A peak-to-average ratio suppression method, system, device and storage medium for OFDM system

Abstract

This invention provides a peak-to-average power ratio (PAPR) suppression method for OFDM systems, belonging to the field of OFDM technology in mobile communication systems. The method includes acquiring the OFDM time-domain signal after inverse fast Fourier transform (IFFT), performing a clipping operation to obtain a clipped signal, and then obtaining clipped noise; scaling the clipped noise and multiplying it by a noise figure to obtain the OFDM system noise; performing piecewise compression on the system noise according to a piecewise compression function and a compression threshold; performing a logarithmic compression operation on system noise greater than the compression threshold, while keeping system noise less than the compression threshold unchanged to obtain compressed noise; and using the compressed noise to correct the clipped signal to obtain the OFDM time-domain signal with PAPR suppression. This solution solves the problem that existing clipping filtering techniques, when using compression functions to compress clipped noise, suffer from the inflexibility and complexity of the compression function structure, making it impossible to control the balance between PAPR suppression performance and bit error rate to meet system performance requirements.

Description

Technical Field

[0001] The present invention belongs to the technical field of OFDM for mobile communication systems, and in particular relates to a peak-to-average ratio suppression method, system, device and storage medium for an OFDM system. Background Art

[0002] OFDM (Orthogonal Frequency Division Multiplexing, OFDM) is one of the key technologies of current wireless digital communication systems. Its essence is to divide the high-speed serial bit stream into low-speed parallel bit streams, and then modulate them onto orthogonal subcarriers. Its biggest disadvantage is the high peak-to-average power ratio (PAPR). When the phases of the subcarriers are the same or similar, the superimposed waveform has a higher peak value, and the peak value is much larger than the average value, which will cause the power amplifier to enter the nonlinear region, causing signal distortion and worsening the system bit error rate (BER).

[0003] The main clipping filtering technologies include direct clipping filtering technology, iterative clipping and filtering (ICF) technology, and simplified clipping and filtering (SCF) technology. Direct clipping filtering directly filters the signal after clipping, which will cause the signal peak to rebound. The ICF algorithm reduces the rebound of the peak through multiple iterations, but a large number of IFFT and FFT transformations are required in the iteration, and the calculation complexity is high. The SCF algorithm simulates the noise after multiple clipping by scaling the noise of the first clipping in the frequency domain. One iteration can achieve the PAPR suppression performance after multiple iterations in ICF, but the bit error rate and PAPR suppression performance of the SCF algorithm are still insufficient. Compression and expansion technology is another classic PAPR suppression technology. This method uses the compression and expansion function at the transmitter to change the amplitude of the time domain signal to reduce the PAPR of the signal. The literature proposes a PAPR suppression method based on clipping-noise compression (CNC), which uses compressed noise to correct the clipped signal, reducing the distortion caused by direct clipping. This method achieves better PAPR suppression effect and bit error rate performance than ICF and SCF. At the same time, this method does not require inverse compression and expansion at the receiving end, reducing the complexity of the receiver.

[0004] However, when the existing limiting filtering technology compresses the limiting noise using a compression function, the compression function structure is complex and the compression conditions are strict and inflexible, so it is impossible to control the balance between PAPR suppression performance and bit error rate to meet the performance requirements of the system. Summary of the invention

[0005] In order to solve the problem that when the existing limiting filtering technology compresses the limiting noise using a compression function, the compression function structure is complex and the compression conditions are strict and not flexible enough, so it is impossible to control the balance between PAPR suppression performance and bit error rate to meet the performance requirements of the system, the present invention provides a peak-to-average ratio suppression method, system, device and storage medium for an OFDM system.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A peak-to-average ratio suppression method for an OFDM system comprises the following steps:

[0008] Obtaining the OFDM time domain signal after the inverse fast Fourier transform, and obtaining the limited signal after performing a limiting operation on the time domain signal, and obtaining the limited noise according to the difference between the time domain signal and the limited signal;

[0009] The limiting noise is scaled and multiplied by the noise coefficient to obtain the scaled limiting noise; according to the piecewise compression function and the compression threshold value, the scaled limiting noise is segmentedly compressed while keeping the phase of the noise unchanged, the scaled limiting noise greater than the compression threshold value is subjected to a logarithmic operation compression, and the scaled limiting noise less than the compression threshold value remains unchanged to obtain the compressed noise;

[0010] The compressed noise is added to the limited signal for correction to obtain an OFDM time domain signal with peak-to-average ratio suppressed.

