WO2007009361A1 - A method and system for inhibiting peak-to-average power ratio in orthogonal frequency division multiplexing system - Google Patents

Abstract

A method and system for inhibiting peak-to-average power ratio in orthogonal frequency division multiplexing system, it including: takes oversample for the inputted data and implements Inverse Discrete Fourier Transform (IDFT); amplifies said oversampled and IDFT data, and takes amplitude limit; takes DFT transform for the limitted data, and removes the outband noise; takes IDFT transform for the data removed outband noise, then collates and outputs it. Correspondingly, there is a system for inhibiting peak-to-average power ratio. The present invention can inhibit PAPP effectly, increase the average power, make the use of the data bit wide more effect, and reduce the error code rate.

Description

 TECHNICAL FIELD The present invention relates to the field of communications technologies, and in particular, to a method and system for suppressing peak power of a signal and reducing a peak-to-peak ratio in an orthogonal frequency division multiplexing communication system. . Background technique

 Orthogonal Frequency Division Multiplexing (OFDM) is a modulation method with high channel utilization. It has good anti-fading performance and can realize parallel transmission of data. With the rapid development of digital signal processing technology and With the widespread use of large-scale integrated circuits, OFDM has received increasing attention, especially in the field of high-speed digital communications.

 In an OFDM communication system, a plurality of carriers having a synchronization relationship in frequency are used to modulate a signal. Since the envelope values of the carriers are statistically independent, the ratio of the peak value to the average power ratio of the superimposed signals increases as the number of carriers increases. The value of the Peak-to-Average power ratio (PAPR) is large. Therefore, the dynamic range of the modulated signal is quite large, which requires the power amplifier in the system to have a high linear dynamic range to avoid spectral spread and nonlinear distortion of the transmitted signal. At the same time, the subsequent D/A converter is required to have a large conversion bandwidth, which increases the cost and implementation difficulty of the system.

 At present, there are many methods for reducing PAPR, which can be roughly divided into two categories: One is to properly process the input stream before the OFDM multiplexer. For example, the method of block coding the input data of OFDM can reduce the independence of the signals of the multiplexer, thereby reducing the probability of superposition of the peak-to-peak value of the signal, thereby reducing the PAPR; or using a partial transmission sequence to reduce the superposition time. The number of signals, thereby reducing the PAPR; there is also a way to select the mapping, that is, through mapping, to generate all possible combinations of the streams, and then select a combination scheme, so that the PAPR of the signals superimposed by the IFFT transform The smallest. Thus, this type of method requires a large amount of computation.

The other type is to process the signal after the OFDM multiplexer. The most direct and effective way is to limit the analog signal. However, clipping is a nonlinear process that causes severe in-band and out-of-band noise, which reduces the bit error rate performance and spectral efficiency of the overall system. An existing method of processing a signal after an OFDM multiplexer is: Ochiai, H., Imai, H. Performance analysis of deliberately clipped OFDM signals (Communication, IEEE Transactions on , Volume: 50 , Issue: 1 , Jan. 2002 Pages: 89 101 ) A method of reducing the PAPR by oversampling and limiting filtering. The structure is shown in Figure 1. The input signal is filled with zeros at the end of the signal, and then the discrete Fourier transform of the JN point is performed. (IDFT, Inverse Discrete Fourier Transform) Transform (J is an oversampling factor), the output data is sent to the limiter for limiting operation, and the data from the limiter is directly sent to the discrete Fourier transform (DFT, Discrete) The Fourier Transform module performs a J-point DFT transform. After the data is restored, the out-of-band noise is removed and N data is converted into an N-point IDFT transform, and the PAPR of the output data is improved.

This method of reducing the PAPR by sampling limiting filtering makes the data bit width more efficient, and the out-of-band interference caused by the clipping is greatly reduced. However, Quadrature Amplitude Modulation (AQ) is adopted. D is 8 bits as an example: Let 64-QAM coding, the maximum amplitude is 127, to obtain a smaller PAPR, A _max is about 10, the bit width is not used more effectively, and the clipping value is too small. A typical receiving circuit will generate a serious error.

