Analysis of Adaptive Multiuser Receivers for DS-CDMA Using Recurrent Neural Networks

Rodrigo C. de Lamare, Raimundo Sampaio-Neto

DOI: 10.14209/its.2002.259

Evento: 2002 International Telecommunications Symposium (ITS2002)

Keywords:

Abstract

"In this paper we investigate adaptive multiuser receivers for DS-CDMA systems using recurrent neural net- works (RNN). A comparative analysis of multiuser detec- tion (MUD) schemes employing linear and non-linear struc- tures is carried out. Adaptive minimum mean squared error (MMSE) linear MUD receivers are examined with the LMS algorithm and compared with MMSE neural MUD receivers operating with the real time recurrent learning (RTRL) al- gorithm. Computer simulation experiments including dif- ferent communication channels and a varying number of users show that the neural MUD receiver operating with the RTRL algorithm outperforms linear MUD receivers with the LMS and the conventional single user detector (SUD)."

Download