[0011] Furthermore, it also includes scaling the clipping noise after normalizing and denormalizing the clipping noise by the expected value of the clipping noise. The specific formula is as follows:

[0012]

[0013] Where f(n) is the clipping noise and E[] is the expectation.

[0014] Furthermore, the expression of the piecewise compression function is:

[0015]

[0016] G=α·δ

[0017] δ=max(f(n))

[0018] Among them, a and b are compression and expansion coefficients, G is the compression threshold value of segmented compression, δ is the maximum value of the limiting noise amplitude, and α is the threshold parameter.

[0019] Furthermore, the step of obtaining a limited signal after performing a limiting operation on the time domain signal, and obtaining the limited noise according to the difference between the time domain signal and the limited signal comprises:

[0020] Set a threshold value T. For signals whose amplitude exceeds the threshold value, fix its amplitude to the threshold value T. For signals whose amplitude is less than or equal to the threshold value, keep it unchanged. The specific formula is as follows:

[0021]

[0022] Where x(n) is the OFDM time domain signal, y(n) represents the limiting signal, and φ(x) is the phase of the OFDM time domain signal;

[0023] Subtract the time domain signal from the limiting signal to get the limiting noise. The specific formula is as follows:

[0024] f(n)=x(n)-y(n).

[0025] Furthermore, the relevant formula for calculating the OFDM time domain signal after the peak-to-average ratio is suppressed is:

[0026]

[0027] Among them, f c (n) is the compressed noise, and y(n) is the limited signal.

[0028] A peak-to-average ratio suppression system for an OFDM system, comprising:

[0029] The signal preprocessing module is used to obtain the OFDM time domain signal after the inverse fast Fourier transform, and obtain the limited signal after performing a limiting operation on the time domain signal, and obtain the limited noise according to the difference between the time domain signal and the limited signal;

[0030] The noise compression module is used to scale the clipping noise and multiply it by the noise coefficient to obtain the scaled clipping noise; according to the segmented compression function and the compression threshold value, the scaled clipping noise is segmentedly compressed under the condition of keeping the phase of the noise unchanged, the scaled clipping noise greater than the compression threshold value is subjected to a logarithmic operation compression, and the scaled clipping noise less than the compression threshold value remains unchanged to obtain the compressed noise;

[0031] The suppression module is used to obtain an OFDM time domain signal with a peak-to-average ratio suppressed by adding and correcting the compressed noise and the limiting signal.

[0032] A computer device comprises a memory and a processor, wherein the memory stores computer-executable instructions, and the processor executes the computer-executable instructions stored in the memory to implement a peak-to-average ratio suppression method for an OFDM system as described above.

[0033] A computer-readable storage medium is used to store computer-executable instructions. When the computer-executable instructions are executed by a processor, they are used to implement the peak-to-average ratio suppression method of an OFDM system as described above.

[0034] The peak-to-average ratio suppression method of an OFDM system provided by the present invention has the following beneficial effects:

[0035] The invention obtains an OFDM time domain signal after an inverse fast Fourier transform, performs a limiting operation on the time domain signal to obtain a limiting signal, and then obtains limiting noise according to the limiting signal; scales the limiting noise and multiplies it by a noise coefficient to obtain scaled limiting noise; according to a segmented compression function and a compression threshold value, the scaled limiting noise is segmentedly compressed under the condition of keeping the phase of the noise unchanged, the scaled limiting noise greater than the compression threshold value is subjected to a logarithmic operation compression, and the scaled limiting noise less than the compression threshold value remains unchanged to obtain compressed noise; and the compressed noise is used to correct the limiting signal to obtain an OFDM time domain signal after peak-to-average ratio suppression.