 For this reason, it has been proposed to use a specially designed receiving circuit to solve the bit error problem. Hangjun of clipped and filtered OFDM signals. ( Communications, 2003. ICC '03. IEEE International Conference on , Volume: 5 , 11-15 May 2003 Pages : 3438 - 3442 vol. 5 ) proposes a receiver, as shown in Figure 2, where a = l- E_, the receiver requires the receiving part to contain the information of the transmitting part, which cannot fully adapt to the requirements of multiple transmitting systems, and the receiving circuit also uses the same structure as the transmitting circuit, which causes the circuit to be too complicated. Summary of the invention

 SUMMARY OF THE INVENTION An object of the present invention is to provide a peak-to-average ratio suppression method and system for limiting filtering which can greatly improve the PAPR value of a signal.

A method for suppressing a peak-to-average ratio in an orthogonal frequency division multiplexing system according to the present invention includes: The input data is oversampled and subjected to discrete Fourier transform IDFT transform; the oversampled and IDFT transformed data is amplified by nl times and subjected to clipping; the sliced data is subjected to discrete Fourier transform DFT transform, and the strip is removed External noise; IDFT transform is performed on the data after removing the out-of-band noise, and is sorted and output. Preferably, the data after removing the out-of-band noise is performed on the signal in the output module.

Reduce the ID3 by n3 times before the IDFT conversion.

 More preferably, the magnification nl and the reduction factor n3 are equal.

 More preferably, after IDFT conversion is performed on the data after the out-of-band noise is removed, the data is amplified by n2 times before the data is output.

 Preferably, the limiting operation in the limiting is performed according to the following steps:

Assume that the nth sample output after the clipping is %= Re ( + j lm(S ), and before the clipping operation, the sample is Sn, = Re ( Sn, ) + jlm ( Sn, ), then Have:

 Where 4 is the maximum value of the signal amplitude.

A preferred range of the maximum value of the signal amplitude of 4 _nax is 60 to 80.

When nl=8, n2 = 4, 4, _ax is equal to 70.

 According to the present invention, there is also provided a system for suppressing a peak-to-peak ratio in an orthogonal frequency division multiplexing system, comprising the following sequenced data processing modules:

 An oversampling module, oversampling the input data and performing an inverse discrete Fourier transform IDFT;

 The limiting module amplifies the transformed oversampled data by nl times and performs clipping; the denoising module performs discrete Fourier transform DFT transform on the data after clipping to remove out-of-band noise;

 The output module performs IDFT transformation on the data after the out-of-band noise is removed, and performs processing and output.

 Preferably, the denoising module reduces the data after removing the out-of-band noise by n3 times and then supplies the data to the output module; or

The output module performs IDFT conversion on the signal from the output module for the data from the denoising module. Reduced by n3 times before.

 More preferably, the magnification nl and the reduction factor are equal.

 Preferably, after the IDFT conversion of the data, the output module amplifies the data by n2 times and outputs the data.

 The limiting module performs limiting operation by the following steps:

限幅运 算以前该样点为 Sn， = Re ( Sn，） +jlm ( Sn， ), 则有：

i_m, {^Im )，^Im(d 其中 ^_max为信号幅度的最大值。 Suppose the 笫n samples output by the limiting module are

Before the clipping operation, the sample is Sn, = Re ( Sn,) + jlm ( Sn, ), then:

i _m , { ^Im ), ^Im( d where ^ _max is the maximum value of the signal amplitude.

The preferred range of the maximum value _ax of the limiting signal amplitude of the limiting module is 60 to 80. When nl 2, n2 = 4, 4 Li is equal to 70.

 In summary, the method and system for suppressing the peak-to-peak ratio in the orthogonal frequency division multiplexing system provided by the present invention amplifies the data before the limiting operation on the input data, the parameter is nl; after the JN point DFT transform After that, the data is restored, and the parameter is l/nl, so that the average power is increased, so that the data bit width is more effectively utilized and the bit error rate is reduced; the peak value in the limiter is the value set by the simulation, and the limit is limited. The range of magnitude is the component (real and imaginary), and the balance is chosen between the bit error rate and the PAPR suppression effect. Before the data output after the N-point IDFT is amplified, the maximum use of the data range is ensured, and the bit error rate is not large, and a general receiving circuit can be used. DRAWINGS

 In order to further illustrate the principles of the present invention, the following detailed description will be made in conjunction with the embodiments and the drawings, in which -

