1 . Scenarios and Requirements for Next Generation Mobile Radio

The future success of the mobile communications revolution strongly depends on increasing the data rate available to the mobile user. This will enable value added services, that are not possible with current state of the art mobile wireless radio systems. While an increase in data rate could be achieved by combining techniques in the form of multi mode terminals (e.g. UMTS and wireless LAN), this has the disadvantage that for each new scenario another mode has to be added. The combination may also not be suitable for the mobility requirements of wireless radio systems. This paper proposes using a broadband multi-carrier based air interface, including discussion of the enabling technologies for multi-carrier mobile systems and discussion of technology trends for future research. The multi-carrier technology offers the desired high data rates for the 4G mobile environments and also has advantages for spectral efficiency and low-cost implementation. WWRF/WG4/Subgroup on New Air Interfaces White Paper Version 1.2 6/27/2002 Page 3 of 10 1. Scenarios and Requirements for Next Generation Mobile Radio Systems The concept of “anytime, anywhere and anyone” was realized by current mobile radio systems for voice communications. The UMTS network will offer 2 Mbit/s maximum data rate and enhance this concept to realize not only voice, but also data communication such as WWW and e-mail at the vehicular speeds. Thus, 4G should be designed to offer significantly higher bit rates in a vehicular environment to achieve this concept more effectively [1]. Higher data rate is needed considering some scenarios in future, e.g. highspeed download electric book and newspaper with PDA, intranet access from outside and video phone and conference with high reality. The terminology 4G mobile radio system is used in this paper as a general description of a future wireless system. It is based on a completely new air interface and should enable new broadband services and scenarios with high mobility and should also cover today’s 3G and wireless LAN services. So this term can be seen as the successor of both 3G as well as current wireless LAN systems. Considering the market trends, mobile communication traffic in 2010 and 2015 increases 5.4 and 23 fold, respectively, because multimedia traffic will increase 40% per year [1]. 4G systems should accommodate this dramatically increasing amount of multimedia traffic. Therefore, enhancing system capacity as well as achieving a higher bit rate transmission is an important requirement for the 4G system. The main task is to investigate and develop a new broadband air interface which can deal with high data rates in the order of 100 Mbit/s, high mobility and high capacity. Since the available frequency spectrum is limited, high spectral efficiency is the major task of 4G mobile radio systems. The asymmetric nature of traffic in the downlink and the uplink should be considered when designing the new air interface. Another important target of the 4G new air interface is the ability to provide efficient support for applications requiring simultaneous transmission of several bits of streams with possibly different Quality of Service (QoS) targets. This research should complement the various approaches relying on the increase of the throughput capitalizing on the higher capacity provided by using a MIMO channel. The goal is to renew the yet becoming classical OFDM transmission scheme and challenge its domination in standards bodies. 2. Broadband Multi-Carrier Solution Most of the standards developed for wireless high rate data transmission in the recent years have been based on multi-carrier modulation (i.e. orthogonal frequency division multiplexing (OFDM) [2],[3]). Among the earliest of these were the digital audio broadcasting standard (DAB) [4] and the terrestrial digital video broadcasting standard (DVB-T) [5]. Work on these started in the 1990s and was followed by the wireless local area network standards (WLANs) IEEE 802.11a, ETSI HIPERLAN/2 [6] and MMAC and the recently developed wireless metropolitan area network (WMAN) standards IEEE 802.16. WWRF/WG4/Subgroup on New Air Interfaces White Paper Version 1.2 6/27/2002 Page 4 of 10 In the development of these systems, the choice of multi carrier modulation for high data rate systems is a consequence of its: • high spectral efficiency due to a nearly rectangular frequency spectrum; • reduced receiver complexity due to simple elimination of inter symbol interference (ISI). This is efficiently achieved by cyclically extending each OFDM symbol and thus enabling very high rate data transmission in multipath channels; • simple implementation of OFDM by using the Fast Fourier Transform (FFT); • low complexity multiple access schemes exist like OFDMA, MC-TDMA, MCCDMA, MC-DS-CDMA, and TFL-CDMA exist [7][8][9][10][14][20] from which some