Advances in Multiuser Detection

4

PERFORMANCE WITH RANDOM SIGNATURES

Matthew J. M. Peacock, Iain B. Collings, and Michael L. Honig

4.1 RANDOM SIGNATURES AND LARGE SYSTEM ANALYSIS

In this chapter, we study the performance of multiuser detectors where the users are assigned random signatures. The signatures may result from spreading in time and frequency, as in variants of Code-Division Multiple Access (CDMA), and spreading in space when multiple antennas are available at the transmitter and/or receiver. Random signatures are commonly used in cellular CDMA systems, and are often used to model channel gains across multiple receive antennas.

In general, there are three main objectives of this performance analysis: (1) It should provide an efficient method for estimating performance metrics such as Signal-to-Interference Plus Noise Ratio (SINR) and bit error rate as a function of system parameters; (2) It should provide a means for optimizing system parameters in different scenarios; and (3) It should provide insight into the relative performance of different receivers. Meeting these objectives becomes challenging when the system model has many parameters (e.g., number of users, bandwidth, spreading factor, power and signature assignments, channel gains, modulation format, etc.), and the performance metrics take on complicated forms. Furthermore, what is often desired are averages of the performance metrics over realizations of random parameters, such as channels and signatures, which can be quite difficult without ...