Copyright 1999 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
This article was published in the September 1999 issue of
IEEE Communications Magazine.

CIRULE1.GIF (372 bytes)

Abstract
The third-generation IMT-2000 initiative of the International Telecommunication Union is aiming at the provision of a limited family of global standards providing worldwide multimedia wireless services in a host of environments encompassing indoor picocells to satellite megacells. The ITU IMT-2000 initiative has been supported by several regional standardization bodies such as the European Telecommunication Standard Institute with its Universal Mobile Telecommunication System. In this article a few air interface proposals for the satellite component of UMTS/IMT-2000 based on adaptation of the emerging terrestrial wideband CDMA standards are reviewed. In particular, S-UMTS requirements are examined together with W-CDMA applicability to the satellite environment. It is shown that with minor adaptations, the terrestrial W-CDMA air interface provides an efficient solution for the satellite UMTS component. This commonality will certainly help in the realization of low-cost low-size dual-mode user terminals encompassing both terrestrial and satellite components.

 

CIRULE2.GIF (100 bytes)


Satellite UMTS/IMT2000 W-CDMA Air Interfaces

CIRULE3.GIF (212 bytes)

Payam Taaghol and Barry G. Evans, University of Surrey
Enrico Buracchini, CSELT
Riccardo De Gaudenzi, ESTEC, European Space Agency
Gennaro Gallinaro, Space Engineering S.p.A.
Joon Ho Lee, Research and Development Group of Korea Telecom
Chung Gu Kang, Korea University

 

Over the last decade digital cellular networks made mobile (voice) communication accessible to almost anyone. This has been possible thanks to early agreement on a common standard. To complement current terrestrial cellular networks, several systems based on low/mid earth orbit (LEO/MEO) constellations operating at the L/S-band have been or are being deployed to provide global mobile satellite personal communications (GMPCS or S-PCN).
But this is only the beginning. In recent years the staggering growth of the Internet paralleled by the success of second-generation mobile cellular systems highlighted an interesting trend. Mobile multimedia services are therefore expected to be in high demand by mobile wireless users on a global scale. While second-generation digital cellular networks can already cope with a large variety of requirements, the inherent bandwidth limitations make these networks less suitable for high-speed applications. These limitations are even more severe in the case of first-generation S-PCN. Consequently, a new mobile communication network is required to satisfy the needs of high-speed customers. This new global mobile telecommunication system, referred to as Universal Mobile Telecommunications Service (UMTS) or International Mobile Telecommunications -- 2000 (IMT-2000), should be available on a global basis at the beginning of the next century. The major service objectives set forth by the European Telecommunications Standards Institute (ETSI) and International Telecommunication Union (ITU) are: The S-UMTS component will make outdoor coverage globally seamless while maintaining terminal compatibility and service portability. More specifically, the S-UMTS component's major objectives are: The selected technology for third-generation terrestrial wireless systems will generate great industrial momentum and competition to the customers' advantage. In fact, large industrial investments justified for the huge IMT-2000 market size will make low-cost user terminal chipsets and equipment a reality. Commonality between the satellite radio transmission technology and the terrestrial IMT-2000 component will certainly be the key enabling factor in provision of low-cost small-sized dual-mode terminals. This aim is further enhanced by the availability of multivendor terminals through cross-licensing the IPR while taking advantage of a common air interface specification. The S-UMTS access network is expected to interface with the UMTS core network through standardized interfaces, thereby making efficient use of the high-speed ground segment networking infrastructure.
In this contribution we review the main features of a few satellite UMTS (S-UMTS) air interfaces, all of which are based on wideband code-division multiple access (W-CDMA) technologies very much derived from the emerging terrestrial UMTS (T-UMTS) set. It is shown how a common state-of-the-art W-CDMA air interface can efficiently cope with two different operational environments (e.g., satellite and terrestrial) to the benefit of UMTS users, equipment manufacturers, and service providers.

Requirements

Services

A key requirement for third-generation systems is service flexibility. In such scenarios, the definition of service types is carried out in a new, more open way than in second-generation systems. Third-generation services are defined in terms of transport quantity requirement (bit rate), transport quality requirement (bit error rate), and maximum transport delivery time requirement (delay).
In terms of air interface transportability, ETSI S-UMTS requirements [1] clearly highlight that S-UMTS systems should at least be able to support user bit rates of up to 144 kb/s.1 This bit rate is considered sufficient to offer multimedia applications based on the H.320, H.323, H.324, as well as the emerging MPEG-4 standards. As far as the QoS and end-to-end delay requirements are concerned, for speech services a target BER equal to 10–3 and a maximum delay of 400 ms have been envisaged, while for data applications a BER of 10–6 and a 400 ms maximum delay (for delay-sensitive applications) have been retained. Different delay figures have been considered for each class of data service (e.g., a few seconds for Internet access and a few minutes for e-mail delivery). In terms of service requirements, based on examinations of the ETSI SMG1 WPC work on "UMTS Service Principles" in the satellite context, it is clear that the definition of S-UMTS services should be based on service capabilities rather than strict service-specific transport mechanisms.
It is expected that the S-UMTS system will be capable of adapting to a range of existing and future applications. From the above, three key requirements are derived:
  • Flexibility in terms of the offered bit rate before and during a call (variable bit rate services)
  • Flexibility in terms of the BER provided for a particular service connection
  • Flexibility in terms of delivery delay
S-UMTS is expected not only to complement the coverage of T-UMTS, but also to possibly extend its services. As an example, the S-UMTS wide coverage capabilities may turn out to be more efficient for one-directional applications involving users spread over large geographical areas.

Operational Environments

As discussed earlier, S-UMTS is envisaged to play a complementary role to its terrestrial counterpart. The operational environments are therefore complementary by nature. The UMTS operational environments can broadly be categorized into the following six main classes, shown in Table 1.
Satellite services (with the exception of low-rate services such as paging) will mainly be provided under line-of-sight propagation conditions. The typical operational environments for S-UMTS are therefore areas where provision of terrestrial coverage is either technically or economically not viable.2 Nevertheless, the provision of seamless global coverage calls for overlapping between the terrestrial and satellite UMTS components, particularly during inter-access-network handover.

