CN102238126B - Method for reducing peak-to-average power ratio of OFDM (orthogonal frequency division multiplexing)/OQAM system based on selective sequence - Google Patents

Abstract

The invention discloses a method for reducing a peak-to-average power ratio of an OFDM (orthogonal frequency division multiplexing)/OQAM (orthogonally multiplexed QAM) system based on a selective sequence. The method comprises the following steps of: coding, interleaving and modulating an input data stream, performing up-sampling on the obtained information source data packets, and then, enablingthe information source data packets to respectively pass through a filter bank; respectively selecting an optimized phase rotation sequence for a time domain signal of each obtained data block in order; and finally, selecting a signal, of which the peak-to-average power ratio is lowest, for transmission, and transmitting information of the phase rotation sequence as sideband information. The method provided by the invention can effectively lower the peak-to-average power ratio of the OFDM/OQAM system without influencing the power spectral density and the error rate performance of the system, improve work efficiency an HPA (high-power amplifier) and reduce energy consumption.

Description

Method for reducing peak-to-average power ratio of OFDM/OQAM system based on selective sequence

Technical Field

The invention belongs to the technical field of wireless and wired communication of an orthogonal frequency division multiplexing (OFDM/OQAM) system based on offset quadrature amplitude modulation, and particularly relates to a method for reducing the peak-to-average power ratio of the OFDM/OQAM system by adopting a selective sequence.

Background

An orthogonal frequency division multiplexing (OFDM/OQAM) technique based on offset quadrature amplitude modulation is a promising multi-carrier modulation technique, which has attracted attention because it has advantages such as high spectral resolution and large data transmission rate. Some of the technical advantages of an OFDM/OQAM system compared to a conventional OFDM system can be summarized as follows: 1) when the users are not synchronous, the allocation of the sub-carriers can be well performed; 2) by designing a prototype filter pulse which meets (approximately) perfect reshaping conditions, smaller intersymbol interference and channel interference can still be obtained on the premise of not needing a cyclic prefix; 3) by combining with the filter bank modulated by OQAM, the maximized data transmission rate can be obtained; 4) high-performance spectrum sensing can be completed; 5) the design of the prototype filter can make the sidelobe of the frequency spectrum lower, so that the interference between adjacent frequency bands of the system is lower; 6) the spectrum sensing and data transmission can be synchronized. Therefore, for LTE-advanced and IMT-advanced, etc., OFDM/OQAM is proposed as a good candidate. For example, for dynamic spectrum access and cognitive radio, OFDM/OQAM has been selected as a promising technology for the physical layer, where spectrum sensing and data transmission can be performed simultaneously, that is, the same filter bank can be used for both spectrum sensing and data transmission, which ensures the compatibility performance of the system. This is a key facility for efficient opportunistic communication systems.

However, in OFDM/OQAM systems, some challenging problems remain unsolved. One of the most critical issues is the design of its High Power Amplifier (HPA) because the peak-to-average power ratio (PAPR) of OFDM/OQAM signals is too large. HPAs typically operate in or near saturation regions where output power efficiency is greatest. Thus, due to the high PAPR of the input signal, memoryless non-linear distortion will be introduced into the communication channel, resulting in out-of-band spectral regrowth. Therefore, in the OFDM/OQAM wireless communication system, it is important to reduce the high PAPR of the signal.

Many techniques have been proposed to reduce PAPR of conventional OFDM systems. For example: coding technology, polyphonic reservation, companding technology and the like. However, all methods of reducing PAPR of an OFDM signal cannot be directly applied to an OFDM/OQAM system. Because, in a conventional OFDM system, each OFDM symbol is independent in the time domain, there is no overlap between adjacent OFDM symbols. Therefore, the reduction of the PAPR of one OFDM symbol does not affect the reduction of the PAPR of a preceding or following symbol. However, in OFDM/OQAM systems, there is an overlap of time domain symbols. Therefore, we independently reduce the PAPR of each OFDM/OQAM symbol to affect other symbols, resulting in poor PAPR reduction.

Disclosure of Invention

Aiming at the problem that the peak-to-average power ratio of a signal in an OFDM/OQAM system is higher, but the existing method for controlling and reducing the peak-to-average power ratio of the OFDM signal cannot be directly applied to the OFDM/OQAM system, the invention provides a method for reducing the peak-to-average power ratio of the OFDM/OQAM system based on a selective sequence.

