-CDMA MSS band S-MBMS Terrestrial repeater S-MBMS HUB UMTS/GPRS Multi mode terminal «S-MBMS enabled» Node B 3GPP RAN RNC 3GPP CORE NETWORK S-MBMS server Figure 2: Satellite based multicast layer architecture for mobile networks CDN Content provider Content provider It consists of a space segment, made of satellites, and a number of terrestrial repeaters. Mobile terminals receive appealing multimedia contents prepared or forwarded by the S-MBMS server and conveyed in a point-to-multipoint connection provided by the space segment. The service is delivered with a point-to-point connection provided by a mobile network (2G or 3G). The S-MBMS server interacts with the mobile network via standard interfaces. It consists in a central part colocated with a hub in charge of content multicasting and secondary entities installed in different countries to handle the point-to-point connections. The S-MBMS server is responsible for the content acquisition from external content providers, via content distribution networks or from other sources. To adapt to the satellite channel characteristics, the server applies on the multimedia content some efficient transport techniques relevant to the targeted service requirements. It relies on preventive techniques such as forward error correction (FEC) coding, interleaving (defined in the IETF standard Reliable MulTicast Protocol, RMTP) as well as content retransmissions. If still some blocks are 3

4 lost during the transmission, selective block recovery mechanisms are provided by making use of the point-to-point connectivity via the terrestrial mobile network. The server implements routing features to allow interoperability between the S-MBMS space segment and the mobile cellular network. This heterogeneous network interconnection allows establishing the point-to-point connectivity also referred to as return link. To counteract heavy shadowing and provide coverage continuity in urban and sub-urban areas, low cost terrestrial repeaters will be deployed in the Line Of Sight (LOS) of the satellite. The repeaters are designed to be smoothly co-sited with 3G/2G base stations. S-MBMS signal reception requires standard 3GPP terminal parts. S-MBMS requires a terminal to accommodate local storage capability which is no issue considering the technology trends. The S-MBMS software package is designed to run in any mobile terminal environment. It shall also allow operation on a 3GPP network (GPRS or T- UMTS) to establish a point-to-point session with the server for the interactive link. The S-MBMS system is designed to ensure a combination of the S-MBMS signals received from the satellite as well as the terrestrial repeaters with a standard rake receiver. The terminal establishes a S-MBMS point-to-point connection with the server via a 2G or 3G mobile network, to achieve retrieval of missing packets to reconstruct damaged files, data collection for audience and service usage measurement, user profiling management, exchange of MMI commands and m-commerce transaction on the received contents. The proposed S-MBMS consists of three segments: Space, Ground and User Segment. The Space Segment consists of a single transparent (i.e. not re-generative) satellite(s) characterized by high EIRP (i.e. large antenna and High Power Amplifiers) The Ground Segment is composed by: the Broadcast/Multicast Service Control Centres, providing the functionality of Contents Aggregator, scheduling the transmissions and implementing the recovery procedures to guarantee data integrity: programs carousel (long range time diversity) channel repetition (short range time diversity) macro-blocks coding the Feeder Link Station (FSs), widespread over the territory and providing the satellite interface to the Broadcast/Multicast Service Control Centre(s). A single operator, as well as a multiple operators configurations can be easily supported by the system. The architecture can be derived from the one of T- UMTS Node-B the terrestrial repeaters (TRs) re-transmitting the same broadcast/multicast content delivered by the satellite (in the same S-band used by the satellite) within the areas where the satellite signal cannot penetrate (e.g. urban environment) the Network Control Centre (NCC) providing monitoring and control functions, resource management in terms of frequency and codes assignment to FSs and Gateways the Satellite Control Centre (SCC) in charge of control and monitoring the satellite and of communicating to the satellite by means of the Tracking Control and Ranging (TCR) station. C. Coverage requirements The target area service is still Western Europe but the coverage is realized by means of either single national spots covering a given region or language homogeneous area or multiple spots. As a minimum the following countries shall be covered by the S-MBMS service: France, Italy, Germany, United Kingdom, Spain, part of Eastern Europe. Tentatively the satellite will be located at 31 degrees E. Figure 3: Coverage (Pictorial View) 4