 Figure 1 is a schematic block diagram of the sample-like limiting filter;

 2 is a receiver of the prior art;

 Figure 3 is a functional block diagram of an embodiment in accordance with the present invention;

4 is a simulation effect in accordance with an embodiment of the present invention. detailed description

 The invention provides an achievable method and system for suppressing the peak-to-level ratio based on oversampling and limiting filtering. The basic idea is: for each transform unit, the input signal is filled with zeros to obtain JN data, and then JN is performed. The discrete Fourier transform IDFT transform of the point (J is the oversampling factor), the output data is amplified according to the simulation result by a certain multiple, the amplification factor is nl, and then sent to the limiter for limiting operation, from the limiter The output data is directly sent to the discrete Fourier transform DFT module for JN point DFT transformation, and the output data is removed from the out-of-band noise, then reduced to N data, then N-point IDFT transform, and then amplified by n2 times. The PAPR of the data is significantly improved over the original PAPR.

 In order to make the principles, features and advantages of the present invention more apparent, the invention will be described with reference to the accompanying drawings.

 For ease of understanding, a system for suppressing peak-to-amplitude ratios provided in accordance with the present invention is first described, the system comprising:

 Oversampling module: Oversampling the input data and performing JN point IDFT transformation When sampling the signal according to the Nyquist sampling frequency, the sampling points are independent of each other, and it is easy to introduce the statistical characteristics of the signal PAPR at this time. The theoretical formula has become the theoretical basis of many PAPR suppression methods. The results show that when the discrete signal is restored to continuous signal by D/A, PAPR rebound phenomenon occurs, especially the discrete signal obtained by Nyquist sampling frequency, PAPR rebound is the most serious. Therefore, the effect of PAPR suppression using the sputum-like signal will be more obvious.

 In the present invention, some 0 information (that is, too) is added to the original data stream, and 0 information can be added between each adjacent valid data, or after the data is concentrated, and the ID point of the J point is performed. Transform. The following is an example of adding 0 information after the data is concentrated.

As shown in FIG. 3, the original signal sequence for transmission in the OFDM system (the number of subcarriers is N) is represented by A={A0, Al, A2, AN-1}, where Ak is complex data on the subcarrier k. After sampling, modulation, and phase shifting, the transmitted signal becomes:

 Limiter module: Performs a clipping operation after amplifying the oversampled data after the transformation. As mentioned earlier, if you directly limit the data, you will find that to obtain a lower PAPR, the utilization of the bit width is lower, and most of the transformed data is destroyed due to the small clipping value. Conventional receiving circuits will generate severe errors, so the data needs to be processed accordingly.

 The data is amplified to increase the average power, so that the limit value can be appropriately increased to obtain a better PAPR suppression effect. In order to avoid introducing the square root in the hardware circuit implementation, the limiting object is the output value. Department and imaginary department.

 As shown in Figure 3, the signal is as follows:

为信号幅度的最大值。

Suppose its nth output sample is + J'lm ( , then:

Is the maximum value of the signal amplitude.

Amplifying the signal after the JN point IDFT by nl times can increase the Pin value of the signal, thereby reducing the clipping rate γ value, thereby effectively reducing the PAPR value. Therefore, from the viewpoint of lowering the PAPR value, it is desirable that the larger the value of nl, the better. The principle of A _max value selection is to ensure that most of the data will not be lost after the limiting operation, which is beneficial to the correct transmission of data, so it is desirable that the value of nl is as small as possible. Based on the above factors, it can be seen that the selection of nl values is contradictory, so in practical applications, a compromise solution is needed.

 Denoising module: After the clipped data is subjected to J-point DFT transform, the out-of-band noise is removed, and then the multiple is reduced by n3 times, and more preferably, n3 is equal to nl.

 The general PAPL limiting filtering method is a nonlinear process that introduces severe out-of-band noise, which reduces the bit error rate performance and spectral efficiency of the overall system.

After the previous oversampling, J-point IDFT conversion, multiple amplification, clipping, and J-point DFT operation, the effect of suppressing PAPR has been achieved. This step is to introduce the limiting band into the out-of-band. Noise is removed. As shown in Figure 3, the method of implementation is as follows:

 After the JN point DFT processing, the J-point data sequence that produces distortion is: To eliminate the out-of-band interference, the sequence of length N after the original data is distorted is:

Output module: Perform OFDM modulation (N-point IDFT) on the data and appropriately amplify the output data, and sort the IDFT-transformed data and output it in the port matching format.

