Sphere Decoding of Multi-Antenna Systems

Tutor: Dirk W├╝bben
Type of Thesis: Projektarbeit (Dipl.)
date of end: 08/2004
Student: Xiaolong Ren
Status: finished
ANT-shelfmark: P-04/08/1

The maximum-likelihood criteria achieves the optimal detection of transmitted signal at a receiver. Unfortunately, the effort of using the ML criteria for signal detection is prohibitly large, thuas subotimal algorihtms are applied.

In the case of multi antenna systems, each receive signal constist of a liner superposition of the different transmitted signals weighted by the dedicated fading coefficient. In principel, ML criteria can be applyied for signal detection, but the effort is prohibitevely large even for small number of transmit antennas and small modulation alphabets.

An interesting alternative to ML detection is given by the so called sphere decoding. Here the received signal is mapped to the lattice of possible receive symbols and compared ony to lattice points inside a distinct radius. By reducing the number of required comparisons, the decoding complexity is reduced significantly, wherby the pefromance is comparable to ML detection.

In this thesis the basics of sphere decoding should be inverstigated and adopted to the detction of multi antenna systems. The algorithms have to be programmed in MATLAB and the simulation results should be compared to the BLAST detection algorithms.

Last change on 06.05.2008 by
AIT ieee GOC tzi ith Fachbereich 1
© Department of Communications Engineering - University of BremenImprint / Contact