5 D. Satellite Terminal requirements The User Segment includes several types of User Equipment (UE) (i.e., handheld, vehicular and portable terminals) characterized by different antenna and receiver performance and able to receive both terrestrial and satellite signals. In case of environments where the satellite signal is obstructed (e.g. buildings) or it is insufficient for the terminal G/T (e.g. handheld), a wireless interface (e.g. Bluetooth interface (BI)) in conjunction with a satellite interface is adopted. The user terminals antenna, power, dimension shall be compatible with those already exiting on the market for 2./ 3 generation communications systems and shall have large memory storage capabilities to support infomediary, non-real time multicast services. Table 2 summarizes the typical user terminals RF performances (as reported in the ETSI TR 101 ). It assumes that the terminals are featuring circular polarized satellite antenna and dedicated RF front-end. The reuse of the same terrestrial hand-held mobile front-end will significantly reduce the achievable G/T (e.g. by to db) thus will have a major impact on the link margin. Handheld Nomadic (integral antenna) Vehicular / Nomadic (external antenna) Antenna Gain [dbi] System Temp. [K] G/T [db/k] ~ Table 2: S-MBMS Terminals Typical RF Performances E. Frequency allocation User link: The proposed user downlink is S-Band: MHz but to allow the coexistence of multiple operators as the S-MBMS system will use 1 MHz (contiguous) i.e. half of the allocated band. Feeder link: The 2./30 GHz Ka-band shall be considered as baseline for both the broadcast uplink feeder link and the 1./20.2 GHz Ka-band for the terrestrial downlink feeder link. The Ka-band sharing with possible missions exploiting the same band is TBD. The utilized frequency bands are summarized in Table 3. Fixed links Feeder stations satellite Satellite terrestrial gap fillers Frequency Band Ka-band Frequency GHz GHz Mobile links Satellite users S-band MHz Table 3: Feeder Link Frequency Bands F. Power Flux density limits The following table summarizes the PFD limits in the band MHz: PFD limit db(w/(m 2 MHz)) 0 δ Angle of arrival δ (degree) PFD limit db(w/(m 2 4 khz)) δ δ Table 4: PFD limits in the band MHz:

6 G. Physical layer 3GPP W-CDMA baseline Taking into account investigations carried out in several ESA ARTES 1 studies [3], it could be stated that the 3GPP UTRA FDD mode W-CDMA radio interface could be used for the broadcast mission of the S-MBMS system. The 3GPP W-CDMA reuse for the S-MBMS advantage resides in: Large commonality with currently deployed European UMTS standard Full exploitation of the recently standardized terrestrial MBMS mode in 3GPP W-CDMA release Short-term availability of mass market UMTS/MBMS terminals The main drawback of the W-CDM(A)/MBMS adoption is related to the limited throughput achievable in the terrestrial gap-fillers Single Frequency Network (SFN) exploitation due to the co-channel interference and CDM orthogonality loss due to the channel frequency selectivity. It is expected that the development of more advanced CDM(A) demodulators than current rake receivers with interference mitigation capability may allow to enhance the SFN performance. In the short term a possible solution is to limit the terrestrial SFN CDM load. The satellite CDM multiplex is then terrestrially re-transmitted on the three frequencies thus limiting to 1/3 the number of CDM Walsh codes per frequency slot compared to the satellite. The satellite will be able to transmit a higher number of Walsh codes/frequency slot as the satellite channel is not frequency selective thus is better preserving the CDM multiplex orthogonality. This approach may lead to some performance degradation in the hand-off region between the satellite and the terrestrial gap-fillers. An alternative solution is to avoid to reuse the satellite frequency terrestrially and to terrestrially transmit half of the satellite Walsh on the two other frequency slots. In the following we limit the analysis to the satellite component disregarding possible terrestrial gap-filler interference. Alternative physical layer schemes As the satellite payload is transparent alternative modulation schemes to the 3GPP W-CDMA can be considered. In particular, a review of terrestrial digital broadcasting standards in [] showed the following results: DAB DVB-T DVB-H Occupied Bandwidth 1. MHz MHz MHz (down to MHz optional) Suitable mode Mode III 2K mode 4K mode # of OFDM Carriers Carrier Spacing khz 4. khz 2.2 khz FEC scheme Convolutional Reed-Solomon & Convolutional Reed-Solomon & Convolutional Coding rate 1/2 1/2 1/2 Useful Data Rate 1.12 Mbps 4. Mbps 4. Mbps Detection mode Differential Pilot-aided coherent Pilot-aided coherent Maximum Allowable Vehicle Speed in S-band (2.2 GHz) ~20 km/h ~100 km/h ~0 km/h Cell Radius max 12 km 1 km 33 km Modulation D-QPSK QPSK QPSK Useful Symbol Duration 12 µsec 224 µs 44 µs Guard Time T U / 4 = 31 µsec max T U /4 = µs max T U /4 = 112 µs Bandwidth Efficiency [bps/hz] Table : Alternative physical layer schemes

