Optimal Design of Pilot Symbols for Channel Estimation in CDD-OFDM Systems
Date Issued
2010
Date
2010
Author(s)
Huang, Wei-Chieh
Abstract
Multiple-input multiple-output (MIMO) technique is an attractive candidate for performance improvement in orthogonal frequency division multiplexing (OFDM) systems. Recently, cyclic delay diversity (CDD) scheme is developed in order to exploit the transmit diversity offered by multiple transmit antennas. The main advantages of CDD over traditional transmit diversity technique, e.g., space-time block coding (STBC), are the easier requirement for channel response and the fact that transmitter and receiver both have relatively lower complexity.
Channel state information (CSI) is generally required for coherent demodulation technique. In multi-path environments, channel estimation is typically performed by using pilot symbol assisted modulation (PSAM) scheme. The accuracy of channel estimation is critically dependent on the structure of pilot sequences. However, the optimal pilot sequences, being derived to minimize the mean square error (MSE) of channel estimation in traditional MIMO-OFDM systems, are not suitable for CDD-OFDM systems. In this dissertation, we present a general expression of the optimal pilot sequences for channel estimation in CDD-OFDM systems. The derived results illustrate that the MSE of channel estimation can be minimized if pilot sequences and cyclic-delays are appropriately designed. Through simulation experiments, we confirm that the MSE of channel estimation using the proposed pilot sequences achieves the lower bound in CDD-OFDM systems.
Since PSAM schemes require the use of dedicated pilot sub-carriers, the bandwidth utilization is substantially reduced. Therefore, the dissertation further investigates a channel estimation approach for CDD-OFDM systems using a superimposed training (ST) scheme, in which pilot symbols are superimposed onto data streams prior to transmission. It is revealed that the developed pilot sequences are also optimal for ST-based channel estimation in CDD-OFDM systems. Additionally, under the assumption that the total power of pilot sequence is a constant, we show that the performance of channel estimation is independent of the length of pilot sequence.
Under both PSAM and ST schemes, the average channel capacity of CDD-OFDM systems is investigated. Moreover, the optimal ratios of pilot symbol power to total transmission power are analyzed in order to maximize the lower bound of average channel capacity. The current results show that, when channel is assumed to be stationary over a number of OFDM symbols, the ST-based channel estimation schemes yield a higher system capacity than PSAM-based channel estimation schemes.
Subjects
Orthogonal frequency division multiplexing (OFDM)
multiple input multiple output (MIMO)
cyclic delay diversity (CDD)
channel estimation
pilot symbol assisted modulation (PSAM)
superimposed training (ST)
system capacity
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-99-D95942017-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):6bafcd6ec1d6ec3a21f7ea6d30c08b62