User Terminal Categories

S-UMTS user terminal types are very much driven by the required set of services. Similar to the terrestrial case, a range of terminal types with different capabilities would best fulfill the varying user requirements. However, in light of the power constraints introduced by terminal and satellite technology in addition to safety implications, it is very likely that S-UMTS terminals will differ from their terrestrial counterparts, particularly in regard to antenna type and battery life. Different types of terminals are therefore associated with diverse user mobility classes as well as peak data rates, as shown in Table 2.
It is important to point out that in the case of vehicular, aeronautical, and maritime terminals, an external vehicle-mounted antenna would be required.
S-UMTS commonality in terms of frequency band and air interface technology can greatly reduce terminal complexity, size, and weight. A dual-mode (terrestrial and satellite) user terminal (UT) should eventually reach a cost comparable to a high-end T-UMTS UT.

Network Compatibility

Subscribing to the Generic Radio Access Network (GRAN) philosophy in which the radio-dependent and radio-independent functionality are handled by the access and core network, respectively, the S-UMTS access network needs to be designed with compatibility in mind. Network compatibility can be reached by designing the higher layers of the protocol stack independent of the radio access technique adopted in the lower layers, following, for example, the approach of the Advanced Communications Technologies and Services (ACTS) Rainbow project. The feasibility of a generic higher-layer protocol architecture capable of handling both the terrestrial and satellite radio interface will simplify the protocol stack of an integrated dual-mode terminal and the interaccess network handover procedures [2]. Furthermore, S-UMTS will benefit from the enormous investments in the common UMTS core network, making S-UMTS a truly seamless extension of T-UMTS services and applications at a customer advantage.

The Choice of Satellite Constellation

The choice of satellite constellation plays a pivotal role in service availability and quality. However, optimization of the constellation parameters represents a major multidimensional challenge well beyond the scope of this article. Nevertheless, it is important to stress that the efficient exploitation of a W-CDMA air interface requires, in addition to conventional constellation optimization issues, such as populated Earth coverage, satellite mass and weight, launch costs, ground segment complexity and cost, network topology, and handover rate, the careful investigation of specific constellation aspects discussed hereafter.
W-CDMA multiple satellite diversity exploitation provides a practical method for reducing blockage effects with little or no impact on capacity compared to a system operating only on a line-of-sight basis (thus not truly mobile). This allows the achievement of superior QoS without affecting system capacity, as is the case for line-of-sight systems requiring large link margins to combat shadowing. The intuition that signal blockage probability greatly reduces with the number of satellites in simultaneous view, recently found confirmation during experimental campaigns whose results are summarized in [3] and illustrated in Fig. 1.
From Fig. 1 it can be observed how, in a typical suburban environment, the probability of blockage varies with the minimum elevation angle and the number of simultaneous satellites in view. With triple satellite diversity almost a tenfold blockage probability reduction is obtained at 10s elevation. For a global coverage system exploiting W-CDMA the constellation shall be optimized trading off multiple satellite visibility for regions with the highest density of mobile users against overall system cost.
From another prospective, different service rates are associated with different levels of mobility. High data rates are generally required for portable multimedia type terminals with very low mobility and larger directional active or passive antennas, making a geostationary earth orbit (GEO)-based solution more attractive. On the other hand, lower-bit-rate services are associated with truly mobile handheld terminals requiring multiple satellite visibility at medium/high elevation angles which can only be offered by LEO or MEO constellations. For this reason the best solution may be represented by a hybrid constellation with regional GEOs complementing the global non-GEO constellation.