The invention provides a method for reducing the peak-to-average power ratio of an OFDM/OQAM system based on a selective sequence, which is characterized in that a signal source data packet of the OFDM/OQAM system comprises M data blocks, and each data block comprises N q-ary symbols; wherein q is 2^lL, M and N are positive integers, characterized in that the method comprises the following steps:

(1) transfusion systemAn input data stream is coded, interleaved and modulated to obtain a signal source data packet X ═ X₀，X₁，...，X_M-1]Wherein X is_mIs the mth data block in X, denoted as X_m＝[X_m(0)，X_m(1)，...，X_m(N-1)]^TWherein X is_m(n) denotes an nth data symbol of an mth data block, and X_m(n)＝a_m(n)+jb_m(N), N ═ 0,1,. and N-1, where a is_m(n) and b_m(n) are each X_m(n) real and imaginary parts, j being the imaginary sign, T representing the transpose of the matrix;

(2) up-sampling the input data symbols by a factor of N, X_mIs obtained from the nth data symbol of the data stream

Is marked as $x_m^n\left(t\right)=X_m\left(n\right)\mathrm{\δ}(t-\mathrm{mT}),$ Where T is a data symbol duration, T is a time variable, T ≧ 0, an impact function $\mathrm{\δ}\left(t\right)=\left\{\begin{array}{cc}1,& t=0\\ 0,& \mathrm{else}\end{array}\right.,$ The resulting signal is then passed through a prototype filter h (t) and

delay of imaginary part of

Then X_mAfter the nth data symbol in the data signal passes through the filter and is modulated to the symbol signal on the carrier wave

Is shown as

$s_m^n\left(t\right)=\{a_m\left(n\right)h(t-\mathrm{mT})+j{b}_m\left(n\right)h(t-\frac{T}{2}-\mathrm{mT})\}{e}^{\mathrm{jn}(\frac{2\mathrm{\π}}{T}t+\frac{T}{2})}$

Wherein $e^{\mathrm{jn}(\frac{2\mathrm{\π}}{T}t+\frac{T}{2})}=\sin \left(n(\frac{2\mathrm{\π}}{T}t+\frac{T}{2})\right)+j\cos \left(n(\frac{2\mathrm{\π}}{T}t+\frac{T}{2})\right);$

(3) Reducing the peak-to-average power ratio by using a selective sequence:

(3.1) generating U phase rotation sequences, wherein each phase rotation factor is selected from a preset set {1, -1}, and recording the U phase rotation sequences as

U-1, wherein

N-0, 1., N-1, noteRepresenting a sequence of phase rotations applied to the m-th block of data, and let P_m，u＝P_u；

(3.2) initializing m to 0, and multiplying the m-th data block by the U phase rotation sequences to obtain U candidate signals

${\tilde{s}}_{0,u}\left(t\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{s}_0^n\left(t\right){\mathrm{<figure-callout\; id="0"\; label="P"\; filenames="HDA0000070294060000021.png"\; state="\{\{state\}\}">P</figure-callout>}}_{0,u}^n,u=\mathrm{0,1},...,U-1$

And calculating the U candidate signals

${\mathrm{PAPR}}_0^u={10\log }_{10}\frac{\underset{0\&le;t\&le;(K-1)T}{\max }\left[{\left|{\tilde{s}}_{0,u}\left(t\right)\right|}^2\right]}{E\left[{\left|{\tilde{s}}_{0,u}\left(t\right)\right|}^2\right]},u=\mathrm{0,1},...,U-1$

Wherein E [ □ ] represents the averaging;

(3.3) minimum of notes

Is composed of

Corresponding to a signal block of

The corresponding phase rotation sequence number is denoted as S₀And let m be m + 1;

(3.4) multiplying the mth data block by the U phase rotation sequences respectively to obtain U candidate signals

${\tilde{s}}_{m,u}\left(t\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{s}_m^n\left(t\right)P_{m,u}^n,u=\mathrm{0,1},...,U-1$

And calculating the U candidate signals

${\mathrm{PAPR}}_m^u={10\log }_{10}\frac{\underset{\mathrm{mT}\&le;t\&le;(m+K-1)T}{\max }\left[{|\underset{p=0}{\overset{m-1}{\mathrm{\&Sigma;}}}{\tilde{s}}_p^{*}\left(t\right)+{\tilde{s}}_{m,u}\left(t\right)|}^2\right]}{E\left[{|\underset{p=0}{\overset{m-1}{\mathrm{\&Sigma;}}}{\tilde{s}}_p^{*}\left(t\right)+{\tilde{s}}_{m,u}\left(t\right)|}^2\right]},u=\mathrm{0,1},...,U-1;$

(3.5) minimum

Is composed of

Corresponding to a signal block of

Correspond toThe sequence number of the phase rotation sequence of (1) is denoted by S_mAnd let m be m + 1;

(3.6) if M < M, going to step (3.4); otherwise, the final output signal is calculated

$\tilde{s}\left(t\right)=\underset{m=0}{\overset{M-1}{\mathrm{\&Sigma;}}}{\tilde{s}}_m^{*}\left(t\right)$

Turning to step (4);

(4) let the optimized phase rotation sequence number be S ═ S₀，S₁，...，S_M-1]Signal to be output

Is sent into the channel and the sequence S is sent into the channel as side information.

The invention relates to a method for reducing the peak-to-average power ratio of OFDM/OQAM signals based on a selective sequence, which not only can better reduce the peak-to-average power ratio of the OFDM/OQAM signals, but also can keep the power spectral density of the OFDM/OQAM signals unchanged, thereby obtaining better error rate performance. After data is encoded, interleaved and modulated, the obtained signal source data packets are up-sampled and then respectively pass through a filter bank. And then, selecting an optimized phase rotation sequence for the obtained OFDM/OQAM signal according to a rule, finally selecting a signal with the minimum peak-to-average power ratio for transmission, and sending information of the phase rotation sequence as sideband information to a receiving end. The invention can effectively reduce the peak-to-average power ratio of OFDM/OQAM signals, simultaneously can ensure that the power spectral density and the bit error rate performance of a system are not influenced, improves the working efficiency of HPA and saves power consumption.

Drawings

FIG. 1 is a system flow diagram of the method of the present invention;

FIG. 2 is a block diagram of a process for reducing peak-to-average power ratio using selective sequences in accordance with the present invention.

Detailed Description

The invention is further described with reference to the accompanying drawings and a set of specific parameters:

the processing flow of the method for reducing the peak-to-average power ratio of the OFDM/OQAM system by adopting the selective sequence is shown in figure 1, wherein the processing flow for reducing the peak-to-average power ratio by adopting the selective sequence is shown in figure 2.

The duration of the prototype filter h (t) is KT, where K is a positive even number; generally, the larger the K, the smaller the side lobe; however, an increase in K results in higher computational complexity; therefore, the value of K chosen in practice usually does not exceed 8.

The steps (1) and (2) of the method are the same as the original processing steps of the transmitting end of the OFDM/OQAM system.

The present invention is explained in more detail below by means of examples, which are only illustrative and the scope of protection of the present invention is not limited by these examples.

Example (c):

description of the parameters: in the OFDM/OQAM system with q-4 QAM modulation, the number of subcarriers N-64 and the prototype filter parameter K-4, the number of phase rotation sequences generated in the selective sequence is respectively U-16 and U-128.

Simulation results show that the capability of reducing the PAPR of the invention is far better than the capability of directly applying the traditional selective mapping method to an OFDM/OQAM system to reduce the PAPR. Meanwhile, the power spectral density of the OFDM/OQAM signal is not influenced by the method. Moreover, the PAPR is reduced, so that the work efficiency of the HPA is improved, and the consumed power is saved.

When U is 128, at Pr { PAPR > PAPR₀}＝10^-3When the selective sequence method provided by the invention is adopted, the reduction of the PAPR can be obtained to be 4.7 dB; the reduction of PAPR by 1.3dB can be obtained by adopting the original selective mapping method. Obviously, the present invention greatly improves the PAPR reduction capability when the parameters are the same. Meanwhile, the power spectral density of the OFDM/OQAM signal after the PAPR is reduced by the method completely coincides with the ideal power spectral density of the OFDM/OQAM original signal under the condition that the OFDM/OQAM original signal does not pass through the HPA. When U is 16, the proposed method saves 45% of the energy compared to the original signal, whereas the conventional selective mapping method saves only 10% of the energy. Meanwhile, the capability of reducing the PAPR of the invention is far better than the capability of directly applying the traditional selective mapping method to an OFDM/OQAM system to reduce the PAPR, and simultaneously, the capability of saving energy is greatly improved.

The above description is a preferred embodiment of the present invention, but the present invention should not be limited to the disclosure of the embodiment and the drawings. Therefore, it is intended that all equivalents and modifications which do not depart from the spirit of the invention disclosed herein are deemed to be within the scope of the invention.

Claims (1)

1. A method for reducing peak-to-average power ratio of OFDM/OQAM system based on selective sequence, set up the signal source data packet of an OFDM/OQAM system to include M data blocks, each data block includes N q system symbols; wherein q is 2^lL, M and N are positive integers, characterized in that the method comprises the following steps:

(1) an input data stream is coded, interleaved and modulated to obtain a signal source data packet X ═ X₀,X₁,...,X_M-1]Wherein X is_mIs the mth data block in X, denoted as X_m＝[X_m(0),X_m(1),...,X_m(N-1)]^TWherein X is_m(n) denotes an nth data symbol of an mth data block, and X_m(n)＝a_m(n)+jb_m(N), N ═ 0,1,. and N-1, where a is_m(n) and b_m(n) are each X_m(n) real and imaginary parts, j being the imaginary sign,^Trepresents a transpose of a matrix;

