Skip to Main content Skip to Navigation

Advanced Techniques for Achieving Near Maximum-Likelihood Soft Detection in MIMO-OFDM Systems and Implementation Aspects for LTE/LTE-A

Abstract : In this thesis, spatial multiplexing-MIMO communication schemes with OFDM modulation are considered with a focus on 4x4 systems. In order to achieve the 3GPP LTE and 3GPP LTE-A requirements, the design of detector's architecture with high performance, low latency and applicable computational complexity is a challenging research topic at the receiver due to the power and latency limitations of mobile communication systems. The challenge consists in providing a detector that achieves near-optimum performance, while offering polynomial computational complexity. Simultaneously, implementation aspects are carefully considered, thus parallel and fixed computational complexity algorithms are strongly favoured. As an answer to the encountered issues, particular attention is paid to the detector's architecture. In particular, the adopted strategy lies in shifting as much as possible of the computational complexity to the preprocessing step that is dataindependent. Throughout the present work and in addition to the background and preliminaries introduction, an inventory of the main existing hard-decision solutions is introduced. It presents the heart of the subject and an original detector is subsequently proposed. It includes - among others - both lattice reduction and sphere decoding aspects. Its advantage over the existing techniques is presented and is shown to be preserved through its softdecision extension. As expected, the strategic choice of shifting the computational complexity to the preprocessing step is shown to be winning. It is due to significant decrease in the computational complexity, which is addressed in this thesis. As a result, the present work has led to the proposal of an original detector that is promising from both the computational complexity and the performance points of view. Namely, it leads to an almost constant - in the constellation size - computational complexity, while offering near-maximum likelihood performance. Consequently, the proposed soft-decision detector compares to existing 4x4 detectors as a very efficient solution.
Document type :
Complete list of metadatas

Cited literature [166 references]  Display  Hide  Download
Contributor : Yolande Sambin <>
Submitted on : Tuesday, September 11, 2012 - 10:50:00 AM
Last modification on : Monday, October 5, 2020 - 9:50:17 AM
Long-term archiving on: : Friday, December 16, 2016 - 12:07:56 PM


  • HAL Id : tel-00730797, version 1


Sébastien Aubert. Advanced Techniques for Achieving Near Maximum-Likelihood Soft Detection in MIMO-OFDM Systems and Implementation Aspects for LTE/LTE-A. Electronique. INSA de Rennes, 2011. Français. ⟨tel-00730797⟩



Record views


Files downloads