最后输出数据乘以一个系数 η2的目的是保证数据范围的最大利用。 下面通过具体的实施例对本发明的技术方案作进一步描述。 实施例 为应用了本发明的无线局域网 802.11a协议系统， 这一系统釆用 64QAM 调制方式。 数据位宽为 8位， 过釆样因子 J=2,J 点 IFFT变换后放缩的 倍数为 nl=8，限幅值的取值范围为 60 - 80, 优选值为 70, N点 IDFT之 后放大的倍数是 n2=4倍。 为了观察 PAPR的改善效果， 以峰平比超过某 一门限值 z的概率 (CCDF)-PAPR值的曲线来衡量本发明所提供的方法的 效果。 具体实施步骤如下： After the previous processing, the PAPR value of the data at this time has been greatly reduced, and the real OFDM modulation is performed on the data. As shown in FIG. 2, the sequence uses the N-point IDFT for ordinary OFDM modulation and then amplifies the n2 output.

The purpose of multiplying the final output data by a factor η2 is to ensure maximum utilization of the data range. The technical solutions of the present invention are further described below by way of specific embodiments. The embodiment is a wireless local area network 802.11a protocol system to which the present invention is applied, which uses a 64QAM modulation scheme. The data bit width is 8 bits, the over-sampling factor J=2, and the multiple of the J-point IFFT transform is nl=8, and the limit value ranges from 60 to 80, preferably 70, after the N-point IDFT The magnification is magnified by n2=4 times. In order to observe the improvement effect of PAPR, the effect of the method provided by the present invention is measured by a curve of a probability that the peak-to-peak ratio exceeds a certain threshold z (CCDF)-PAPR value. The specific implementation steps are as follows:

 Oversampling the input data and performing a JN point IDFT transformation;

 After analysis, for the oversampling factor, J=2 and J=16 have the same effect on suppressing PAPR, but the latter is difficult to implement by hardware (16x64 point IDFT transform), so the oversampling factor is chosen to be 2, ie in each The IDFT transform unit (64) adds 64 zeros and then feeds them into the 128-point IDFT transform unit.

 Performing a clipping operation after amplifying the transformed oversampled data;

 Determine several sets of parameters after modeling the model

First group: n 1 = 8 n2 = 4 A _max = 70; Second group: nl-8 n2=4 A _max =127;

The third group: nl=16 n2=4 A _max =127;

The fourth group: nl=8 n2=4 A _max =40;

 The fourth group of parameters is suppressed. PAPR has the best effect, but its bit error rate is high and cannot be recovered. In the comparison between the first group and the third group, although the first group's PAPR suppression effect is worse than the third group, the received 64QAM constellation effect is significantly better than the third group. Therefore, the first group of parameters is used as Preferred parameter combinations. The simulation results are shown in Figure 4.

 After the clipped data is subjected to J-point DFT transform, the multiple is restored and the out-of-band noise is removed;

 For each IDFT unit of 128 data, the next 64 points are removed directly, and each processing unit becomes 64 data.

 The data is OFDM-modulated (N-point IDFT) and appropriately amplified, and the IDFT-converted data is collated and output in a port-matching format.

 Each 64 data from which the out-of-band noise is removed is subjected to 64-point IDFT conversion, and the data is output after being amplified by 4 times.

 The system peak-to-peak ratio of the PAPR-free process is about lldb, and the PAPR is reduced to about 7 db after the above method, as shown in Fig. 4, which is a simulation effect according to the present invention. The ordinate is the probability that the peak-to-level ratio exceeds a certain threshold z (CCDF), and the abscissa is the PAPR value. As can be seen from the graph shown in Fig. 4, the present invention can effectively suppress PAPR, and the receiving circuit does not require specific requirements and is easy to implement.

 The above embodiments are intended to illustrate and explain the principles of the invention. It is to be understood that the specific embodiments of the present invention are not limited thereto. Various changes and modifications may be made without departing from the spirit and scope of the invention.