7 In order to better compare the two standards performances, the last row of the table shows the bandwidth efficiency defined as Useful Data Rate / Bandwidth [bps/hz]. In DAB the bits are differentially encoded and hence the system does not need channel estimation and is simpler to design. Since in DVB-T system, a coherent demodulation has been chosen, a channel estimation is necessary to compensate the distortions caused by the transmission channel. If the channel estimation is done correctly, it may outperform incoherent detection. But the channel estimation will not be perfect, especially in mobile environment, and will result in an estimation error that will increase the BER. On the other hand it has been found that a BER of after a Viterbi decoding may not be needed for a subjective high quality video for mobiles. The actual BER may be , which can result in a SNR margin improvement of up to 3 db, that can compensate the non-perfect channel estimation. A time interleaving, which is available in DAB would help to mitigate the effects of short and deep fades caused by shadowing. However, time interleaving has disadvantages like increased system latency and greater reacquisition time after the signal is lost. The presence of a frequency interleaver allows both the systems (DAB & DVB-T) to benefit from the frequency diversity. The DAB is designed for frequency up to 3 GHz whereas the DVB-T is intended to operate in terrestrial TV bands i.e. up to Band V (below 0 MHz). For this reason the DAB system is more robust to the S band mobile environment effects than DVB. Moreover, the DVB-T signal has a bandwidth that is 4 - times larger than DAB, therefore the user terminal processing capabilities requirements is more stringent for DVB standard than in case of DAB standard. The results of the performed comparison are summarized hereafter: 1. DAB mode III and DVB-H are the most suitable standards for the mobile broadcasting 2. The performances of the two system are comparable in mobile environment with the following features: DAB gives higher spectral efficiency (~% DAB vs. 2% DVB-H) DVB-H 4K mode gives larger SFN cell radius (~33 km DVB-H vs. 12 km DAB) DVB-H is more power efficient than DAB 3. The main issue is about the frequency band: assuming the terrestrial gap filler operating in S band, the DAB system is more robust against the high Doppler effects 4. The DAB standard allows to operate with a less stringent user terminal processing capabilities (1. MHz vs. / MHz of DVB-H). DVB-H features as well the time slicing capability that can greatly reduce the user terminal power consumption. Furthermore the DVB-H MPE-FEC allows to enhance the OFDM performance in case of high mobility conditions. Results reported in [] showed the superior DAB single frequency gap-filler network efficiency compared to W- CDMA scheme. OFDM performance over nonlinear channel can get very close to the CDM one if a powerful FEC is adopted. The evolution of DVB-H to better support satellite operations at S-band is currently being investigated within the DVB Technical Module. Another alternative is represented by the use of TDM scheme for the satellite to mobile user link and OFDM for the terrestrial gap fillers. However, this approach which has been adopted by US DARS systems, has limited advantage as the multibeam S-MBMS payload will anyway feature HPA multi-carrier operation, thus CDM, OFDM, TDM difference will be limited. Furthermore the DARS physical layer approach will require a dedicated frequency band for the terrestrial gap-fillers as well as two dedicated satellite and terrestrial demodulators. H. Channelization Detailed simulation results [3] proved that the best freq. reuse scheme is the 1:3 frequency reuse as it offers the highest spectral efficiency for the broadcast mission. This approach is compliant with the 1 MHz intended S-band occupancy as it requires three slots of MHz in each beam. Another advantage is that it allows terrestrial gap fillers to use the 1 MHz band to reduce the number of active Walsh sequences present in each MHz FDM slot. This approach helps reducing the co-channel interference thus improving the system availability. This is possible if the gap filler demodulates at bit level the Ka-band MHz CDM corresponding to the S-band national/linguistic beam coming from the satellite. Then the gap filler re-modulates the satellite CDM multiplex over three adjacent MHz CDMs each containing one third of the satellite Walsh codes.

