On DS/CDMA for wireless multirate applications Ann-Louise Johansson, Tony Ottosson and Arne Svensson Chalmers University of Technology Department of Information Theory S-412 96 Gothenburg Sweden phone: +46 31 772 5188, +46 31 772 5189, +46 31 772 1751 fax: +46 31 772 1748 e-mail:{Anne.Johansson, Tony.Ottosson, Arne.Svensson}@It.Chalmers.Se Abstract: In this paper multirate DS/CDMA systems for multimedia applications are discussed. Several modulation schemes for handling multiple bit rates in DS/CDMA are described. An overview of multiuser detectors in general and successive interference cancellation (IC) schemes for multirate schemes in particular is given. Numerical bit error probability results are summarized for some examples.

Introduction Second generation digital mobile telephone systems have now been in public use for several years. Research is now focusing on third generation systems. These are mostly referred to as Personal Communications Networks (PCN), Universal Mobile Telecommunication Systems (UMTS) or Future Public Land Mobile Telecommunication Systems (FPLMTS) [1]-[4]. Speech will still be an important application in these systems, but a wide range of other services will also be included. Some of these services are likely to be very bit rate demanding; examples are video, image and data transmission. These services are sometimes referred to as multimedia applications.

Multirate modulation schemes in DS/CDMA In order to design these new systems, the air interface must be able to handle multiple data rates. We expect rates on the order of 10 kbps up to around 2 Mbps to be demanded in the near future. CDMA is a very strong candidate for such an air interface, especially due to it universal frequency reuse which makes high capacity feasible [5]. In this paper we will focus on DS/ CDMA schemes that support multiple data rates. Examples are multimodulation (mixed modulation), where signalling constellations of various sizes are used to obtain different rates, and multicode (multichannel, parallel channels), where a high bit rate service of a user is demultiplexed into several low rate data streams, which are transmitted in parallel on several codes. Other schemes that will be briefly discussed are multiple chip rate [6], multiple processing gain and multicode with parallel combinatory spread spectrum (PC/SS) [7]. The interested reader is referred to [6] to [9] for more details on these multirate schemes. We find that multiple chip rate and multicode schemes obtain almost identical bit error rate performance, while the performance of high rate users in a multimodulation scheme is reduced due to their use of large constellation sizes. An advantage with multicode is that the symbol periods are of equal length for all rates, which simplifies the code design. Since each high rate user in a multicode scheme transmits on several codes in parallel, the transmitted signal will have large envelope variations which requires rather linear amplifiers. From this point of This work was supported by the Swedish National Board for Industrial Technical Development, NUTEK, under contract 9303363-2.

1 Invited paper to Radiovetenskaplig konferens RVK’96, June 1996, Luleå, Sweden

view, a multiple chip rate scheme has an advantage, but on the other hand high rate users in a multiple chip rate scheme use a very small spreading factor which make them susceptible to external interference. In conclusion we find that a multicode scheme is preferred when the ratio between the largest and smallest data rate does not become too large, otherwise a combination of multicode and multimodulation should be used.

Multiuser detectors in DS/CDMA The conventional detector is composed of a filter matched to the desired user’s signal, which is the optimal structure for a single user channel corrupted only by additive white Gaussian noise (AWGN) [10]. The presence of a number of users in the system often introduces multiple access interference (MAI), which may lead to an irreducible error probability [11] in the conventional detector, unless all users are orthogonal at the receiver. Typically, MAI arise in an asynchronous system when the users’ signals are received with different time lags, which usually increases the cross-correlation between the spreading sequences. The optimum multiuser detector, on the other hand, consists of a bank of filters matched to each of the users’ signals combined with a decision device that performs maximum likelihood decoding [12]. This detector is, however, very complex to implement. Another important issue for CDMA systems, in addition to the MAI, is the near-far problem [13]. In the reverse link (mobile to base station), the signals are received with different powers. The currently most considered solution to this problem is stringent power control [15], though over the last few years a lot of attention has been given to the area of less complex (compared to the optimum multiuser detector) multiuser detectors [11]. These have the prospect of both mitigating the near-far problem and cancelling the MAI. The first major contribution was the so called decorrelator detector [14], which is near-far resistant. Decisions are made in a subspace, which is orthogonal to all the interference, but this has the drawback that the noise is enhanced. One subclass of multiuser detectors is the class of detectors that employ interference cancellation (IC) schemes. The basic principle is to reduce the interference at the detector without noise enhancement by in principle estimate each user’s interfering signal and subtract these estimates from the received signal before detection. The structure of the IC schemes differs depending on how the users are detected and cancelled. In [16]-[18] the detection and cancellation is done in parallel for all the users while in [19]-[23] the detection and cancellation is done successively (or serially), one user after the other. There are also hybrid schemes where the detection and cancellation is done both successively and in parallel [24]. The proposed methods also differ in the way an interfering signal is estimated, some use channel estimation in combination with previous decisions, while others use matched filter outputs directly to estimate the interfering signal. Most work on simplified multiuser detectors have considered single rate systems. The only exception we know of is our own work and a recent contribution by Saquib et. al. [25], where two versions of the decorrelation detector is proposed for a dual rate multiple chip rate scheme. In this paper we will shortly describe various multiuser detectors, but the main emphasis is put on successive IC for multirate systems of the type discussed above. We consider a scheme which is a generalization of the Patel-Holtzman approach for BPSK [19]. This scheme has been generalized to any level of rectangular lattice type QAM, multicode type of schemes and into a multistage scheme, and it is now also capable of handling multipath fading channels [22], [23]. Initial stages of soft decision IC where channel estimates are not needed are combined with final stages of decision directed IC employing channel estimates. This combination proves to be very powerful in multipath fading channels. Analytical bit error probability results are obtained using a Gaussian approximation of the 2

