Pre-coding for MIMO Broadcast Channels Using Dirty Paper Coding Techniques
No Thumbnail Available
Date
2011-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Wireless communication systems with multiple antennas are the focus of many applications
nowadays due to the higher throughput and/or more robust performance than single antenna
communication. One variety of these systems is the multi-user system in which many users
share the same wireless environment. In this thesis, communication strategies that use dirty
paper coding (DPC) techniques to form independent spatial streams to the Multiple Input
Multiple Output broadcast channel (MIMO BC) are studied. Unlike previous studies that
assume ideal channel conditions, the performance of DPC algorithms under frequency
selective channels having spatially correlated path gains is investigated. These algorithms are
found to be equally applicable to frequency selective channels if orthogonal frequency
division multiplexing (OFDM) is employed with 1.5 dB power loss due to correlation for
realistic correlation values.
The second part of the thesis considers issues that arise in practical implementation of the
DPC algorithms. The effect of imperfect channel estimates at the transmitter and the problem
of scheduling users for transmission are considered. For time division duplex (TDD) systems,
we show that errors in channel estimation result in performance degradation that can almost
completely be eliminated above a SNR of 25 dB. For frequency division duplex (FDD)
systems, it is shown that delay in estimation feedback results in severe performance
degradation that becomes unacceptable above some delay depending on the algorithm used.
User scheduling is formulated as user selection for single carrier systems. We adapt an
algorithm proposed for linear pre-coding techniques and show there is an increase in sum rate
in using this algorithm over uniform scheduling of users. For OFDM systems, applying the
user selection algorithm proposed for single carrier systems to each subcarrier results in
increased sum rate with lower fairness index compared to uniform scheduling. Allowing
users to select their strongest subcarriers, on the other hand, is shown to give lower sum rate
and improved fairness compared with the sum rate maximizing algorithm.
Key words: Dirty Paper Coding, MIMO Broadcast Channel, Pre-coding
Description
Keywords
Dirty Paper Coding, Mimo Broadcast Channel, Pre-coding