Scope

The course introduces multiuser detection (MUD) to increase the capacity of personal communication systems, both CDMA and narrowband. Where possible, the orientation is to third generation systems. Because the success of such methods rests on exploitation of diversity, the use of antenna arrays is a second theme.

It was originally developed as a second-level graduate course. However, this offering recognises that many students will not have studied communication in fading channels. Consequently, there will be an extensive introductory section to cover fading channel propagation phenomena, detection of single-user signals in fading and corresponding performance analysis.

As for MUD, several technical approaches have been proposed: decorrelation, interference cancellation and joint sequence estimation are just a few. Each places its own demands on computation and on accuracy of the channel estimates and each provides its own level of performance. However, despite almost a decade of very active investigation in the research and industrial communities, there are still significant open questions associated with every method, and the collective state of knowledge continues to evolve.

How should the course be structured in order to have continuing utility in such an evolutionary environment? The strategy proposed here is to establish a core of general principles and analysis of the most common and/or most promising methods. More detailed analyses of specific methods or development of new methods would make excellent thesis topics.

The course contains only about 22 hours of instruction. It does cover a lot of ground, but the time should be sufficient for an accelerated tour of MUD.

Preparation Required of Students

The treatment will be at the level of a typical graduate course. Mathematics is inevitable, and students should be comfortable with signal detection and estimation at the graduate level. The minimum requirement is a senior undergraduate course in digital communications and students should already have a solid understanding of:

detection of isolated pulses in white and coloured noise, signal constellations, matched filter, maximum likelihood detection;
detection of pulse sequences, intersymbol interference, equalization;
the Viterbi algorithm;
basic convolutional encoding;

and have absorbed the baseline CDMA performance paper by K.S. Gilhousen et al., “On the Capacity of a Cellular CDMA System,” IEEE Trans. Veh. Technol, May 1991.

The course will begin with a brief review of some of these topics, plus an introduction to fading channels.

References

Good reference works for the course are:

· P. van Rooyen, M. Lotter and D. van Wyk, Space-Time Processing for CDMA Mobile Communications, Kluwer Academic Publishers, 2000.

· J.G. Proakis, Digital Communications, third ed., McGraw-Hill, 1995 (essential).

· S. Haykin, Adaptive Filter Theory, third ed. Prentice-Hall, 1995.

· F. Swarts et al., CDMA Techniques for Third Generation Mobile Systems, Kluwer, 1999.

· S. Glisic and B. Vucetic, Spread Spectrum CDMA Systems for Wireless Communications, Artech House, 1997.

and, of course,

· J.K. Cavers, Mobile Channel Characteristics, Kluwer, 2000.

In addition, key journal papers will be identified as references.