[0036] The present invention adopts an enhanced logarithmic operation segmented compression function, and realizes flexible adjustment and compression of system noise by setting a segmented compression threshold value. The size of the threshold value can adjust the proportion of the signal amount that remains unchanged, thereby flexibly controlling the signal integrity and achieving a balance between the system's PAPR suppression capability and the bit error rate. The compression function used in the present invention only needs to perform logarithmic operation compression once, and compared with the existing compression function, the algorithm complexity is also lower and the compression conditions are more relaxed.

[0037] This solution comprehensively solves the problem that when the existing limiting filtering technology uses a compression function to compress the limiting noise, it is impossible to control the balance between PAPR suppression performance and bit error rate to meet the performance requirements of the system due to the complex structure of the compression function and the strict and inflexible compression conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiment of the present invention and its design scheme, the following briefly introduces the drawings required for this embodiment. The drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0039] Figure 1This is the existing OFDM basic modulation and demodulation block diagram;

[0040] Figure 2 It is a schematic diagram of the existing OFDM signal spectrum;

[0041] Figure 3 It is a schematic diagram of the existing direct limiting principle;

[0042] Figure 4 A diagram showing a comparison of compression characteristics of two compression functions according to an embodiment of the method of the present invention;

[0043] Figure 5 A flowchart of a suppression method according to an embodiment of the present invention;

[0044] Figure 6 This is a PAPR comparison diagram of different algorithms with CR=1.4 in an embodiment of the method of the present invention;

[0045] Figure 7 BER comparison diagram of different algorithms with CR=1.4 in the embodiment of the method of the present invention;

[0046] Figure 8 This is a PAPR comparison diagram of different algorithms with CR=1.2 in an embodiment of the method of the present invention;

[0047] Fig. 9 BER comparison diagram of different algorithms for CR=1.2 of the method embodiment of the present invention;

[0048] Fig.10 A PAPR comparison diagram of different segmented compression threshold parameters α in an embodiment of the method of the present invention;

[0049] Fig.11 BER comparison diagram of different segment compression threshold parameters α in the embodiment of the method of the present invention;

[0050] Fig.12 This is a comparison diagram of the power spectra of OFDM signals after being processed by different algorithms in the method embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to enable those skilled in the art to better understand the technical solution of the present invention and implement it, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the scope of protection of the present invention.

[0052] OFDM technology modulates the bit stream onto multiple non-interfering subcarriers to increase data transmission rate and system capacity. The subcarriers must be strictly orthogonal. Figure 1 As shown in the figure, it is the basic modulation and demodulation block diagram of OFDM.

[0053] like Figure 2 As shown, it is a spectrum diagram of an OFDM signal, the number of subcarriers is set to N = 5, and the subcarriers use the same modulation method. Because the subcarriers in the OFDM signal are orthogonal to each other, the maximum value of the spectrum is at the center frequency of each carrier, and the spectrum of other subcarriers at the maximum value of the spectrum is zero.

[0054] In practical applications, continuous signals need to be digitized. Fast Fourier transform (IFFT) can be used to reduce the computational complexity, as shown below:

[0055]

[0056] In order to restore the information of the continuous signal to the greatest extent from the discrete signal, the signal needs to be oversampled L times during the digitization process, and the oversampling factor L is generally taken as ≥ 4. Oversampling is completed by interpolating the frequency domain signal. At this time, the oversampled signal becomes:

[0057]

[0058] The probability of large-amplitude signals in OFDM signals is small, and most signals have small amplitudes, which leads to large amplitude fluctuations in OFDM signals. This is why OFDM has high peaks. The amplitude fluctuation range of the signal is very large. When it exceeds the linear amplification working area of ​​the power amplifier and enters the nonlinear area, it will introduce nonlinear distortion, resulting in a decrease in BER performance and power spectrum (PSD) performance. There are many indicators to measure the amplitude fluctuation of OFDM signals. The peak-to-average power ratio (PAPR) is the most widely used and simple to calculate. PAPR is expressed as:

[0059]

[0060] Usually dB is used as the unit of PAPR, and the PAPR expression is:

[0061]

[0062] Complementary Cumulative Distribution Function (CCDF) is usually used to measure the size of PAPR, which indicates the probability that the signal is greater than a certain threshold value PAPR0. It can be specifically expressed as:

[0063] CCDF=P(PAPR>PAPR0)

[0064] =1-P(PAPR≤PAPR0)(5)

[0065] =1-(1-exp(-PAPR0))N

[0066] When L times oversampling is performed, the CCDF function becomes:

[0067] CCDF = 1-(1-exp(-PAPR0)) LN (6)

[0068] Direct limiting has simple operation steps and excellent PAPR suppression effect. However, its biggest disadvantage is that it will produce distortion, which will seriously affect the performance of the communication system. Figure 3 shown.