W-CDMA Applicability to S-UMTS

In-depth capacity and QoS assessment performed during studies sponsored by various organizations [4, 5] indicated that CDMA/frequency-division multiple access (FDMA) has the highest potential for a global satellite system. For the selected W-CDMA option, the main technical issues are related to the need to offer large system capacity while allowing flexible provision of a comprehensive set of multimedia services. In particular, optimization of system capacity may lead to the selection of slightly different approaches depending on the required system features (i.e., coverage). For regional systems based on GEO or high earth orbit (HEO) the adoption of hybrid code/time-division multiple access (C/TDMA) represents an interesting option [4]. However, in this section only aspects related to W-CDMA applicability to global satellite networks are briefly discussed.
Power Flux Density -- Compared to narrowband systems, CDMA low-power flux density tends to vanish when the system loading in terms of active co-channels is relatively large. In this case the aggregate power flux density approaches the level of narrowband systems. Nevertheless, it is important to recall that CDMA provides an additional dimension for controlling interference to other in-band networks with minimal service disruption. Eventually this capability allows for possible frequency band sharing among different networks.
The Propagation Channel -- In relatively narrowband terrestrial communication systems, the existence of multipath causes severe fading and intersymbol interference (ISI). In terrestrial W-CDMA systems (with bandwidths larger than the channel coherence bandwidth), the multipath environment can be exploited through a RAKE receiver architecture [6], allowing the coherent combining of multipath signal replicas arriving with a differential delay larger than one chip. In addition to coherent multipath rays power combining, this technique is able to mitigate the channel Rayleigh fading effects. This is a unique feature of direct sequence CDMA (DS/CDMA) referred to as multipath diversity, which is essentially a space diversity technique.
Because of the large satellite free space loss and onboard radio frequency (RF) power scarcity, mobile satellite systems normally operate under line-of-sight propagation conditions. This typically results in a mild Ricean (or at most Rice-lognormal) channel with Rice factors ranging from 7 to 15 dB [7], thus causing limited signal envelope fluctuations. Recent wideband channel measurement campaigns [8], showed that for typical low-earth altitude satellite channels the delay spread has an average value of about 100 ns. Hence, any W-CDMA system designed to combat multipath with a RAKE demodulator would have to adopt chip rates in excess of 10 Mchips/s. Furthermore, it has been found that resolvable multipath rays for a 10–15 Mchip/s signal are characterized by much lower power than the line of sight (typically 15–30 dB less), thereby significantly reducing any potential multipath diversity gain for S-UMTS.
Satellite Path Diversity -- Although it appears that for S-UMTS mobile channels CDMA loses its most important advantage (i.e., the capability to resolve and combine multipath), it should be realized that independent fading paths with sufficient path delays can be introduced artificially in the form of satellite path diversity. Satellite path diversity, providing (quasi)-permanent softer handoff conditions, aims at increased QoS [3] by largely reducing the link margins required to combat slow fading and partial link obstruction. In addition to exploitation of satellite path diversity, the RAKE receiver also allows implementing seamless beam-to-beam softer handover. In a CDMA mobile system the path diversity advantage comes almost for free in the return link, due to UT antenna omnidirectionality. On the contrary, in the forward link, to provide satellite path diversity the gateway must purposely deliver the same signal toward two (or three) satellites, and must do so judiciously not to waste system capacity. It was found that satellite path diversity is indeed essential to provide high QoS with minimum RF power requirements from both the satellite and the UT. While this conclusion is quite intuitive for the reverse link, the same cannot easily be argued for the forward link, where path diversity requires resource assignment to the different satellites. The forward link diversity advantage from the capacity point of view has been rigorously demonstrated for realistic mobile multisatellite multibeam working conditions in [8]. This novel important result has been confirmed by detailed system-level simulations.
CDMA full frequency reuse permits reverse link satellite antenna arraying (similar to deep space ground reception techniques) whereby the different replicas of the same signal transponded by the different satellites are independently demodulated, time aligned, and coherently combined at the gateway station. This more sophisticated detection technique requires some additional gateway hardware (essentially a RAKE receiver as a channel unit demodulator), but results in a drastic reduction in the UT equivalent isotropic radiated power (EIRP) even under nonblocked conditions. This CDMA unique feature is felt to be essential in combination with payload improvements to achieve the increased bit rate capability sought by IMT-2000 with affordable terminal EIRP. Another important advantage also related to path diversity consists of the Eb/N0 reduction achieved by spatial diversity signal combining in the presence of slow fading. This is quite important because, different from terrestrial systems, slow fading is counteracted by neither the power control (due to its limited tracking speed) nor the finite size interleaver. For mobile satellite systems, slow fading, characterizing low-mobility users, represents the most power-demanding link condition. With path diversity it is possible to largely counteract the adverse slow fading effects with very modest power margins.
Power Control -- Although the so-called near-far effect in S-UMTS is not as severe as in the terrestrial case (due to the limited link margin and line-of-sight operational conditions), adaptive power control must still be implemented to preserve precious power and therefore system capacity. Slow (trackable) link power level variations are due to different causes such as satellite motion (path loss changes), satellite and user antenna gain variations, specular reflections and shadowing (environmental effect), user speed changes, and time-varying co-channel interference. As in T-UMTS, a combination of open loop for random access channels and closed loop power control for connection-oriented links is required. However, due to the longer satellite propagation delay, a power control correction per frame (not per time slot) is sufficient for S-UMTS. The signal propagation time in satellite systems is not negligible as in terrestrial cellular networks. This makes power control less responsive to fast dynamic changes in the signal received power level. It has been experimentally shown that, for a LEO constellation, closed-loop power control can accurately track shadowing variations with a bandwidth of about 0.1 Hz or less. It has also been found that the power control residual error is only slightly dependent on the satellite altitude. There are several advanced power control techniques capable of improving power control performance. Among these techniques, adaptive [9], multistep size (discussed later), and predictive are the most effective for S-UMTS.
System Capacity -- Another peculiarity of a multisatellite W-CDMA-based S-UMTS system is that the capacity bottleneck is typically represented by the forward link. This is quite in contrast with what is normally assumed for terrestrial cellular systems, where the reverse link optimization is of paramount importance to the system. The main reasons for the forward link S-UMTS capacity limitation are the scarce satellite RF onboard power available for the downlink and the (quasi)-permanent uplink satellite path diversity conditions that make reverse link capacity well in excess of that on the forward link. This explains the interest in CDMA interference mitigation techniques that can be applied to the mobile UT, thus reducing average forward link power consumption. In-depth system investigations showed that under practical scenarios minimum-output-energy-based blind CDMA interference mitigation [10] can boost system capacity 50 percent or more.
User Positioning -- Future wireless communication networks will have increasing requirements in terms of user positioning capabilities. User positioning is important not only for operator billing purposes but also for customer localization (e.g., emergency calls). CDMA adoption in a satellite network provides powerful user localization capabilities through exploiting the available communication signals. Standalone positioning has to be preferred over the exploitation of external means such as Global Positioning System (GPS), because it reduces terminal complexity and power consumption. Multisatellite constellations adopting W-CDMA allow accurate user positioning based on time of arrival or frequency of arrival techniques even when the number of satellites in simultaneous view is limited. This information can also be used for optimization of handover performance.

S-UMTS Proposals

Six radio transmission technology (RTT) proposals for the UMTS/IMT200 satellite component have been submitted to ITU.
The submissions significantly vary in detail, making any objective comparison virtually impossible. Nevertheless, all the CDMA proposals aim to provision services up to 144 kb/s and represent adaptation of terrestrial RTTs to the satellite environment. From this point of view, (1) and (2) in Table 3 showed that it is possible to easily tailor terrestrial technologies for effective support of the S-UMTS component. More detailed information about CDMA-based proposals is provided in later sections of this article.

