Minimum bit error rate multiuser detection techniques
Minimum bit error rate multiuser detection techniques
The Minimum Bit Error Rate (MBER) linear MUDs considered are designed for the synchronous downlink of Direct Sequence CDMA (DS-CDMA) systems, employing Binary Phase Shift Keying (BPSK) modulation as well as 4-level Quadrature Amplitude Modulation (4-QAM). An iterative non-adaptive algorithm capable of converging to the MBER solution was developed based on the Steepest Descent Gradient algorithm. A Simplified Conjugate Gradient algorithm was also proposed for the sake of improving the convergence speed of the Steepest Descent Gradient MBER algorithm. Our simulation results showed that the MBER MUD is capable of outperforming the MMSE MUD in terms of the achievable BER under various channel conditions. Furthermore, the MBER MUD is combined with Forward Error Correction (FEC) coding and several coded modulation schemes are invoked for the sake of enhancing the achievable BER performance.
Adaptive versions of the MBER algorithm are also presented, which are initialised for example to the MMSE MUD weights and then iteratively adjust the weights until the MBER solution is reached. More specifically, two classes of adaptive algorithms are presented, namely block adaptive and sample-by-sample adaptive algorithms. Two algorithms belonging to the block adaptive category are referred to as the Block Adaptive Steepest-descent Gradient (BASG) and the Block Adaptive Conjugate Gradient (BACG) algorithms. Although the BACG algorithm is capable of a faster convergence to the MBER solution, this algorithm exhibits a higher sensitivity to the choice of the adaptive step size. Hence, a hybrid solution is preferred. Similarly, two sample-by-sample adaptive algorithms were derived, namely the Least Bit Error Rate (LBER) algorithm and the Approximate LBER (ALBER) arragement. Our simulation results demonstrated that both algorithms outperformed the Least Mean Square (LMS), the Difference approximation MBER (DMBER) and the Approximate MBER (AMBER) algorithms m terms of the attainable convergence speed and the steady state BER.
University of Southampton
2003
Samingan, Ahmad Kamsani
(2003)
Minimum bit error rate multiuser detection techniques.
University of Southampton, Doctoral Thesis.
Record type:
Thesis
(Doctoral)
Abstract
The Minimum Bit Error Rate (MBER) linear MUDs considered are designed for the synchronous downlink of Direct Sequence CDMA (DS-CDMA) systems, employing Binary Phase Shift Keying (BPSK) modulation as well as 4-level Quadrature Amplitude Modulation (4-QAM). An iterative non-adaptive algorithm capable of converging to the MBER solution was developed based on the Steepest Descent Gradient algorithm. A Simplified Conjugate Gradient algorithm was also proposed for the sake of improving the convergence speed of the Steepest Descent Gradient MBER algorithm. Our simulation results showed that the MBER MUD is capable of outperforming the MMSE MUD in terms of the achievable BER under various channel conditions. Furthermore, the MBER MUD is combined with Forward Error Correction (FEC) coding and several coded modulation schemes are invoked for the sake of enhancing the achievable BER performance.
Adaptive versions of the MBER algorithm are also presented, which are initialised for example to the MMSE MUD weights and then iteratively adjust the weights until the MBER solution is reached. More specifically, two classes of adaptive algorithms are presented, namely block adaptive and sample-by-sample adaptive algorithms. Two algorithms belonging to the block adaptive category are referred to as the Block Adaptive Steepest-descent Gradient (BASG) and the Block Adaptive Conjugate Gradient (BACG) algorithms. Although the BACG algorithm is capable of a faster convergence to the MBER solution, this algorithm exhibits a higher sensitivity to the choice of the adaptive step size. Hence, a hybrid solution is preferred. Similarly, two sample-by-sample adaptive algorithms were derived, namely the Least Bit Error Rate (LBER) algorithm and the Approximate LBER (ALBER) arragement. Our simulation results demonstrated that both algorithms outperformed the Least Mean Square (LMS), the Difference approximation MBER (DMBER) and the Approximate MBER (AMBER) algorithms m terms of the attainable convergence speed and the steady state BER.
This record has no associated files available for download.
More information
Published date: 2003
Identifiers
Local EPrints ID: 465143
URI: http://eprints.soton.ac.uk/id/eprint/465143
PURE UUID: 272e829b-3466-44db-a5d4-bd99c49532a2
Catalogue record
Date deposited: 05 Jul 2022 00:25
Last modified: 05 Jul 2022 00:25
Export record
Contributors
Author:
Ahmad Kamsani Samingan
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics