Publications

Detailed Information

Low-Complexity Schemes for Class-III and CORR SLM in OFDM Systems

DC Field Value Language
dc.contributor.advisor노종선-
dc.contributor.author우준영-
dc.date.accessioned2017-07-13T07:11:20Z-
dc.date.available2017-07-13T07:11:20Z-
dc.date.issued2015-08-
dc.identifier.other000000067001-
dc.identifier.urihttps://hdl.handle.net/10371/119120-
dc.description학위논문 (박사)-- 서울대학교 대학원 : 전기·컴퓨터공학부, 2015. 8. 노종선.-
dc.description.abstractIn this dissertation, orthogonal frequency division multiplexing (OFDM) system
is studied. Since OFDM signal sequence undergoes high peak-to-average power ratio
(PAPR), several schemes are proposed to mitigate the PAPR problem. PAPR reduction
schemes such as selected mapping (SLM) and partial transmit sequence (PTS) are introduced.
Due to the high computational complexity of the SLM scheme, low-complexity
SLM schemes have been proposed by many researchers. Class-III SLM scheme [55]
requires only one inverse fast Fourier transform (IFFT) operation, whereas the conventional
scheme needs U IFFT operations. By randomly selecting the cyclic shift and rotation
values, this scheme can generate up to N3 alternative OFDM signal sequences.
The PAPR reduction performance of Class-III SLM scheme is little degraded compared
to the conventional SLM scheme. Recently, instead of PAPR reduction, the different
performance criteria for SLM scheme are proposed such as inter modulation distortion
[38] and correlation (CORR) [56]. The objective of these schemes are enhancing the bit
error rate (BER) performance instead of PAPR reduction performance.
In the first part of this dissertation, a deterministic selection method of phase sequences is proposed for Class-III SLM scheme [55]. First, the optimal condition of
cyclic shift values in the Class-III SLM scheme is proposed. Then, the cyclic shift values
satisfying the optimal condition is also derived. Compared to the random selection
method, the proposed selection method guarantees the optimal PAPR reduction performance.
Second, two generation methods for good alternative OFDM signal sequences
are proposed, one by using rotation values which do not have linear relation and the
other with no rotation values. The advantages of the proposed selection schemes are:
(a) The second proposed selection scheme does not need the rotation values. (b) Both
of the proposed selection schemes require less side information than random selection
scheme. (c) The first proposed selection scheme guarantees the optimal PAPR reduction
performance in terms of variance of correlation.
In the second part of this dissertation, the proper oversampling rate for the CORR
SLMscheme is proposed. It is known that four times oversampling is enough to estimate
the PAPR of the continuous OFDM signal. By calculating the correlation coefficient
between the continuous and two times oversampled OFDMsignal sequences, it is found
that two times oversampling is enough to achieve the same BER performance as four
times oversampling case in the CORR SLM scheme. In the simulation results, the same
BER performance can be achieved by the proposed two times oversampling rate as four
times oversampling case.
-
dc.description.tableofcontentsAbstract i
Contents iii
List of Tables vii
List of Figures ix
1. Introduction 1
1.1. Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2. Overview of Dissertation . . . . . . . . . . . . . . . . . . . . . . . . . 4
2. OFDM System Model 7
2.1. OFDM System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2. Modulation and Demodulation of OFDM Signal . . . . . . . . . . . . 9
2.2.1. Orthogonality Principle . . . . . . . . . . . . . . . . . . . . . . 9
2.2.2. OFDM Signal Modulation and Demodulation . . . . . . . . . . 10
2.3. Fast Fourier Transform . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4. Guard Interval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.5. Peak-to-Average Power Ratio . . . . . . . . . . . . . . . . . . . . . . . 13
2.5.1. Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.5.2. The distribution of PAPR . . . . . . . . . . . . . . . . . . . . . 13
2.5.3. PAPR of Oversampled Signal . . . . . . . . . . . . . . . . . . 15
3. PAPR Reduction Schemes 17
3.1. Clipping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2. Tone Reservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3. Partial Transmit Sequence . . . . . . . . . . . . . . . . . . . . . . . . 19
3.4. Selected Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.5. Low-Complexity SLM Schemes . . . . . . . . . . . . . . . . . . . . . 24
3.5.1. SLM Scheme with Divided IFFT Stages . . . . . . . . . . . . . 24
3.5.2. Modified SLM Scheme . . . . . . . . . . . . . . . . . . . . . . 25
3.5.3. SLM Scheme with Conversion Matrices . . . . . . . . . . . . . 26
3.6. Considerations for PAPR Reduction Schemes . . . . . . . . . . . . . . 28
4. BER Reduction Schemes 30
4.1. PTS Scheme with PICR Metric . . . . . . . . . . . . . . . . . . . . . . 30
4.2. IMD Reduction Scheme . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3. PTS Scheme with MSE Metric . . . . . . . . . . . . . . . . . . . . . . 33
4.4. DSR Reduction Scheme with Distortion Prediction . . . . . . . . . . . 34
5. Low-Complexity Class-III SLM Scheme 37
5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.2. Overview of Class-III SLM Scheme . . . . . . . . . . . . . . . . . . . 39
5.3. Selection of Optimal Alternative OFDM Signal Sequences for Class-III
SLM Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.3.1. Correlation Analysis . . . . . . . . . . . . . . . . . . . . . . . 41
5.3.2. Selection of Optimal Cyclic Shift Values . . . . . . . . . . . . 44
5.3.3. Maximum Number of Optimal Alternative OFDM Signal Sequences
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.3.4. Selection of Additional Alternative OFDM Signal Sequences . . 49
5.4. Side Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.5. Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.6. Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
6. Low-Complexity CORR SLM Scheme 61
6.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
6.2. Overview of SLM Scheme Using CORR Metric . . . . . . . . . . . . . 62
6.2.1. Overview of CORR Metric . . . . . . . . . . . . . . . . . . . . 62
6.2.2. BER Performance of SLM Scheme under HPA . . . . . . . . . 65
6.3. Oversampling Effect on SLM Scheme Using CORR Metric . . . . . . . 67
6.3.1. Expression of Oversampled Signal and CORR Metric . . . . . . 67
6.3.2. Correlation Coefficients between Coefficient Sequences Derived
from CORR Metric Computation . . . . . . . . . . . . . . . . 70
6.4. Computational Complexity . . . . . . . . . . . . . . . . . . . . . . . . 72
6.5. Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
6.6. Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
6.6.1. Effect of 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
6.6.2. Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.6.2.1. Comparative CORR . . . . . . . . . . . . . . . . . . 82
6.6.2.2. Low Sampled CORR . . . . . . . . . . . . . . . . . 83
6.7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
7. Conclusions 86
Bibliography 88
초록 96
-
dc.formatapplication/pdf-
dc.format.extent2589804 bytes-
dc.format.mediumapplication/pdf-
dc.language.isoen-
dc.publisher서울대학교 대학원-
dc.subjectOFDM-
dc.subject.ddc621-
dc.titleLow-Complexity Schemes for Class-III and CORR SLM in OFDM Systems-
dc.typeThesis-
dc.description.degreeDoctor-
dc.citation.pages97-
dc.contributor.affiliation공과대학 전기·컴퓨터공학부-
dc.date.awarded2015-08-
Appears in Collections:
Files in This Item:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share