Advanced decoding algorithms for MIMO systems

Contact: Raj Kumar
Room: BC 160
Tel: 021-693-6791
Email: raj [dot] kumar [at] epfl [dot] ch

Wireless systems employing multiple antennas at both the transmitter and receiver (multiple-input multiple-output (MIMO) systems) have the potential to dramatically increase both the rates and reliability of transmission. MIMO systems are a part of the recently standardized IEEE 802.11n standard for Wireless LANs, and the IEEE 802.16 WiMax standard.

A key challenge towards the widespread implementation of large MIMO systems is the development of low-complexity decoding algorithms. An optimal decoder for MIMO systems reduces to a closest lattice point search over a subset of a lattice, that is implemented using class of decoders known as “sphere decoders”. Recently, several low complexity versions of such algorithms (such as regularized lattice decoders and their lattice reduction aided linear counterparts) have been shown to be information theoretically optimal, at high signal-to-noise ratios (SNR). The aim of this project will be to:

• Understand sphere decoding algorithms for MIMO systems
• Study and compare through simulations the performance of these techniques at practical values of SNR that are commonly encountered in actual systems
• Use insights gained from such a study to try to improve these algorithms

Prerequisites:

• Knowledge of digital communication
• Good mathematical aptitude
• Knowledge of Matlab