The European Approach

EC S-UMTS Activities: The SINUS Project

Since IMT-2000/UMTS aims at being both universal and mobile, a great deal of work within the ACTS European R&D projects was and is dedicated to integration of the satellite component into the global system. Work on this topic was initiated in the framework of the former Research in Advanced Communications in Europe (RACE) program through projects such as MONET and SAINT. In line with the requirements of ACTS, Satellite Integration into Network for UMTS Services (SINUS) exploited the results of RACE II on S-UMTS as a technical baseline for design and implementation of the first S-UMTS real-time testbed.
The SINUS system demonstrator implements the various communication layers from physical all the way to application. The demonstrator has been designed and implemented with high flexibility in mind. An all-digital FES and the UT implementation allow for future modifications to the lower layers. The advanced satellite channel emulator enables very flexible real-time emulation of the constellation, spotbeam, and propagation characteristics for any orbit type and operational environment [11].
Along this line, an air interface with some level of programmability to different constellations, capable of supporting service rates up to 144 kb/s, has been designed. Since a high degree of commonality between the terrestrial and satellite radio interface of UMTS is desired, during the air interface selection process particular consideration was given to the work carried out within FRAMES, an ACTS project aimed at defining a suitable air interface for terrestrial UMTS. The FRAMES project defined two operational modes; FMA1, based on the slotted CDMA access technique, and FMA2, based on W-CDMA [12]. After thorough analysis of these two FRAMES modes, a W-CDMA access scheme similar to that of FMA2 [12] was chosen. The scheme provides high data rates by adoption of a variable spreading rate for the uplink (MS/SAT/FES) and a multicode option for the downlink (FES/SAT/MS).
Details of the SINUS air interface can be found in [13]. A summary is included in the following sections of this article.
Doppler Countermeasures -- A simple yet effective Doppler compensation technique is proposed and evaluated for all the constellations. Through the use of such a scheme, the UT Doppler shift range can be reduced; nevertheless, the Doppler rate remains unchanged and relatively fast. Synchronization and tracking loops capable of coping with residual Doppler have been investigated and implemented within the demonstrator modems.
Satellite Diversity Exploitation -- Satellite diversity has been adopted in light of the advantages outlined earlier. The ability to resolve and combine different satellite diversity paths with different constellations has been considered within SINUS. It is preferable to minimize the residual delay between the two paths to be as low as possible to reduce memory requirements in the terminal and the length of the spreading sequence used on the pilot channel [6]. Implementing a simple delay precompensation technique significantly reduces the residual path delay to about 4.1 ms and 6.5 ms in typical LEO and MEO systems, respectively.
Satellite and Spotbeam Handover -- The constellation type and spotbeam pattern influence the duration of a communication within a spotbeam of a single satellite (e.g., an average satellite visibility of around 10 min, and a single spotbeam visibility of 1–2 min). Long calls require several spotbeam handovers and in some instances even handover between satellites. This implies a relatively high in-band signaling load, which reduces the traffic resource efficiency. Within the SINUS project investigation of techniques to reduce this signaling load through the use of accurate positioning algorithms has been carried out.
Adaptable Power Control and Interleaving -- An important feature of the SINUS air interface is the fact that it can be adapted to three different constellation types: LEO, MEO, and GEO. Parameters such as the closed loop power control command rate, step sizes, number of power control command bits, and interleaving depth can all be varied to optimize performance for any given constellation.
Main Air Interface Characteristics -- The frame structure has been designed with the following specifications:
  • Frame duration: 10 ms
  • Channel spacing: 4.8 MHz
  • Chip rate: 4.3008 Mchips/s (a revised solution of 4.096 Mchips/s is foreseen)
  • Rolloff factor: 0.12 (0.172 in the alternative solution)
  • In-frame signaling: 24 bits
Both the 10 and 20 ms frame duration options were analyzed. The 10 ms frame format was eventually chosen because it allows greater flexibility in addition to increased commonality with FMA2. Moreover, a higher granularity in terms of interleaving depth (multiples of 10 ms) and signaling such as power control commands can be achieved.
Within the SINUS air interface definition, 11 different user bit rates have been specified. Supported data rates range from 0 b/s (which implies that only signaling is transmitted to maintain mobile and FES synchronization) to 144 kb/s. In the downlink (FES to mobile) a multicode concept as in FRAMES [12] is proposed. With the multicode option, the information stream at R kb/s is split into N parallel data flows (with a rate of R/N), each spread by using a different 128-bit-long Walsh function. The adopted solution has been optimized in terms of number of users for medium-rate data services by using 128-bit Walsh functions. Data rates up to 16 kb/s are supported by associating a 128-bit Walsh function to one link. Higher data rates are provided by associating the physical link with more than one Walsh function (e.g., the 64 kb/s service can be served by using four codes). Repetition of the information bits is used to have a constant bit rate at the channel coder input (16.8 kb/s), as in the SAINT project. In this way, it is possible to use the same error correcting code for all the services. The adoption of the same coding scheme (a convolutional code with rate 1/3 and constraint length 9) for all services guarantees the same protection to all information bits, reduces hardware complexity (there is no need to implement different coders/decoders for different services), and allows low-bit-rate services to be transmitted at a lower power level in order to reduce power consumption and interference levels. Note that in a CDMA system it is possible to use very-low-rate codes without diminishing the spectral efficiency, since the coding rate contributes to the spreading process.
For the reverse link, a Walsh-Hadamard block code/orthogonal modulation is used with rate 6/64 for incoherent detection; a convolutional encoding is performed with rate 1/3 and constraint length 9 for all the services. The overall processing gain is obtained combining the channel code, the Walsh-Hadamard code/modulation block, and pseudo-noise (PN) sequences with different processing gains.
In order to converge toward a solution adopting 4.096 Mchips/s, as proposed by UTRA, the SINUS project has taken harmonization steps. A harmonized satellite/terrestrial RTT implies only minimum changes with respect to the above-mentioned. In particular, the only difference is the code rate, which changes from 1/3 to 7/20 by puncturing (1 of every 21 bits deleted). The simulation results indicated that link-level performance are practically unaffected by this change in the coding scheme [12, 13]. The modulation schemes adopted for the SINUS air interface are orthogonal quadrature phase shift keying (OQPSK) for the reverse link and QPSK for the forward link, based on the SAINT project results.
Through simulations it has been found that over satellite fading channels with one frame interleaving span, the adoption of convolutional coding only provides better performance than concatenated convolutional and Reed-Solomon codes [12, 13]. As a consequence, only one coding scheme (convolutional codes) for both speech and data services (with different interleaving depth, depending on delay requirements) has been adopted.