[0069] First, a threshold value T is set. For signals whose amplitude exceeds the threshold value, the amplitude is fixed to the threshold value T. For signals whose amplitude is less than or equal to the threshold value, it remains unchanged. During the limiting process, the phase is always kept unchanged. The limiting process is shown in equation (7), where x(n) is the OFDM time domain signal, y(n) represents the signal after limiting, and the phase of the OFDM time domain signal is represented by φ(x).

[0070]

[0071] Generally, the PAPR suppression capability and bit error rate are balanced by selecting an appropriate clipping ratio (CR). The clipping ratio CR is shown in formula (8), P AVG is the average signal power.

[0072]

[0073] Analyzing the expression of the clipping ratio CR, we can get that the larger the CR is, the larger the threshold value T is, the fewer signals that exceed the threshold value and need to be clipped, the less signal distortion is, the lower the bit error rate is, and the weaker the PAPR suppression performance is. On the contrary, the more signals are clipped, the higher the bit error rate is, and the better the PAPR suppression performance is.

[0074] Direct limiting technology is simple and efficient, but its in-band and out-of-band distortion will cause system performance to deteriorate. Generally, filtering the limited signal can effectively remove out-of-band distortion, reduce interference to other frequency band signals, and improve the system's bit error rate performance. The transfer function of a typical filter that removes out-of-band distortion is as follows, where N is the number of subcarriers and L is the oversampling factor:

[0075]

[0076] The signal in the filter passband remains unchanged, and the out-of-band signal is filtered out. The advantage of post-limiting filtering is that out-of-band interference is removed, but the disadvantage is that the filtering operation will cause the peak of the signal to rise, and the peak-to-average ratio suppression performance of the signal will decrease. In general, iterative limiting filtering, that is, ICF peak-to-average ratio suppression technology, can be adopted. Repeated limiting filtering can meet the PAPR requirements and out-of-band radiation requirements of the system. The disadvantage is that the calculation complexity is high. Because filtering the signal requires the signal to be transformed into the frequency domain through FFT, and then transformed into the time domain through IFFT after filtering for the next limiting, and the filtering continues in the same way. The signal is converted back and forth between the frequency domain and the time domain, which requires multiple FFT and IFFT operations.

[0077] The iterative clipping filter algorithm ICF requires a lot of computational complexity. Some studies have improved this and proposed a simplified clipping filter SCF. The SCF method points out that the clipping noise obtained by the first clipping can be multiplied by a coefficient to simulate the clipping noise after multiple iterations. This can reduce the number of iterations in ICF to 1, reduce the number of FFT and IFFT operations, and reduce computational complexity. The noise coefficient is as follows:

[0078]

[0079] K is the number of iterations in ICF, where β is as in equation (11), T is the limit threshold, and σ is the standard deviation of the OFDM signal x(n).

[0080]

[0081] The limiting noise is the difference between the original signal and the signal after limiting, that is, the limiting noise z(n) = x(n) - y(n). The SCF method performs FFT on the limiting noise z(n) and converts it to the frequency domain to obtain the frequency domain form of the limiting noise Z(k). Formula (9) is used to filter out the out-of-band noise from Z(k) to obtain the in-band noise Will Multiplying by the noise coefficient λ, we can get the clipping noise after multiple simulation iterations. The frequency domain signal after simplified clipping filtering is:

[0082]

[0083] Finally, the frequency domain Convert to time domain through IFFT to get the final transmission signal The SCF algorithm only needs one additional FFT and one IFFT to obtain the final transmitted limiting filtered signal, while ICF requires multiple times. Compared with ICF, it saves a lot of computational complexity.