have already been applied in recent wireless standards; • high flexibility in terms of subcarrier allocation strongly supports data rate and service adaptation; • very low transmission power and non-bursty transmission modes keeping radiation effects at very low level. These many advantages of OFDM have already been proven in standardized systems and demonstrators and this is a strong motivation to build on this success in the design of a new broadband air interface for a 4G mobile radio system. The research in OFDM based multiple access was at its beginnings when the important decisions in which direction the 3G mobile radio standards will go has to be decided and, thus, OFDM did not become part of 3G systems. However, this situation has changed and today many research has been carried out on multi-carrier based multiple access schemes and the basic knowledge on these new multiple access concepts is existing to apply it for broadband 4G mobile radio systems. There is also a scope for future research to focus on a sophisticated design of the upand downlink in order to further address the new 4G system requirements of high data rates, high spectral efficiency and low power consumption limits. The following research topics should be addressed for the design of a new multi-carrier 4G mobile radio system: • Which multiple access schemes achieve the highest spectral and system efficiency for the uplink and the downlink? • How can the demands on flexibility in data rate and robustness best be met? To what extent is flexible subcarrier allocation feasible in multi-carrier systems? • Which multiple access scheme can best cope with the significant power consumption limits? This topic also addresses the PAPR issue, which has to be considered in multicarrier systems. • What could be gained in system capacity when using a more complex synchronous uplink like in GSM compared to a quasi-synchronous or completely asynchronous uplink? • Is FDD or TDD the right decision for asymmetric traffic? • Which frequency reuse factor is achievable in multi-carrier based cellular systems and what is the appropriate cell size? WWRF/WG4/Subgroup on New Air Interfaces White Paper Version 1.2 6/27/2002 Page 5 of 10 • How can the system efficiency be improved by making channel state information available to the system and the transmitter through a feedback channel? This topic also addresses channel prediction, adaptive coding/modulation, and fast scheduling between users. 3. System Enabling Technologies in Multi-Carrier Multiple Access System Scalability The 4G mobile radio systems should be scalable with respect to bandwidth and cell site deployment (macro/micro/pico). These radio systems should be also scalable in terms of the service requirement, including both symmetrical and asymmetrical services. To meet this objective, the physical layer design including the slot/frame structure hierarchy, should be adaptable to the different bandwidths, deployment scenarios, traffic conditions and a wide range of future application requirements. Asymmetrical Service and Access Schemes The uplink and downlink for the 4G mobile radio system may utilize different techniques. For example, the downlink may use OFDM and the uplink MC-DS-CDMA in order to optimize both the spectral efficiency and the mobile power consumption. The 4G mobile radio system should be suitable for operation in unpaired spectrum. The 4G mobile radio system should also support multi-cast and broadcast services. High Speed Data Rate Coverage and Uniform Distribution The 4G mobile radio system should deliver uniform data rate coverage and ubiquity of service. It is important to study techniques to achieve uniform data rate access across both radio system and the access network. These studies should include research on the access architecture and control for factors such as data rate, packet delay and power control. Soft Hand-off and Macro/Network Diversity Soft hand-off is a proven technique for mobile systems. It is a fundamental technology to improve the coverage and reduce the outage. It is important to develop the soft hand-off technique for the 4G radio systems and to exploit macro/network diversity. Application to High-speed Mobility To support the high-speed mobility application for the 4G radio system, several fundamental aspects of the OFDM physical layer need to be studied including the pilot structure and adaptation schemes for coding and modulation. MIMO-OFDM and Smart Antennas The recent developments for space-time coding and smart antennas should also be taken into account in the design of the 4G mobile radio system and air interface. The combination of MIMO with OFDM is an enabling technology to fundamentally improve the spectral efficiency of the 4G mobile radio system. OFDM is well suited for the implementation of MIMO technology. The advantage of OFDM over CDMA to implement MIMO technology is significant. The