8 Note that each beam is assumed to be covered by a different physical layer scrambling sequence at physical layer (as done in terrestrial 3GPP W-CDMA for differentiating base stations) thus allowing full reuse of the channelization Walsh sequences. Consequently, in the following the CDM isolation among beams has been considered to be corresponding to a random-like spreading. Alternative approaches proposed in the past by companies working in S-MBMS have not been retained because: Orthogonality among a subset of beams using the same physical layer scrambling sequence will not allow to reuse the same pilot Walsh among beams, thus violating the 3GPP W-CDMA standard. Furthermore, simulations indicated that this approach can create correlation among signals going to different beams and thus problematic to be used in payloads with nonlinear HPAs shared among beams. Orthogonality among beams using optimized physical layer scrambling sequences is not considered feasible. In fact, the cross-correlation properties of the scrambling sequence will be largely impacted by the chip X-oring with the channel dependent Walsh sequence used for CDM multiplexing. Thus the optimization can not be performed on the scrambling sequences but shall account all possible combination of the Walsh superimposition at chip level. This is considered to be unfeasible in practice. A. Feeder link The feeder uplink station generates the signals to be uplinked to the satellite at Ka-band according to two different formats: a) FDM multiplexed W-CDM 3GPP (or alternative physical layer configurations) for the user feeder link, b) DVB-S2 single carrier TDM CCM format for the gap-filler feeder link. The reason for doing that is the need to have a low-cost consumer type of Ka-band receive equipment at the terrestrial gap-filler. Furthermore, for UMTS user terminal rake correct operations, the gap-filler Ka-band signal shall be advanced compared to S- band one to compensate for the demodulation delay at the gap-filler demodulator site. In fact, to reduce the CDM load on the terrestrial frequency slot one shall demodulate and demultiplex the satellite feeder link multiplex to regenerate three parallel streams at its output. It is then proposed to use a TVRO DVB-S2 equipment for that purpose. The bit streams corresponding to the different terrestrial FDMs can be labelled using the MPEG2 PID TS information. Buffering will then be required to align the three streams in time. The proposed frequency plan for the feeder uplink is shown in Figure 4. The CDM carriers feeder user uplink is transmitted at Ka-band in X-polar. Each carrier occupies MHz with a 1 MHz spacing. The TDM feeder DVB-S2 uplink is transmitted in Y-polar in a 2 MHz bandwidth. 30 MHz MHz MHz MHz MHz MHz MHz MHz MHz MHz MHz X-polar CDM/FDM feeder user uplink Y-polar DVB-S2 TDM feeder gap-filler uplink 2 MHz Figure 4: Ka-band feeder uplink frequency plan It is assumed that the CDM Ka-band feeder uplink multiplexes (one CDM/beam FDM multiplexed) are on-board transparently demultiplexed and then converted to the corresponding S-band user downlink beam frequency.