Claims

-9- Claims  A method for suppressing a peak-to-average ratio in an Orthogonal Frequency Division Multiplexing system, comprising: oversampling input data and performing inverse discrete Fourier transform (IDFT) transform; and performing the over-the-counter and IDFT transform The data is amplified by nl times and subjected to clipping; the limited data is subjected to discrete Fourier transform DFT transform, and the out-of-band noise is removed; the data after removing the out-of-band noise is subjected to IDFT transform, and is sorted and output. 2. The method of suppressing a peak-to-peak ratio according to claim 1, wherein the data obtained by removing the out-of-band noise is reduced by n3 times before the IDFT conversion.  The method of suppressing a peak-to-peak ratio according to claim 2, wherein the amplification factor n1 and the reduction factor n3 are equal.  The method of suppressing a peak-to-peak ratio according to claim 1, wherein after the IDFT conversion is performed on the data after the out-of-band noise is removed, the data is amplified by n2 times before the data is output.  The method of suppressing a peak-to-peak ratio according to any one of claims 1 to 4, characterized in that the limiting operation in the limiting is performed according to the following steps: The nth sample output after the clipping is supported is ^ H Re( /o) + j Im . ), before the limiting operation, the sample is Sn = Re ( Sn,) + jlm ( Sn, ), Bay ¹ J : = { 》 Im, <f ^{Im :} ') ^; ' ^{) ≤} where ^ is the maximum amplitude of the signal value.  6. The method of suppressing a peak-to-peak ratio according to claim 5, wherein a maximum value of the signal amplitude ^^ ranges from 60 to 80.  7. The method of suppressing a peak-to-peak ratio according to claim 4, wherein the limiting operation in the limiting is performed according to the following steps:  Assume that the ηth sample output after the clipping is 3⁄4»=: Re . ) + _/ Im(3⁄4i) , before the limit operation, the sample is Sn = Re ( Sn' ) +jlm ( Sn' ), then: Κ· { 1 ^ΚΦ; ImdiO ;')  D ( > Jobs „)>4 Where 4 „. _x is the maximum value of the signal amplitude, when nl=8, n2 = 4, ^ equals 70 8. A system for suppressing a peak-to-peak ratio in an orthogonal frequency division multiplexing system, comprising: a data processing module connected in the following order:  An oversampling module, oversampling the input data and performing an inverse discrete Fourier transform IDFT;  The limiting module amplifies the transformed oversampled data by nl times and performs clipping; the denoising module performs discrete Fourier transform DFT transform on the data after clipping to remove out-of-band noise;  The output module performs IDFT transformation on the data after the out-of-band noise is removed, and performs processing and output.  The system for suppressing the peak-to-peak ratio according to claim 8, wherein the denoising module reduces the data after removing the out-of-band noise by n3 times and then supplies the data to the output module; or the output module pair is from the denoising The module's data is reduced by n3 times before the output module performs IDFT conversion on the signal.  The system for suppressing a peak level ratio according to claim 9, wherein the magnification factor n1 and the reduction factor n3 are equal.  The system for suppressing a peak-to-peak ratio according to claim 8, wherein the output module amplifies the data by n2 times and then outputs the data after performing IDFT conversion on the data.  A system for suppressing a peak-to-peak ratio according to any one of claims 8 to 11, wherein the limiting module performs a clipping operation by the following steps:  Assume that the nth sample output by the limiting module is ^ ^ Re ( + · Ιηι(3⁄4>), and the sample point is Sn, = Re ( Sn,) + jlm ( Sn, ) before the clipping operation. :

Where ^ _ax is the maximum value of the signal amplitude. 13. The system for suppressing peak-to-peak ratio according to claim 12, wherein the limit signal amplitude of the limiting module has a maximum value of 4 _ax ranging from 60 to 80. 14. The system for suppressing a peak to ceiling ratio according to claim 11, wherein: The limiting module performs the limiting operation by the following steps:  Assume that the nth sample output by the limiter module is ^ HRe ( + im ( ' before the limit operation , the sample is Sn , = Re (Sn, ) + jlm ( Sn, )), then: R#i ^Re '), ;' ) " { 4 ^Re OX Im ₍ = { ^{Im ;} ') where 4 is the maximum value of the signal amplitude. When nl=8, π2 = 4, Λ, _ηχ is equal to 70