effects of MIMO schemes on the system capacity of an OFDM based air interface are of high importance for the design of the detection and decoding scheme. It is important to study techniques to combine MIMO WWRF/WG4/Subgroup on New Air Interfaces White Paper Version 1.2 6/27/2002 Page 6 of 10 and smart antenna schemes to allow the antenna technology to be optimized and closely integrated into the 4G air interface. Co-channel Interference Cancellation/Avoidance In mobile cellular networks a frequency reuse of one is an important factor in the system capacity. It is important to study techniques for interference mitigation and cancellation that will increase the spectral efficiency. It is also important to study techniques for avoiding interference through e.g. fast scheduling between users or new innovative ways of combining multi-carrier and CDMA such that interference is limited or avoided [20]. Medium Access Control (MAC) Design The new air interface should be designed to carry different types of traffic including real time traffic and non-real time traffic. This will require the design of an efficient MAC that maximizes the system throughput and minimizes the overhead. The MAC design should carry in addition to data, control information that requires more robustness to errors. With the vision that the 4G system will have different types physical layer concepts, the MAC should be generic to support the different physical layers. Radio Resource Management With the new air interface supporting many high data rate users, resources should be managed carefully. Different schemes can be studied to maximize the system throughput/capacity in an OFDM system. These includes but not limited to, OFDMA techniques where frequency selectivity is exploited, fast scheduling techniques exploiting time and spatial dimensions in addition to the frequency dimension. Also, adaptive modulation/coding techniques over all or a group of subcarriers should be studied. Studies comparing feedback and feedforward schemes are required to enhance the system throughput without creating excessive overhead. Channel prediction Large spectrum and system efficiency improvements are possible by making channel state information available to the system and the transmitters [21][22]. This however requires that channel prediction algorithms that can predict the channel impulse response well into the future can be designed and that these have reasonable good performance [23]. Furthermore, adaptive coding/modulation schemes, fast scheduling algorithms, and multiple access schemes that can use the information (predicted channel impulse response and the accuracy of it) provided by the channel prediction algorithm must be developed. A feedback channel is needed in the system and this needs to be designed such that it does not consume too much of the bandwidth resources. 4. Technology Trends in Multi-Carrier Multiple Access Besides the general system aspects addressed in the previous section, the focus of future research should also include the technical aspects of the individual access schemes. An overview of the technology trends summarizing important research results is given here followed by the identification of future research topics. WWRF/WG4/Subgroup on New Air Interfaces White Paper Version 1.2 6/27/2002 Page 7 of 10 Coding versus Spreading in Future OFDM Systems It is well known that current OFDM systems require some forward error correction (FEC) coding in order to achieve excellent performance results in fading channels. FEC coding can be classical convolutional coding as it is used in DAB, DVB-T, IEEE 802.11a or HIPERLAN/2 or it can be spreading or a combination of both. It is of importance to investigate the trade-off between channel code rate, spreading code length, and symbol allocation on the channel for typical 4G mobile radio scenarios. With channel state information available on the transmitting side of the system, completely new designs are possible [20]. The fading can to a large extent be counteracted already in the transmitter and diversity (through e.g. channel coding) becomes less important. Alternatively, it is possible to exploit the redundancy in the prefix or postfix [17]. Singleand Multi-user Detection in MC-CDMA and MC-DS-CDMA Investigations have shown that the most promising single user detection technique is MMSE equalization. It can be implemented in a suboptimum version with very low performance degradation and with similar complexity to zero forcing or equal gain combining. Multi-user detection techniques can significantly improve the system performance at the expense in complexity. Various interference cancellation schemes and block linear equalizers with and without feedback have been investigated and proposals have been made, how the optimum maximum likelihood detector can be realized with reasonable complexity in MC-CDMA systems. Results for the upand