interference. For a single path Rayleigh fading channel, these results show that with random or Gold codes of length 127 in a multimodulation scheme composed of BPSK, QPSK and 16QAM, the performance degradation (average bit error probability) after three stages of soft decision IC at 10-3 compared to a single user BPSK system is less than 2 dB in a system with 70 users. The degradation for BPSK and QPSK users is even smaller while the degradation for the 16-QAM users are somewhat larger. Furthermore, for the same parameters, the error floor due to residual interference (not yet cancelled) is well below 10-4 for all rates. Simulated bit error probability have also been obtained. These results show that the analytical bit error probability is optimistic for large signal to noise ratios but rather good for small signal to noise ratios. In a system with 20 BPSK, 10 QPSK and 5 16-QAM users with Gold sequences of length 127, the performance degradation compared to a single user BPSK scheme is less than 1 dB at an error probability of 10-2 after 2 stages. The irreducible error floor is, however, above 10-3 after 2 stages but may be reduced to well below 10-4 after 5 stages of IC. The bit error probability of a multicode scheme with 15 QPSK users, each transmitting on two codes, is slightly smaller. On a two path, equal strength, Rayleigh fading channel with Gold codes of length 127 and 35 users in a multimodulation scheme composed of BPSK, QPSK and 16-QAM, the bit error probability with four stages of soft decision IC followed by one stage of decision directed IC using perfect channel estimation is degraded about 2 dB compared to the single user BPSK bound. The difference between analytical results based on a Gaussian approximation and simulated results is well within 1 dB. In a multicode scheme with 15 QPSK users, each transmitting on two codes, the single user BPSK bound is almost reached with perfect channel estimates, while a degradation of about 1 dB occurs when pilot symbol channel estimation is used. Again, the advantage of multicode is that all users obtain the same bit error probability independent of their rate, in contrast to multimodulation where the bit error probability of the high rates users is increased. The advantage of both multimodulation and multicode over multiple processing gain is that large spreading gains may be used also for high data rates in channel bandwidths of only a few MHz. More details on the discussed schemes may be found in [26] and [27], which were submitted (and hopefully accepted) to this same conference.

References: [1] D. C. Cox, “Wireless Personal Communications: What is it?,” IEEE Personal Communications, Vol. 2, No. 2, pp. 20-35, April 1995. [2] Special Issue on the European Path Towards UMTS, IEEE Personal Communications, Vol. 2, No. 1, Febr. 1995. [3] M. H. Callendar, “Future Public Land Mobile Telecommunication Systems,” IEEE Personal Communications, Vol. 1, No. 4, pp. 18-22, Fourth Quarter 1994. [4] V. O. K. Li and X. Qiu, “Personal Communication Systems (PCS),” IEEE Proceedings, Vol. 83, No. 9, pp. 1210-1243, Sept. 1995. [5] A. J. Viterbi, “The Orthogonal-Random Waveform Dichotomy for Digital Mobile Personal Communication,” IEEE Personal Communications, Vol. 1, No. 1, pp. 18-24, First Quarter 1994. [6] T.H. Wu and E. Geraniotis, “CDMA with Multiple Chip Rates for Multi-Media Communications”, Proceedings CISS’94, Conference on Information Sciences and Systems, Princeton, New Jersey, 1994, pp. 992-997. [7] S. Sasaki, H. Kikuchi, J. Zhu, and G. Marubayashi, “Multiple Access Performance of 3

[8]

[9]

[10] [11] [12] [13] [14]

[15] [16]

[17]

[18]