ESA S-UMTS Activities

In the IMT-2000/UMTS framework, the European Space Agency (ESA) has undertaken several study activities on the S-UMTS air interface [4, 5] and related technologies, presently heading to a testbed demonstration activity [14]. The ESA S-UMTS W-CDMA air interface definition, dubbed Satellite W-CDMA (SW-CDMA), has been performed with a close look at the ongoing ETSI T-UMTS and ARIB standardization activities in order to maximize commonality. As discussed in more depth in [15], ESA SW-CDMA represents the ETSI UTRA terrestrial frequency-division duplex (FDD) W-CDMA proposal adaptation to the satellite environment. For this reason and lack of space, no detailed air interface description is provided here except for the summary features given below and in a later section. Furthermore, as stated before, emphasis is given to the SW-CDMA RTT devised for global coverage systems that shows the highest commonality with emerging T-UMTS FDD mode proposals such as UTRA. The reader can refer to the above references for detailed information on the proposed air interfaces. The main SW-CDMA features can be summarized as:
  • Support for a wide range of bearer services (from 2.4 kb/s through 32 kb/s to handheld and 144 kb/s to transportable terminals) through orthogonal variable spreading factor (OVSF) spreading techniques
  • Support of packet services through random access or connection-oriented packet-activated physical links
  • Compatible with a wide range of satellite constellations but best suited to global systems
  • High capacity and QoS when applied to LEO/MEO constellations providing multisatellite visibility
  • Full frequency reuse in all beams and satellites, thus simplifying the onboard antenna design
  • Fast initial acquisition thanks to the use of bursty pilot
  • Compatibility with onboard adaptive antenna systems
  • Support of optional forward link regenerative payloads
  • Provision of a high-penetration paging channel and user localization service
  • High power and spectral efficiency obtained by means of:
  • -- Satellite path diversity and coherent combing for both forward and reverse links
  • -- Exploitation of powerful forward error correction (FEC) codes and concatenated coding in case of data transmission
  • -- Introduction of a power/spectrally efficient modulation/spreading format (QPSK with binary spreading and rolloff 0.22 square-root raised-cosine chip shaping)
  • -- Pilot-aided coherent return link with reduced out-of-band emissions with in-quadrature code-division multiplexed control channel
  • -- Combination of open and multilevel predictive closed-loop power control
  • -- Optional adoption of MOE-based blind interference mitigation techniques
The main ESA-SW-CDMA deviations from the ETSI UTRA FDD mode can be summarized as:
  • Optional provision of a half chip rate option to ease frequency band sharing by different operators
  • Feeder link full Doppler compensation and center-of-beam frequency Doppler compensation in the forward link
  • Different (two steps instead of three) initial signal acquisition procedure compared to terrestrial
  • Optional inclusion in the forward link of a short (256 chips) scrambling (randomization) sequence to allow utilization of adaptive linear CDMA interference mitigation detectors
  • Permanent forward link softer handover condition for mobile users
  • Permanent multiple satellite antenna diversity combining in the reverse link
  • Introduction of a high-power paging channel
  • Reduced power control rate to match the satellite propagation delay
  • Multilevel predictive power control loop
  • Longer random access channel preamble

The proposed approach has been validated in depth through comprehensive software simulations at the source coding, physical, and system levels [14, 15]. Each analyzed configuration has been characterized with a merit factor, which can be found as a compromise between spectral efficiency and power efficiency; according to ITU Recommendation M.1225, the following expressions were considered:
Forward Link --


Return Link --


where r is the frequency reuse factor (1 for CDMA), nc is the number of carriers per beam, b is the data rate, ns is the number of overlapping satellites, k is the diversity order, v is the voice activity factor, and B is the bandwidth.
In the following numerical results, the worst case of pedestrian handheld users using pilot symbols in the traffic channels3 experiencing slow Rician fading with no interference mitigating detectors has been considered. Figure 2 (forward link) and Figure 3 (reverse link) show that power efficiency decreases with increasing spectrum efficiency; the choice of these parameters, then, should be exploited considering all system constraints and traffic loading. The lower reverse link efficiency (Fig. 3), in the absence of satellite diversity, is mainly related to the nonorthogonal intrabeam CDMA interference not present in the forward link. In the presence of satellite diversity instead, the capacity of the reverse link considerably exceeds that of the forward link. Concerning slow fading cases, some considerations can be made on the benefits related to diversity adoption. Considering the forward link, the improvement achieved with double diversity in terms of spectrum efficiency is shown in Fig. 3 (referring to an 8 kb/s data rate). It can be observed that the same values of power efficiency correspond to higher values of spectrum efficiency. This benefit could also be expressed in terms of power efficiency improvement due to diversity for the same spectral efficiency. Link budgets relative to the return link clearly show macrodiversity benefits, in terms of terminal EIRP reduction in cases where diversity is not employed. Diversity advantage increases with actual diversity order. It should be observed that, contrary to the forward link, the power efficiency curve of Fig. 2 applies to all diversity orders.
In parallel to air interface activities ESA is presently investigating advanced multibeam payload architectures that will provide higher capacity and be able to support the higher-data-rate services required by S-UMTS.