9 III. Payload Characteristics The block diagram of the S-MBMS baseline payload is presented in Figure. It comprises a Ka-Band feeder link receive section, an S-Band section and a Ka-Band gap filler section LO2 F LO3 LO1 F3 F OMT LO2 F LO3 F DIP Figure : S-MBMS Payload Block Diagram The Ka-band feeder link signal for the S-band is composed of -frequency slots ( user beams) of MHz in linear vertical polarization (V), while the feeder link signal for the gap filler is in the orthogonal linear polarization (H). At the S-Band section output the signal is composed by three frequency slots (FDM). These three frequency slots are used to perform the frequency reuse scheme on the user coverage. The gap filler section performs the broadcast operation of the DVB-S2 stream. A Ka band Tx/Rx antenna operates polarization separation of the orthogonal received signals (via the orthomode transducer OMT). One linear polarization of the OMT (H) will also be employed for Ka-Band gap filler transmission and a diplexer is needed before low noise amplification. After the LNA, the Ka-Band gap filler signal is directly converted to the Ka-band Tx frequency (single conversion), it is amplified, filtered, fed to the diplexer Tx input port, then to the OMT H-port and finally reradiated by the Ka-band antenna. After amplification, the S-Band feeding signal is converted to IF and spited to parallel branches where it is appropriately filtered and converted to the 3 S-band frequency slots (xmhz filters of 3 types and xlo of 3 frequencies). The power level of each beam channel can be independently adjusted by a settable gain amplifier. Afterwards, the required RF power is obtained by the high power section that represents the central heart of the full payload. The high power section solution shown in Figure is the result of an extensive trade off which are documented in companion paper []. Thanks to an advanced payload and antenna numerical optimization we show how it is possible to achieve good beam isolation making the intra-system interference effect negligible. Key elements of the solution are: A low signal level phase-only BFN. A fully shared stack of 32 TWT amplifiers in a power pooling configuration (in groups of 4: phase tracked redundancy blocks). A stack of Butler-like matrices. An array of 32 feeds appropriately connected to the hybrid matrices. An S-band Large Deployable Reflector of 12 meters projected aperture.

10 A. S-Band Transmit Section Outstanding satellite antennas apertures are required at S-Band. Large deployable reflectors (LDR) are enabling elements to achieve the objective and recognizing the need of European R&D activities on this critical technology, the European Space Agency (ESA) awarded a development contract to an Alcatel Alenia Space (Italy) led team supported by the Russian NPO EGS. The ESA funded program, which is now terminating, aimed at realizing a space-qualified deployable reflector of 12 meters projected aperture diameter and relevant mechanism, adopting lightweight technologies. Considering the schedule criticality of reiterating the LDR development for an optimized optics, we assumed the existing LDR geometry []. In this respect it is worth noting that ongoing LDR development was already targeting similar L/S-band requirements and the optical configuration can be considered well suited for the current mission. The feed array geometrical configuration has been selected based on the following assumptions: One-to-one correspondence of number of feeds to number of amplifier (No TWT power combining). Number of feeds traded-of to achieve a reasonable power at feed level Feed position adequate to cover the target geographical zone with a sufficient number of elementary beams The selected configuration consists of 32 feeds of about 0. λ diameter on a planar equilateral triangular lattice placed in the focal plane of the reflector. Table 11: LDR during PIM testing in RSC Energia facility (Curtesy Alcatel Alenia Space / NPO EGS / RSC Energia [][]) Beam # EOC Directivity C/I UK FR NA SP D IT Table : Baseline Antenna Performances Item Performance TWT Saturated Power 140 W TWT Working OBO 0% Number of Active TWT 32 Output losses 1 db Antenna Ohmic Losses 0.2 db Table : S-Band High Power Section Performances Item Performance Rx Antenna Directivity 3.2 dbi Input Losses 1. Ka Band LNA Noise Figure 2. db Ka band TWT Saturated Power 100 W Output losses 2.3 db Tx Antenna Directivity 3. dbi Table : Ka-Band Section Performances DC Power (W) S-Band Transponders 100 Ka-Band Transponders 210 Total 310 Total (with Miscellanea and Margin) 000 Table : S-MBMS Payload Power Budget Mass (kg) S-Band Transponders 230 S-Band Antenna (Hybrids /Feed Array/LDR) 20 Ka-Band Transponders 10 Ka-Band Antenna Total 4 Total (with Miscellanea and Margin) 0 Table 10: S-MBMS Payload Power Budget 10