downlink have been presented. In future research, it is important to fairly compare all these techniques, to consider their imperfections and to investigate their robustness against different kinds of interference like inter cell interference and near-far resistance. Iterative Detection and Decoding Strategies The iterative combination of detection and decoding has gained increasing interest recently resulting in novel detection schemes like soft-interference cancellation [10]. These schemes exploit reliability information about the coded bits from the outer decoder and feed this back to the detector. The results indicate that a better performance can be obtained with soft interference cancellation than with separated maximum likelihood detection followed by classical convolutional decoding. These promising schemes have to be adapted to the requirements on 4G systems. The benefits of these schemes have to be proven also for system imperfections such as imperfect power control, imperfect channel estimation and interference from neighboring cells. Moreover, in order to improve the multi-carrier detection scheme iterative demodulation algorithms (turbo demodulation) of bit interleaved higher order coded modulation enables significant performance enhancement over standard bit metric derivations. Channel Estimation Pure OFDM systems can easily apply differential modulation (as used in DAB). In the case of multiple access schemes like MC-CDMA, the coherent detection with pilot symbol assisted channel estimation has gained more interest due to its higher user capacity. Efficient two-dimensional channel estimation concepts have been developed WWRF/WG4/Subgroup on New Air Interfaces White Paper Version 1.2 6/27/2002 Page 8 of 10 [12], which can significantly reduce the overhead due to pilot symbols by exploiting the correlations of the mobile radio channel in both the time and frequency dimension. These schemes are especially of interest for downlink systems. For the 4G system, concepts have to be developed for an efficient MC-CDMA uplink channel estimation that can cope with high numbers of active users. Alternatives or extensions of pilot symbol aided channel estimation schemes including blind or semi-blind estimation algorithms [13][18] that reduce the overhead due to pilot symbols have gained high interest in the past years. The suitability of these schemes for 4G broadband systems should also be investigated in detail. Synchronization OFDM inherently offers the possibility to use the guard interval for synchronization [15] and thus can reduce or even completely avoid additional overhead for reference symbols required for synchronization. Investigations are necessary to show whether these algorithms are sufficient to perform acquisition and tracking in a 4G system or whether additional measures are necessary [16]. PAPR Reduction Multi-carrier signals have a non constant envelope which reduces the efficiency of the power amplifiers in the transmitter and due to the nonlinear amplifier characteristics distorts the transmitted signal and enhances the out-of-band power. A multitude of measures have been proposed to reduce the peak-to-average power ratio (PAPR) or to apply predistortion to increase the power efficiency of the transmitter. It is important to study the PAPR techniques to conclude which methods best fit the requirements in the uplink and in the downlink. Thus, deriving digital strategies either in the frequency or time domain conducting to a reduction of a peak to average power allows a significant decrease of the backoff for the power amplifier [19]. This would impact a lot the battery life of the wireless terminals. More generally, the study of digital means for reducing the OFDM constraints on the RF front-end and transceiver architectures that targets lower power operation will be key for reaching high data rate and preserving true wireless operation. 5. Conclusions The future demands on data rate and mobility require the design of completely new air interfaces for the next generation of mobile radio systems. The multi-carrier technology has shown its suitability for very high rate systems in many digital broadcasting and wireless LAN standards and seems to be an excellent candidate to fulfill the requirements on 4G mobile radio systems. Future research is expected to show that multi-carrier will be a superior technology for 4G. We see great potential e.g. in diversity schemes and iterative receiver algorithms which consider the interaction between components which operated in an isolated mode today. An extensive cooperation between research groups is also very important encompassing but not limited to Smart Antennas, Spectrum Issues, Channel Modeling and Propagation Measurements. WWRF/WG4/Subgroup on New Air Interfaces White Paper Version 1.2 6/27/2002 Page 9 of 10 We intend to shape the next generation mobile radio systems by contributing to