Font Size: a A A

Architectures VLSI des recepteurs a usagers multiples pour la communication sans fil de troisieme generation WCDMA

Posted on:2009-05-09Degree:Ph.DType:Thesis
University:Universite du Quebec a Trois-Rivieres (Canada)Candidate:Ho, Quoc-ThaiFull Text:PDF
GTID:2448390005959959Subject:Engineering
Abstract/Summary:
The interferences -- intersymbol interferences (ISI) and multiaccess interferences (MAI) -- are severely problematic for data reconstruction in the third generation (3G) wireless communication systems. In presence of MAI, the multiuser detection (MUD) becomes an essential part in deployment of 3G systems at receivers' level. The MUD has been making a continuous research topic since more twenty years. The algorithmic aspect of MUD is transparent. There exist several sophisticated algorithms permitting to effectively cancel MAI. Nevertheless, due to the implementation complexity while satisfying algorithmic performances, the aspect of VLSI implementation still remains non-trivial. So this thesis deals with VLSI implementation and design methodology of multiuser receivers.;Key words: Multiuser detection (MUD), WCDMA communication, VLSI architecture, design methodology, parallelism, system-on-a-chip (SoC).;The major contribution of this doctoral thesis consists in proposing and developing the parallel VLSI architectures of a class of adaptive MUD algorithms. These architectures satisfy the algorithmic performances for several data rates in several WCDMA communication scenarios. Also, the studied receiver was extended and implemented for data multirate systems. Considering limitations of commercial tools, and limitations of design methodologies actually, the second contribution of this work deals with the design methodologies. It consists of two straightforward design methodologies for complex systems using commercial tools. The first one, namely conventional approach, concerns an improvement in the standard design flow. The second one, namely unified approach, consists in using the high level languages -- C/C++, SystemC and MATLABRTM -- in hardware design of complex systems.
Keywords/Search Tags:VLSI, Communication, MUD, Systems, Architectures
Related items