11 Main performances of the S-band payload are reported in Table and Table and relevant radiation patterns and C/I plots are shown in the following figures... V [deg] U [deg] Figure : Minimum EOC Gain.... V [deg]. V [deg] U [deg] U [deg].... V [deg]. V [deg] U [deg] U [deg] Figure : C/I Maps 11

12 IV. System Performance Assessment The system assumptions made for the feeder link transmit station are reported in Table 12 both for the user and for the gap-filler feeder-link. A. Gap-Filler Analysis Considering the Gap-Filler receiver characteristics reported in Table 13, link budgets were performed for the gap-filler Ka-band feeder link in case of the worst-case fading in the up or in the downlink for an availability of.% for the uplink and. % for the downlink. DVB- S2 QPSK CCM format with coding rate ½ has been assumed. A 4 db worst-case margin was achieved. B. User-Link Analysis As mentioned in section III.G we assumed as baseline the use of the 3GPP W-CDMA physical layer with the following key physical layer parameters: Useful bearer rate per CDM channel = 34 Kbps Code chip rate = 3.4 Mcps Walsh sequence spreading factor = 1 CDM codes non orthogonality factor e= 0. Optimized sat OBO (multicarrier HPA) = 1 db Nonlinear channel Eb/No losses = 1 db Required Eb/No = 2.4 db (BER = 10- with turbo code 1/2) Pilot and signalling overhead = 20 % Feeder Link type (POL) User (X) Gap-filler (Y) Tx frequency GHz TX power at Satur. W Antenna diameter m.0.0 Antenna efficiency Antenna peak gain dbi Output back-off db Output loss db N carriers X-Pol - CDMs 1 EIRP per pol dbw Pointing losses db Eff. EIRP per carrier dbw NPR db Cross-Pol Isolation db Table 12: Feeder Station TX characteristics Carrier frequency GHz 20 Antenna diameter m 0. Antenna efficiency Antenna gain db 41.0 Sky temperature K 4.00 Ground temperature K Antenna temperature K.00 Pointing loss db 1.00 Feed loss db 0.30 Receiver noise figure db 2.00 Receiver temperature dbk 22.2 Clear sky noise temperature dbk 24.1 Terminal G/T (clear sky) db/k 1.22 Terminal G/T (fading) db/k 1.3 Table 13: Gap filler Rx characteristics The broadcast link-budgets reported in Table 14 are based on these assumptions. The CDM analysis outcome is shown in Table 1; the total throughput with about db margin over AWGN is equal to 1.3 Mbps (34 Kbps/Walsh Walsh/CDM beams). This capacity corresponds to a total occupied bandwidth of 1 MHz total for the "linguistic" spots. In case DAB OFDM is used as physical layer scheme instead of CDM, simulation and system analysis reported in [] showed that over the satellite channel OFDM is less power efficient than CDM (about 2 db due to extra backoff required and less powerful FEC). However, when considering the SFN terrestrial gap filler network OFDM has a higher spectral efficiency than CDM. Furthermore, by using DVB-H with coherent detection and more powerful FEC the OFDM is probably a good alternative to CDM for both satellite and terrestrial channels (same OBO can be used). V. Conclusion The paper presented the design of an S-band Satellite Multimedia Broadcasting Mobile System (S-MBMS) as a candidate mission for the program. Mission definition and preliminary payload design was reported together with link budget analysis and capacity assessment. The outcomes of the feasibility study clearly showed the viability of the proposed mission. Acknowledgments The authors thankfully acknowledge the support of Space Engineering (Italy) for the analysis and design of the S-Band mission antenna. 12