The Far East Approach

Relevant R&D Activities in Korea

There are five different study groups in the Technical Assembly of TTA, the telecommunication standardization body in Korea, which is responsible for standardization of radio and network aspects of IMT-2000 [16]. One of these groups, SG4, is the satellite study group, responsible for standardization of the IMT-2000 satellite component. SG4 has derived the RTT specification for the satellite component from the Global CDMA II specification, which is one of two RTT candidate systems submitted to the ITU -- Radiocommunication Standardization Sector (ITU-R) by TTA. One of the driving forces in this effort was Korea Telecom (KT), who undertook a research activity centered on broadband wireless technologies for mobile satellite communications during the period 1997–1998. This particular project has been undertaken in close collaboration with Korea University, who has studied the multiple access options for the IMT-2000 satellite component and evaluated the related technologies in the LEO mobile satellite communication environment. The Wireless Multimedia Team in KT is another active task force who was responsible for the study of satellite constellation configuration and spotbeam/cell layout.
As set out in the initial requirements by ITU-R, TTA believes that the satellite component of IMT-2000 should play a key role in providing global coverage. Furthermore, it should also provide advanced services to end users complementary to the terrestrial IMT-2000 component. In this sprit, the initial approach in TTA is to design a new system which supports a wide range of voice and nonvoice services with a data rate of up to 128 kb/s, thus providing capabilities well beyond the existing GMPCS. The new system is supposed to accommodate handheld, fixed, portable, and vehicular terminals while advanced access technology is employed to support integrated services based on both circuit and packet modes.
After assessing the different multiple access options, W-CDMA has been chosen in SAT-CDMA, since it provides highly efficient spectrum utilization, simple frequency allocation management, and a high degree of commonality with most proposed RTTs for the terrestrial IMT-2000 component. SAT-CDMA implements a variable spreading factor and multicode scheme to support variable data rates while employing satellite diversity through an artificial RAKE receiver to cope with blockage due to low elevation angle and multipath fading in a mobile satellite environment. Furthermore, signal-to-interference ratio (SIR)-based closed loop power control is designed with a variable step size to adapt to relatively long propagation delay of this environment. Throughout preliminary design studies on satellite constellation in terms of beam coverage and implementation cost, a LEO at an altitude of 1600 km has been configured with eight orbit planes and six satellites/orbit plane. Under this condition, link budget analysis has been performed to derive link parameters including the average elevation angle and time visibility of satellites as a function of latitude.

Air Interface Specification for SAT-CDMA: Key Features

SAT-CDMA is based on W-CDMA with FDD mode operating at a chip rate of 4.096 Mchips/s. It adopts the same air interface as Global CDMA II, which is one of the terrestrial W-CDMA IMT-2000 RTTs proposed by TTA. As mentioned earlier for the S-UMTS air interface, exploitation of the same air interface as in the terrestrial system will allow reduction of dual-mode terminal cost and size. Therefore, various design aspects have been examined to maintain commonality while satisfying the satellite system requirements.
In SAT-CDMA, both logical and physical channels have been identified. Each radio frame is fixed to 10 ms and is composed of 16 time slots. The configuration of radio frames or timeslots varies with the symbol rate of the physical channel. SAT-CDMA will support mobile-assisted network-assisted handover. Three different types of handover are considered: interbeam handover, intersatellite handover, and inter-land earth satellite (LES) handover.
We note that the above features closely resemble those characterizing European-originated RTTs. Therefore, similar to terrestrial, there is scope for harmonizing satellite RTTs based on common W-CDMA technology and sharing commonality with terrestrial RTTs.
Simulation Results -- In the multibeam CDMA system, system capacity is governed not only by the elevation angle but also by the number of satellite beams [17]. As shown in Fig. 4, the system capacity dramatically improves by increasing the number of beams. The capacity boost is particularly noticeable by exploiting high elevation angle and/or path diversity gain.
Extensive simulation studies show that the diversity scheme is an essential element to combat the low-elevation performance degradation in the CDMA-based LEO multibeam satellite system, and furthermore warrants the high-quality and low-cost implementation as well as the system capacity [17].
One other research focus in KT is to investigate the effectiveness of closed loop transmit power control (TPC) on DS/CDMA-based mobile satellite communication systems with a LEO satellite constellation for which long round-trip delay (RTD) is a major limiting factor. In SAT-CDMA, both open and closed loop TPC schemes are employed. The initial transmit power of common physical channel is determined by the system and informed on to the LES. In general, open loop power control is useful for tracking slower variations of the received signal due to shadowing and path loss through the use of a pilot signal. Due to rather large frequency spacing (190 MHz) between forward and reverse links, however, sufficient fading correlation between forward and reverse links is not always guaranteed; thus, accurate open loop power control cannot be implemented. For this reason, open loop TPC is mainly used for initial power setting during call setup. A closed loop TPC for a dedicated physical channel is activated after initial transmission power is set based on the open loop estimate. Mobile earth stations (MESs) periodically update their transmit power based on TPC at a rate of 200 b/s. The effectiveness of the closed loop TPC scheme is negatively impacted by the satellite long RTD; consequently, it impairs fast fading compensation [18]. In SAT-CDMA, a variable step-size TPC is adopted as the means to speed up reaction to signal variation. Currently, two different step sizes of 0.25 and 1 dB are supported with 2-bit power control information, which is multiplexed into forward signaling channel and reverse link pilot channel, respectively, for reverse and forward link closed loop TPC.
The variable PC step approach was shown to be a useful means to combat a negative effect of the long RTD on the closed loop TPC.
One of the main SAT-CDMA design objectives is to provide integrated services for both real- and non-real-time services in a flexible manner. Especially, the QoS-oriented next-generation Internet technologies have been perceived as another driving force in the near future. Toward this end, different types of medium access control (MAC) protocols have been developed to support both voice and data services in packet mode. One example is the Prioritized Packet Reservation CDMA (P2R-CDMA) protocol for the uplink in a multicode CDMA system, which employs centralized frame-based slot reservation along with dynamic slot assignment in the base station using the QoS-oriented dynamic priority of an individual terminal. Simulation results for P2R-CDMA have shown the effectiveness of the proposed algorithm for QoS-oriented integrated services over existing approaches such as the ones described in [19, 20].

