Asymptotic multiuser efficiency for 2-stage detectors in AWGN channels

TitleAsymptotic multiuser efficiency for 2-stage detectors in AWGN channels
Publication TypeConference Paper
Year of Publication1993
AuthorsD Brady
Conference NameProceedings of the 1993 Ieee International Symposium on Information Theory
Date Published01/1993
Abstract

In the AWGN multiple-access channel with binary phase-shift keying modulation, the kth user error probability for a given demodulator vanishes exponentially with the noise level as -ηk SNRk/2, where ηk is the asymptotic multiuser efficiency (AME), and SNRk is the received signal-to-background-noise ratio. Thus, the asymptotic multiuser efficiency is an attenuation of the error rate exponent for isolated transmission and maximum a posteriori demodulation, and provides a simple yet precise means of comparing multiuser receivers for sufficiently low noise levels. To date, this parameter is only known for the following receivers in the 2-user, asynchronous AWGN channel: the maximum likelihood sequence detector, the decorrelating detector, the linear MMSE detector, and the conventional detector. In this talk the asymptotic multiuser efficiencies for a class of detectors for the 2-user, asynchronous AWGN channel: will be presented. This class may be loosely described as receivers which estimate and subtract multiple-access interference (MAI) by using tentative data decisions, and includes the two-stage detectors with both conventional or decorrelated tentative decisions. The asymptotic multiuser efficiencies for this class of detectors clearly indicate regions for which a given user should avoid updating tentative decisions and suggest combinations of the above receivers to improve single-user performance. This technique applies to the AME of soft tentative decision strategies as well, and we demonstrate that the near-far resistance of two-stage detectors may be markedly improved using soft decision nonlinearities. Below we present an outline of the approach for conventional tentative decisions.