European research and development of future wireless systems, standards bodies and research forums. Following institutions declared their intention for research in Broadband Multi-Carrier Based Air Interfaces (alphabetical):Carleton University (Canada), CEA-LETI (France), Chalmers University of Technology(Sweden), DoCoMo Eurolabs (Germany), Fujitsu Laboratories of Europe Ltd (UK),German Aerospace Center (DLR) (Germany), Mitsubishi Electric ITE (France), MotorolaLabs (France), Nortel Networks (UK), Royal Institute of Technology (KTH) (Sweden),Telefónica (Spain), Technical University of Ilmenau (Germany), Uppsala University(Sweden) and VTT Electronics (Finland). References[1] T. Otsu, I. Okajima, N. Umeda and Y. Yamao, “Network Architecture for MobileCommunications Systems Beyond IMT-2000”, IEEE Personal Communications, pp.31—37, October, 2001. [2] B. R. Saltzberg. Performance of an Efficient Parallel Data Transmission System.IEEE Trans. on Communications, 15(6):805-811, December 1967. [3] J.A.C. Bingham, “Multicarrier modulation for data transmission: An idea whose timehas come, “ IEEE Communications Magazine, vol. 28, pp. 5-14, May 1990. [4] ETSI ETS 300 401, Radio Broadcasting Systems; Digital Audio Broadcasting (DAB)to mobile, portable and fixed receivers. European Standard (Telecommunicationsseries), Valbonne, France, February 1995. [5] ETSI EN 300 744, Digital Video Broadcasting (DVB); Framing structure, channelcoding and modulation for digital terrestrial television. European Standard(Telecommunications series), Valbonne, France, July 1999. [6] ETSI TS 101 475, Broadband Radio Access Networks (BRAN); HIPERLAN Type 2;Physical (PHY) layer. European Standard (Telecommunications series), Valbonne,France, November 2000. [7] K. Fazel and L. Papke. On the Performance of Convolutionally-CodedCDMA/OFDM for Mobile Communication System. In proceedings of IEEE PIMRC,pages 468-472, Yokohama, Japan, September 1993. [8] European Transactions on Telecommunications (ETT), Special Issues on Multi-Carrier Spread Spectrum, July/Aug. 1999, Nov./Dec. 2000 and Sept./Oct. 2002. [9] K. Fazel and S. Kaiser (Eds.), Multi-Carrier Spread Spectrum & Related Topics.Boston: Kluwer Academic Publishers, 2000 and 2002. [10] S. Kondo and L.B. Milstein, “Performance of multicarrier DS CDMA systems,”IEEE Transactions on Communications, vol. 44 pp. 238-246, Feb. 1996. WWRF/WG4/Subgroup on New Air InterfacesWhite PaperVersion 1.2 6/27/2002Page 10 of 10[11] S. Kaiser and J. Hagenauer, “Multi-carrier CDMA with iterative decoding andsoft-interference cancellation,” in Proceedings IEEE Global TelecommunicationsConference (GLOBECOM’97), Phoenix, USA, pp. 6-10, November 1997. [12] P. Höher, S. Kaiser, P. Robertson, “Pilot-symbol-aided channel estimation in timeand frequency,” In Proc. IEEE Global Telecommunications Conference(GLOBECOM ’97), Communication Theory Mini-Conference, Phoenix, USA, pp. 90-96, Nov. 1997. [13] J. Heath, G. Giannakis, “Exploiting input cyclostationarity for blind channelidentification in OFDM systems,” IEEE Transactions on Signal Processing, Vol. 47,No. 3; pp. 848-856, March 1999. [14] S. Abeta, H. Atarashi, M. Sawahashi and F. Adachi, “ Performance of coherentmulti-carrier/DS-CDMA and MC-CDMA for broadband packet wireless access,”IEICE Transactions on Communications, Vol. E84-B, No. 3, pp. 406-414, March2001. [15] M. Sandell, J.-J. van de Beek and P.O. Börjessen, „Timing and frequencysynchronization in OFDM systems using the cyclic prefix,“ in Proceedings 1995International Symposium on Synchronization, Essen, Germany, pp. 16-19, Dec. 1995. [16] T.M. Schmidl and D.C. Cox, “Robust frequency and timing synchronization forOFDM,” IEEE Transactions on Communications, vol. 45, no. 12, pp. 1613-1621,Dec. 1995. [17] B. Muquet, M. de Courville, P. Duhamel, and G.B. Giannakis. OFDM withTrailing Zeros versus OFDM with Cyclic Prefix: Links, Comparisons andApplication to the HiperLAN/2 System. In Proceedings of the Int. Conf. onCommunications, volume 2, pages 1049-1053, New-Orleans, USA, June 2000. [18] B. Muquet, M. de Courville, P. Duhamel, and V. Buzenac. A Subspace basedblind and semi-blind channel identification method for OFDM systems. In Proc.ofSPAWC’99, pages 170-173, May 1999. [19] J. Tellado-Mourelo. Peak to average power reduction for multicarriermodulation. PhD thesis, Stanford University, September 1999. [20] A. Persson, T. Ottosson, E. Ström, “Time-Frequency Localized CDMA forDownlink Multi-Carrier Systems,“ in Proc. of ISSSTA’2002, Sept. 2002, to appear. [21] S.T. Chung, A.J. Goldsmith, “Degrees of Freedom in Adaptive Modulation: AUnified View,” IEEE Trans. On Communications, vol. 19, no. 9, pp. 1561-1571,Sept. 2001. [22] T. Ottosson, A. Ahlén, A. Brunstrom, M. Sternad, A. Svensson, “A 4G IP BasedWireless System Proposal,” in Proc. of RVK’2002, June 2002, to appear. [23] T. Ekman, Prediction of Mobile Radio Channels. Lic. Thesis, Uppsala University,Sweden, Dec. 2000.