[19]

[20]

[21]

[22]

Parallel Combinatory Spread Spectrum Communication Systems in Nonfading and Rayleigh Fading Channels,” IEICE Trans. on Communications, Vol. E78-B, No. 8, August 1995 pp. 1152-1161. T. Ottosson, “Multirate Schemes and Multiuser Decoding in DS/CDMA Systems,” Technical Report no. 214L, Dept. of Information Theory, Chalmers Univ. of Techn., ISBN 91-7197-217-X, November 1995 (Licentiate Thesis). T. Ottosson and A. Svensson, “Multi-Rate Schemes in DS/CDMA Systems,” Proceedings VTC’95, 1995 IEEE 45th Vehicular Technology Conference, Chicago, Illinois, 1995, pp. 1006-1010. J.G. Proakis, Digital Communications, 3rd ed, McGraw-Hill, 1995. A. Duel-Hallen, et al., “Multiuser detection for CDMA systems,” IEEE Personal Communications, Vol. 2, No. 2, pp. 46-58, 1995. S. Verdú, “Minimum Probability of Error for Asynchronous Gaussian Multiple-Access Channels,” IEEE Trans. on Information Theory, Vol. IT-32, No. 1, pp. 85-96, Jan. 1986. R. Lupas and S. Verdú, “Near-far resistance of multiuser detectors in asynchronous channels,” IEEE Trans. on Communications, Vol. COM-38, pp. 497-507, April 1990. R. Lupas and S. Verdú, “Linear Multiuser Detectors for Synchronous Code-Division Multiple-Access Channels,” IEEE Trans on Information Theory, Vol. 35, No. 1, pp. 123136, Jan. 1989. K. S. Gilhousen, et al., “On the Capacity of a Cellular CDMA System,” IEEE Trans. on Vehicular Technology, Vol. 40, No. 2, pp. 303-311, May 1991. M. Varanasi and B. Aazhang, “Multistage Detection in Asynchronous Code-Division Multiple Access Communications,” IEEE Trans. on Communications, Vol. COM-38, No. 4, pp. 509-519, April 1990. S. Striglis, et al., “A Multistage RAKE Receiver for Improved Capacity of CDMA Systems,” Proceedings VTC’94, 1994 IEEE 44th Vehicular Technology Conference, Stockholm, Sweden, June 1994, pp. 789-793. Y.C. Yoon, et al., “A Spread-Spectrum Multiple Access System with Cochannel Interference Cancellation for Multipath Fading Channels,” IEEE Journal on Sel. Areas in Communications, Vol. 11, No. 7, pp 1067-1075, Sept. 1993. P. Patel and J. Holtzman, “Analysis of a Simple Successive Interference Cancellation Scheme in a DS/CDMA System,” IEEE Journal on Sel. Areas in Communications, Vol. 12, No. 5, pp. 796-807, June 1994. M. Ewerbring, et al., “CDMA with Interference Cancellation: A Technique for High Capacity Wireless Systems,” Proceedings ICC’93, 1993 IEEE International Conference on Communications, Geneva, Switzerland, May 1993, pp. 1901-1906. K. Jamal and E. Dahlman, “Multi-Stage Serial Interference Cancellation for DSCDMA,” Proceedings VTC‘96, 1996 IEEE 46th Vehicular Technology Conference, Atlanta, Georgia, April 1996. A. Johansson and A. Svensson, “Multi-Stage Interference Cancellation in Multi-Rate DS/CDMA Systems,” Proceedings PIMRC‘95, 1995 6th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, Toronto, Canada, Sept. 1995, pp. 965-969.

[23] A. Johansson and A. Svensson, “Multistage Interference Cancellation in Multirate DS/ CDMA on a Mobile Radio Channel”, Proceedings VTC‘96, 1996 IEEE 46th Vehicular Technology Conference, Atlanta, Georgia, April 1996. 4

[24] Y. Li and R. Steel, “Serial Interference Cancellation Method for CDMA,” Electronics Letters, Vol. 30, pp. 1581-1583, Sept. 1994. [25] M. Saquib, R. Yates, and N. Mandayam, “Decorrelating Detectors for a Dual Rate Synchronous DS/CDMA System,” Proceedings VTC‘96, 1996 IEEE 46th Vehicular Technology Conference, Atlanta, Georgia, April 1996. [26] T. Ottosson, A. Svensson, “Multirate schemes for multimedia applications in DS/CDMA systems,” Submitted to RVK’96, Luleå, Sweden, June 1996. [27] A. Johansson, A. Svensson, “Interference cancellation in multirate DS/CDMA systems,” Submitted to RVK’96, Luleå, Sweden, June 1996.

5