Future Perspectives

According to ESA-funded investigations [5, 10], third-generation mobile communication systems shall pave the way for the introduction of adaptive interference mitigation techniques at the UT level. This class of advanced demodulators will ease the cost-effective introduction of higher data rates over mobile satellites. To demonstrate the technical feasibility of these new technologies, ESA is presently funding a complete laboratory testbed [5, 14] and a single application-specific integrated circuit (ASIC) [21] implementing third-generation interference-mitigating CDMA receivers suited for S-UMTS.
Looking further beyond UMTS, the industrial competition between Asia, Europe, and America promises a difficult path toward the definition of a unique standard for future mobile systems, despite the trading benefits derived from a common worldwide standard underlined by market analysts. In this framework, the software radio concept is emerging as a potential pragmatic solution: a software implementation of the UT capable of dynamically adapting to the radio environment in which it exists. In fact, the term software radio stands for "radio functionalities defined by software" and is not only limited to the lower communication layers. The prospect of a software-based radio interface necessarily implies the use of digital signal processing (DSP) engines replacing dedicated hardware to execute, in real time, the necessary radio and network functions and algorithms. Software radio offers the flexibility to operate in multiservice/multistandard environments, without being constrained to any particular standard.
In this field, the ACTS SORT project is currently developing an advanced testbed, dedicated to UTRA and the SINUS air interface, an all-digital reconfigurable channelization and sample rate adaptation demonstrator, one of the most critical enabling functionalities before baseband signal processing.

Summary

The adoption of an international open UMTS standard encompassing both the terrestrial and satellite components will certainly realize the original IMT-2000 vision of a worldwide third-generation wireless system capable of efficient delivery of multimedia services, anytime and anywhere. Exploitation of a common W-CDMA approach for both the satellite and terrestrial UMTS components will certainly bring this eventuality much closer. In this article a collective review of the satellite W-CDMA activities within the framework of UMTS/IMT-2000 was provided. All the proposed W-CDMA-based S-UMTS air interfaces (Table 4) share a common view since they represent an approach whereby maximum commonality with terrestrial W-CDMA RTTs is achieved. Considering the ETSI and ITU guidelines jointly with limitations of satellite systems operating at S-band, the RTTs envisage a maximum service rate of 144 kb/s. There is certainly room for further harmonization among the S-UMTS proposals as well as with their emerging terrestrial counterpart.
The success of the IMT-2000 satellite component to a great extent will depend on how the IMT-2000 terrestrial market develops. What is certain, however, is that the IMT-2000 satellite complement will have a greater chance of success if a family of open and harmonized4 air interfaces become available very soon.

References
[1] ETSI ETR UMTS 12.01, "Framework for satellite integration within the UMTS."
[2] E. Buracchini et al., "Inter-segment handover implementation in the SINUS project and its integration within Rainbow project," ACTS Mobile Summit 1997, Aalborg, Denmark, Oct. 7–10, 1997.
[3] Y. Karasawa et al., "Analysis of Availability Improvement in LMSS by Means of Satellite Diversity Based on Three-State Propagation Channel Model," IEEE Trans. Vehic. Tech., Nov. 1997.
[4] B. Lyons et al., "A High Capacity Third-Generation Mobile Satellite System Design," Euro. Trans. Telecommun., vol. 9, no. 4, July/Aug. 1998.
[5] ESA Contract No. 12497/NL/97/NB, "Robust Modulation and Coding for Personal Communication Systems."
[6] P. Taaghol et al., "Satellite Diversity and its Implications on the RAKE Receiver Architecture for CDMA-Based S-PCNs," IMSC '95, Ottawa, Canada, June 1995.
[7] E. Lutz et al., "The Land Mobile Satellite Channel - Recording, Statistics and Channel Model," IEEE Trans. Vehic. Tech., vol. VT-40, May 1991.
[8] G. E. Corazza and C. Caini, "Satellite Diversity Exploitation in Mobile Satellite CDMA Systems," submitted to IEEE Wireless Commun. and Networking Conf., New Orleans, LA, Sept. 21–24, 1999.
[9] P. Taaghol, S. Nourizadeh, and R. Tafazolli, "An Advanced Power Control Scheme for CDMA-Based Satellite Communication Systems," IMSC '99, Ottawa, Canada, June 15–18, 1999.
[10] R. De Gaudenzi et al., "A Frequency Error Resistant Blind Interference Mitigating CDMA Detector," IEEE 1998 5th Int'l Symp. Spread-Spectrum Techniques and Applications, Sun City, South Africa, Sept. 1998.
[11] P. Taaghol et al., "A Real-time Dynamic Space Segment Emulator," IMSC '99, Ottawa, Canada, June 15–18, 1999.
[12] Toskala et al., "FRAMES FMA2 Wideband CDMA for UMTS," Euro. Trans. Telecommun., vol. 9, no. 4, July 98, pp. 325–35
[13] E. Buracchini et al., "SINUS Air Interface Overview," ACTS Mobile Summit 1997, Aalborg, Denmark, Oct. 7–10, 1997.
[14] G. Caire et al., "Development and Validation of a Wideband CDMA IMT-2000 Physical Layer for Satellite Applications," Proc. Int'l. Mobile Satellite Commun. Conf. '99, Ottawa, Canada, June 1999.
[15] G. Caire et al., "ESA Satellite Wideband CDMA Radio Transmission Technology for the IMT-2000/UMTS Satellite Component: Features & Performance," submitted to IEEE GLOBECOM '99, Rio De Janeiro, Brazil, Dec. 5–9, 1999.
[16] K.-J. Wee and Y.-S. Shin, "Current IMT-2000 R&D Status and Views in Korea," IEEE Commun. Mag., Sept. 1998.
[17] Y. Lee et al., "Channel Capacity of Wideband CDMA-based LEO Multi-beam Satellite System with Diversity Reception," Proc. CDMA Int'l. Conf., Seoul, Korea, 1998.
[18] Y. Lee et al., "Performance of a Closed-Loop Power Control in Mobile Satellite Communications," Proc. APCC '98, Singapore, 1998.
[19] K. Mori and K. Ogura, "An Investigation of Permission Probability Control in Reserved/Random CDMA Packet Radio Communications," PIMRC '97, May, 1997, pp. 933–37.
[20] A. E. Brand and A. H. Aghvani, "Performance of a Joint CDMA/PRMA Protocol for Mixed Voice/Data Transmission for Third Generation Mobile Communication," IEEE JSAC, vol. 14, no. 9, Dec. 1996, pp. 1698–1707.
[21] Centro TEAM and ST Microelectronics, "Multi User Interference Cancellation Demodulator," ESA Contract No. 13905/98/NL/SB.

