Frame detection and estimation of carrier frequency offset and symbol delay in mobile communication systems with OFDM transmission

Purchla-Malanowska Magdalena

Abstract

In the thesis, the problem of time and frequency synchronisation in mobile communication systems with OFDM (orthogonal frequency division multiplexing) transmission is considered. OFDM is a modern modulation technique, which exhibits such advantages as high spectral efficiency, suitability for transmission in frequency-selective channels, possibility to be applied in single frequency networks, or multiuser gain when used in a multiuser scenario. It is already used in broadcasting systems for digital television and radio, in wireless networks, and is a candidate for cellular mobile communication of the fourth generation (4G). In spite of many advantages, it has some important drawbacks, which have to be dealt with. These are a significant peak-to-average power ratio and high requirements on synchronisation. In this dissertation, the problem of synchronisation is addressed with the emphasis on frame detection and estimation of carrier frequency offset (CFO) and symbol delay in mobile wireless communication systems which use OFDM transmission. Since packet-oriented transmission is assumed, the presented algorithms are based on the use of a preamble. In consequence of the fact that the CFO has to be estimated with higher accuracy than the symbol delay, the structure of the preamble is chosen to provide the best possible CFO estimates. For this purpose, the Cramér-Rao bound for CFO estimators is found. Apart from comparing several CFO estimation algorithms found in literature the author proposes some computationally less complex estimators which ensure accuracy close to the Cramér-Rao bound. In the scope of the time synchronisation, the probability of false alarm and missing a preamble when using the Schmidl-Cox metric with various correlator structures is derived in a different way than found in literature. Furthermore, the influence of the correlator structure on the delay estimation is studied, and the author proposes a simple fine delay estimation method for correcting errors of the correlation-based estimator. All theoretical results are verified by means of computer simulations in various propagation conditions, including different multipath scenarios and Rayleigh fading. The experiments show that the accuracy of the proposed estimators is very high. Moreover, the observations based on simulation results, referring to the detection capabilities of the considered correlation-based algorithms, comply with the presented theoretical analysis.
Diploma typeDoctor of Philosophy
Author Purchla-Malanowska Magdalena (FEIT / PE)
Purchla-Malanowska Magdalena,,
- The Institute of Electronic Systems
Title in EnglishFrame detection and estimation of carrier frequency offset and symbol delay in mobile communication systems with OFDM transmission
Languageen angielski
Certifying UnitFaculty of Electronics and Information Technology (FEIT)
Disciplinetelecommunications / (technology domain) / (technological sciences)
Start date26-06-2007
Defense Date26-10-2010
End date23-11-2011
Supervisor Jerzy Szabatin (FEIT / PE)
Jerzy Szabatin,,
- The Institute of Electronic Systems

Internal reviewers Andrzej Dąbrowski (FEIT / IT)
Andrzej Dąbrowski,,
- The Institute of Telecommunications
External reviewers Wesołowski Krzysztof (Poznan University of Technology )
Wesołowski Krzysztof,,
- Poznan University of Technology
Pages161
Keywords in EnglishOFDM, synchronisation, estimation, frame detection, carrier frequency offset.
Abstract in EnglishIn the thesis, the problem of time and frequency synchronisation in mobile communication systems with OFDM (orthogonal frequency division multiplexing) transmission is considered. OFDM is a modern modulation technique, which exhibits such advantages as high spectral efficiency, suitability for transmission in frequency-selective channels, possibility to be applied in single frequency networks, or multiuser gain when used in a multiuser scenario. It is already used in broadcasting systems for digital television and radio, in wireless networks, and is a candidate for cellular mobile communication of the fourth generation (4G). In spite of many advantages, it has some important drawbacks, which have to be dealt with. These are a significant peak-to-average power ratio and high requirements on synchronisation. In this dissertation, the problem of synchronisation is addressed with the emphasis on frame detection and estimation of carrier frequency offset (CFO) and symbol delay in mobile wireless communication systems which use OFDM transmission. Since packet-oriented transmission is assumed, the presented algorithms are based on the use of a preamble. In consequence of the fact that the CFO has to be estimated with higher accuracy than the symbol delay, the structure of the preamble is chosen to provide the best possible CFO estimates. For this purpose, the Cramér-Rao bound for CFO estimators is found. Apart from comparing several CFO estimation algorithms found in literature the author proposes some computationally less complex estimators which ensure accuracy close to the Cramér-Rao bound. In the scope of the time synchronisation, the probability of false alarm and missing a preamble when using the Schmidl-Cox metric with various correlator structures is derived in a different way than found in literature. Furthermore, the influence of the correlator structure on the delay estimation is studied, and the author proposes a simple fine delay estimation method for correcting errors of the correlation-based estimator. All theoretical results are verified by means of computer simulations in various propagation conditions, including different multipath scenarios and Rayleigh fading. The experiments show that the accuracy of the proposed estimators is very high. Moreover, the observations based on simulation results, referring to the detection capabilities of the considered correlation-based algorithms, comply with the presented theoretical analysis.
KBN classification35
Thesis file
doktorat Purchla-Malanowska.pdf 1.53 MB

Get link to the record

Back
Confirmation
Are you sure?