13 Feeder user link Feeder user link Waveform fading up/down fading up only Symbol rate Ms/s Roll-off factor Carrier bandwidth MHz Gateway TX characteristics Tx frequency GHz TX power at saturation W Antenna diameter m Antenna efficiency Antenna peak gain dbi Output back-off db Output loss db Number of carriers EIRP per carrier dbw.3.3 Pointing losses db Effective EIRP per carrier dbw.3.3 Gateway -> Sat propagation Elevation Angle deg Orbital height Km Range km Free Space Loss db Clear sky attenuation db Additional attenuation db Total attenuation db Sat reception Satellite antenna directivity dbi ohmic losses db Satellite RX temperature dbk Satellite G/T db/k TXP noise bandwidth MHz Received noise power dbw C/No Up-Link dbhz C/N up-link db.1.1 Received uplink power (clear sky) db Received uplink power db Overall C/I due to co-frequency channels db C/I due to adjacent-channel interference db C/I due to adjacent satellite systems db NPR db Cross-polar Isolation db Total up-link C/I db Received interference power Up-link results Received signal power dbw Received noise plus interference power dbw Power robbing factor Up-link C/(N+I) db Up-link C/(No+Io) db Satellite transmission Tx frequency GHz Number of carriers per TWTA 1 1 TWTA saturated power W TWTA saturated power dbw OBO db Output losses db TWTA effective output power dbw TWTA effective output power W Tx effective output power/carrier dbw Antenna directivity dbi Ohmic losses db Pointing loss +feed loss db EIRP effective/carrier dbw.0.0 Sat -> Terminal RX propagation Elevation Angle deg Orbital height km Range km Free Space Loss db Atmospheric attenuation db Additional attenuation db Total attenuation db Terminal Rx reception Antenna diameter m Antenna efficiency Antenna gain db Sky temperature K Ground temperature K Antenna temperature K Pointing loss db Feed loss db Receiver noise figure db Receiver temperature dbk Clear sky rx noise temperature K Clear sky rx noise temperature dbk Tm K DeltaT Rx Noise temperature K Rx Noise temperature dbk Terminal clear sky G/T db/k Terminal G/T db/k Received total signal power dbw Received useful signal power dbw Boltzmann constant dbw/k/hz C/No Down Link dbhz 0.. C/N down-link db Number of beams/cdm 1 1 Bit rate per CDM Walsh Kb/s Number of Walsh/CDM multiplex CDM chip rate Mcps TXP power robbing coefficient CDM codes non orthogonality factor Pilot overhead/cdm % DOWNLINK Received power/cdm dbw Pilot power/cdm dbw Power/Walsh dbw CDM Walsh C/No dbhz CDM Walsh C/Io dbhz CDM Walsh C/Nt dbhz Bit rate dbhz.4.4 Downlink Eb/Nt db UPLINK Received power/cdm dbw Pilot power/cdm dbw Power/Walsh dbw CDM Walsh C/No dbhz CDM Walsh C/Io dbhz CDM Walsh C/Nt dbhz Bit rate dbhz.4.4 Uplink Eb/Nt db Total Eb/Nt db Required Eb/Nt (AWGN) db Margin (AWGN) db Table 1: Broadcast Link Budget (CDM analysis) References [1] ESA Telecom Web Announcement of Opportunities [2] N. Chuberre, G.E. Corazza, M.G. Francon, C. Nussli, C. Selier, A. Vanelli-Coralli, P. Vincent, Satellite Digital Multimedia Broadcasting for 3G and beyond 3G systems, 13th IST Mobile & Wireless Communications Summit, Lyon (FR), [3] R. Berto Monleon, A. Bolea Alamanac, R. De Gaudenzi, First Experimental Results of Reliable Satellite Mobile Multimedia Broadcasting via Satellite Exploiting the W- CDMA Air Interface, Submitted to the 24th International Communications Satellite Systems Conference 200. [4] ARTES 1 S-UMTS System Studies by Alcatel (CN 131.NL/US/) and Alenia (ESA CN 13/NL/US/) [] P. Angeletti, P. Gabellini, N. Gatti, Optimization of a Flexible Payload Architecture for S-Band Mobile Broadcasting to Multiple Shaped-Beams, Submitted to the 24th International Communications Satellite Systems Conference 200. [] Alenia Spazio (I) Final Report for CCN 4 to ESA Contract No. 13//US/NL, S-UMTS System Study, 200. [] G.L. Scialino et alii, Presentation of Reflector Dish Development Activities and Achieved Performances, Proceedings of the 2 th ESA Antenna Workshop on Space Antenna Systems and Technologies, 200 [] G.L. Scialino et alii, Large Deployable Reflector Study - Presentation of Achieved Performances on Passive Intermodulation Initial Tests, Proceedings of the 2 th ESA Antenna Workshop on Space Antenna Systems and Technologies, 200 Total C/I cochannel interbeam C/I due to adjacent satellite systems db C/I crosspolar NPR db C/I down-link total db Table 14: Broadcast Link Budget 13