Additional Reading
[1] M. A. N. Parks et al., "Simultaneous Wideband propagation measurement applicable to mobile satellite communication systems as L-band and S-band," AIAA '96.

Biographies
Payam Taaghol [M'96] received a B.Eng. degree in electronic and microwave communications engineering in 1993 from the University of North London, United Kingdom. He then joined the CCSR Ph.D. research program in October 1993. He is currently a lecturer in the Mobile Communications Research Group, and is working toward the completion of his Ph.D. degree. His main research interests and expertise are in CDMA-based land-mobile satellite personal and multimedia communication systems with particular emphasis on advanced receiver architecture, adaptive and predictive power control algorithms, and real-time emulation of the propagation channel. During his time with the group, he has contributed and participated in several European Community RACE II and ACTS programs as well as many industrial projects. He is currently coordinator of the SECOMS and SINUS projects at Surrey. He is also leading the development of an advanced dynamic satellite channel emulator hardware/software (based on experimental results of CCSR's wide and narrowband measurement campaigns) within the framework of SINUS, WISDOM, and ASSET.
Barry G. Evans [SM'98] received B.Sc. and Ph.D. degrees in electrical engineering and microwave systems from the University of Leeds in 1965 and 1968, respectively. He was British Telecom Lecturer-Reader in Telecommunications Systems at the University of Essex from 1969 to 1983. In 1983 he was appointed to the Alec Harley Reeves Chair of Information Systems Engineering at the University of Surrey, and in 1990 became the first director of the postgraduate Center for Satellite Engineering Research, which he built up to about 150 researchers and a spinoff company, Surrey Satellite Technology Ltd. Since 1996 he has been director of the new Center for Communication Research Group at Surrey which is now 90 full-time researchers. He is a Fellow of the Royal Academy of Engineers in the U.K, a Fellow of the IEE and a senior member of the AIAA. He is editor of the International Journal of Satellite Communications and the author of three books and over 300 journal papers. He is currently technical advisor to the Director General of OFTEL, the U.K. telecommunications regulation body, and a member of the U.K. Government Foresight think tank on future IT and communications. He also sits on U.K. ITU and MOD committees. In 1997 he was one of four finalists for the U.K. Royal Academy McRoberts Engineering Award, and in 1998 was awarded the DTI President's Prize in Engineering for CCSR's research collaboration with industry. He is director of SATCONSULLTA and SPECA Ltd.
Enrico Buracchini received, with full marks, a degree in eectronic engineering from the University of Bologna in October 1994. In December 1994 he was employed in the Mobile Services Division of CSELT (R&D labs of TELECOM ITALIA group) as a research engineer. His activity concerns the study of multiple access methods (TDMA, CDMA, SDMA and SWradio), demodulation, and power control techniques for terrestrial and satellite mobile communications systems. He was involved in the framework of the European research programs COST 227, COST 231, and RACE II SAINT. He is now involved in the ACTS SINUS and SORT projects, leading a WP, and in the ESPRIT SLATS project, and COST 252 and 259 activities. He is part, since 1996, of the Italian delegation to the ITU-R TG8/1 standardization group. He is now senior researcher and project manager in CSELT of the R&D project "New techniques and methodologies for radio mobile systems," dedicated to software radio and smart antennas.
Riccardo De Gaudenzi [SM-97] received a Doctor Engineer degree (cum laude) in electronic engineering from the University of Pisa, Italy, in 1985. From 1986 to 1988 he was with the European Space Agency (ESA), Stations and Communications Engineering Department, Darmstadt, Germany, where he was involved in satellite telecommunication ground systems design and testing. In particular he followed the development of two new ESA satellite tracking systems. In 1988 he joined ESA's Research and Technology Center (ESTEC), Noordwijk, The Netherlands where he holds the position of senior telecommunication engineer in the Electrical Systems Department. He has been responsible for the definition and development of advanced satellite communication systems for fixed and mobile applications. He is also involved in the definition of the future European Navigation System. In 1996 he spent one year with Qualcomm Inc., San Diego, California, in the Globalstar LEO project system group under an ESA fellowship. His current interest is mainly related to efficient digital modulation and access techniques for fixed and mobile satellite services, synchronization topics, adaptive interference mitigation techniques, and communication systems simulation techniques.
Gennaro Gallinaro received a Dr. Ing. degree in electronic engineering (magna cum laude) from University of Rome in 1979. Until 1989 he worked in Telespazio, Rome, where he was involved in satellite communication system planning and design. Now he is with Space Engineering, Rome. His main interests are presently in digital communication system analysis and simulation, and digital signal processing.
Joon Ho Lee received B.S. and M.S. degrees, both in electronics engineering, from Korea University, Seoul, in 1987 and 1989, respectively. After researching FH tactical radio at Goldstar Electric Co. Ltd., he has worked for Korea Telecom (KT) as a senior researcher since 1990. He participated in the ITU-R TG8/1 15th meeting held in the United Kingdom as a national delegate. His research interests are in mobile satellite communication systems and wireless communication systems.
Chung Gu Kang [M] received a B.S. degree in electrical engineering from the University of California, San Diego in 1987, and M.S. and Ph.D. degrees, both in electrical and computer engineering, from the University of California, Irvine, in 1989 and 1993, respectively. Since March 1994, he has been with the School of Electrical Engineering at Korea University, Seoul, as an assistant and associate professor. His research interests are in mobile satellite communication and wireless multimedia communication systems development, including wireless ATM.