Return to search

Design of Optimal Precoders for Multiuser OFDM Systems with MMSE Equalization

<p> In this thesis, we consider a multiuser downlink OFDM system for which the channel state information ( CSI) is known to both the transmitter and the receiver. </p> <p> For such a system, we design an optimal precoder that minimizes the total mean square error (MSE) subject to a total power constraint for which a minimum MSE (MMSE) equalizer is employed. We show that, the MMSE precoder can be obtained by optimally allocating the subcarriers and optimally allocating the power. This problem can be solved by a two-stage process, in which we minimize the lower bound of the MSE to obtain the optimal power for each subcarrier, followed by seeking an optimal precoder to achieve this minimized lower bound. Specifically, our subcarrier allocation strategy states that, each subcarrier should be allocated to only one user that has the largest subchannel gain in that subcarrier. </p> <P> Moreover, based on this subcarrier allocation strategy, we perform an optimal power loading and design the corresponding optimal precoder that minimizes the average bit error rate (BER). Here, the MMSE equalizer is also employed. This optimization problem is solved by two stages. In the first stage, we derive the lower bound of the average BER and minimize this lower bound. After we employ the MMSE subcarrier allocation strategy, the optimal power loading problem can be efficiently solved by interior point methods. In order to reduce computation complexity, an alternative, efficient power loading method is proposed here, which is much more efficient when the number of subcarriers is large. In the second stage, to achieve the minimized lower bound, we seek a design of an optimal precoder. Simulation results show that for moderate to high signal-to-noise ratio (SNR), the performance of the minimum BER {MBER) precoder employed with the MMSE equalizer design is superior to several other design methods, including the MMSE precoder design. </p> / Thesis / Master of Applied Science (MASc)

Identiferoai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/21862
Date01 1900
CreatorsWang, Xuan
ContributorsWong, Kon, Electrical and Computer Engineering
Source SetsMcMaster University
LanguageEnglish
Detected LanguageEnglish

Page generated in 0.0024 seconds