(2) up-sampling the input data symbols by a factor of N, X_mIs obtained from the nth data symbol of the data stream

Is marked asWherein T' is a data symbol duration, T is a time variable, T is greater than or equal to 0, and an impact function $\mathrm{\&delta;}\left(t\right)=\left\{\begin{array}{cc}1,& t=0\\ 0,& \mathrm{else}\end{array}\right.,$ The resulting up-sampled signal is then passed through a prototype filter h (t) anddelay of imaginary part of

Then X_mAfter the nth data symbol in the data signal passes through the filter and is modulated to the symbol signal on the carrier wave

Is shown as

$s_m^n\left(t\right)=\{a_m\left(n\right)h(t-m{T}^{\&prime;})+j{b}_m\left(n\right)h(t-\frac{T^{\&prime;}}{2}-m{T}^{\&prime;})\}{e}^{\mathrm{jn}(\frac{2\mathrm{\&pi;}}{T^{\&prime;}}t+\frac{T^{\&prime;}}{2})}$

Wherein, $e^{\mathrm{jn}(\frac{2\mathrm{\&pi;}}{T^{\&prime;}}t+\frac{T^{\&prime;}}{2})}=\sin \left(n(\frac{2\mathrm{\&pi;}}{T^{\&prime;}}t+\frac{T^{\&prime;}}{2})\right)+j\cos \left(n(\frac{2\mathrm{\&pi;}}{T^{\&prime;}}t+\frac{T^{\&prime;}}{2})\right);$

(3) reducing the peak-to-average power ratio by using a selective sequence:

(3.1) generating U phase rotation sequences, wherein each phase rotation factor is selected from a preset set {1, -1}, and recording the U phase rotation sequences as

U-1, whereinN =0, 1.., N-1, note

Representing a sequence of phase rotations applied to the m-th block of data, and let P_m,u＝P_u；

(3.2) initializing m to 0, and setting the m-th data block X_mAfter the nth data symbol in the data signal passes through the filter and is modulated to the symbol signal on the carrier wave

Multiplying the phase rotation sequences by U phases respectively to obtain U candidate signals

${\tilde{s}}_{m,u}\left(t\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{s}_m^n\left(t\right)P_{m,u}^n,u=\mathrm{0,1},...,U-1$

And calculating the U candidate signals

${\mathrm{PAPR}}_m^u=10{\log }_{10}\frac{\underset{0\&le;t\&le;(K-1)T}{\max }\left[{\left|{\tilde{s}}_{m,u}\left(t\right)\right|}^2\right]}{E\left[{\left|{\tilde{s}}_{m,u}\left(t\right)\right|}^2\right]},u=\mathrm{0,1},...,U-1$

Wherein E [. cndot. ] represents averaging;

(3.3) recording the residue obtained in step (3.2)

The smallest value in (1) is noted

The corresponding signal block is

The corresponding phase rotation sequence number is denoted as S₀And let m be m + 1;

(3.4) multiplying the mth data block by the U phase rotation sequences respectively to obtain U candidate signals

${\tilde{s}}_{m,u}\left(t\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{s}_m^n\left(t\right)P_{m,u}^n,u=\mathrm{0,1},...,U-1$

And calculating the U candidate signals

${\mathrm{PAPR}}_m^u=10{\log }_{10}\frac{\underset{\mathrm{mT}\&le;t\&le;(m+K-1)T}{\max }\left[{|\underset{p=0}{\overset{m-1}{\mathrm{\&Sigma;}}}{\tilde{s}}_p^{*}\left(t\right)+{\tilde{s}}_{m,u}\left(t\right)|}^2\right]}{E\left[{\left|\underset{p=0}{\overset{m-1}{\mathrm{\&Sigma;}}}{\tilde{s}}_p^{*}\left(t\right){\tilde{s}}_{m,u}\left(t\right)\right|}^2\right]},u=\mathrm{0,1},...,U-1;$

(3.5) minimum

Is composed of

The corresponding signal block is

The corresponding phase rotation sequence number is denoted as S_mAnd let m be m + 1;

(3.6) if M < M, going to step (3.4); otherwise, the final output signal is calculated

$\tilde{s}\left(t\right)=\underset{m=0}{\overset{M-1}{\mathrm{\&Sigma;}}}{\tilde{s}}_m^{*}\left(t\right)$

Turning to step (4);

(4) let the optimized phase rotation sequence number be S ═ S₀,S₁,...,S_M-1]Signal to be output

Is sent into the channel and the sequence S is sent into the channel as side information.

Images