[0084] ICF technology achieves the desired peak-to-average ratio and communication performance requirements through multiple iterations of limiting filtering. SCF makes improvements on this basis and reduces the computational complexity. However, SCF still requires an additional FFT and IFFT module to complete limiting filtering, and its PAPR suppression capability and bit error rate can still be improved. A new limiting noise processing method is proposed in a document. The document proposes to compress the limiting noise instead of compressing the signal itself. The compression algorithm CNC based on limiting noise proposed in the document first performs a limiting operation on the OFDM time domain signal according to formula (7), and then obtains the limiting noise:

[0085]

[0086] f(n)=x(n)-y(n)(14)

[0087] The limiting noise f(n) is scaled in the time domain and multiplied by the noise coefficient λ to approximate the noise after multiple iterations of limiting. Then, the scaled limiting noise is compressed using the μ-law compression function. The difference between this algorithm and the SCF algorithm is that this algorithm operates on the signal in the time domain, while the SCF needs to transform the signal into the frequency domain and then transform it into the time domain. The μ-law compression function is shown in (15):

[0088]

[0089] When compressing noise, the scaled noise fed into the compression function is first normalized by equation (16), and then the compressed noise is denormalized to obtain the noise compression function as shown in equation (17), maintaining the noise phase φ f (n) remains unchanged. The scaled noise is compressed using equation (17) to obtain the compressed noise.

[0090]

[0091] Finally, the compressed noise is used to correct the limited signal, and the final transmitted time domain signal is obtained according to the following formula (18):

[0092]

[0093] Method Embodiment

[0094] The present invention provides a peak-to-average ratio suppression method for an OFDM system, denoted as MCNC. The proposed improved algorithm adopts an enhanced logarithmic piecewise compression function such as formula (19) when compressing the limiting noise. Wherein, a and b are compression coefficients, which jointly determine the shape of the compression function, generally taking a·b=1, G is the threshold value of the piecewise compression, G=α·δ, α∈[0,1), δ is the maximum value of the limiting noise amplitude |f(n)|, that is, δ=max(f(n)).

[0095]

[0096] The following is a comparative analysis of the compression characteristics of the μ-law compression function and the proposed enhanced logarithmic piecewise compression function. In order to maintain consistency with the CNC algorithm, it is assumed that α=0 in the proposed compression function, at which time the compression threshold G=0, and all noise is compressed without piecewise compression processing.

[0097] Figure 4 The compression characteristics of the two compression functions are compared. The black solid line in the figure is the uncompressed signal, the three dotted lines are the compression curves in the CNC method, and the three solid lines are the compression curves of the MCNC of the present invention. For the μ-law compression function in the CNC method, its compression interval is x>1. In order to make the limiting noise in the compression interval, the CNC method normalizes the limiting noise by formula (16). From the perspective of the probability of peak occurrence, the amplitude of the OFDM signal obeys the Rayleigh distribution, and the probability of peak occurrence is small. Because the noise f(n) has few non-zero components, E[|f(n)|]<|f(n)|, then the probability of {|f(n)| / E[|f(n)|]}>1 is high. Through normalization, the noise falls in the compression interval. For the compression function in the MCNC of the present invention, the compression interval is x>0, and the limiting noise all works in the compression interval.

[0098] Generally, the amplitude of the limiting noise is small. In order to better compress the noise, the MCNC method of the present invention also adopts the normalization method of formula (16) and denormalization for the limiting noise, keeping the phase unchanged during compression. The final compression function is shown in formula (20):

[0099]

[0100] Considering the difficulty and complexity of implementation, the μ-law compression function is a division operation of two logarithms, while the proposed enhanced logarithmic segmented compression function has only one logarithmic operation. Therefore, the compression function structure of the present invention is simpler, the computational complexity is lower, and it is easier to implement.

[0101] From the perspective of compression conditions, the limiting noise in the MCNC method of the present invention all works in the compression interval, and the CNC method only compresses the noise with {|f(n)| / E[|f(n)|]}>1. Although the probability of {|f(n)| / E[|f(n)|]}>1 is high, as the limiting ratio CR decreases, the peak value to be cut becomes larger, and this condition becomes more stringent. The compression function proposed in the MCNC of the present invention has a looser value condition for |f(n)| / E[|f(n)|], and the compression interval is x>0. Even if {|f(n)| / E[|f(n)|]}<1, the normalized noise still works in the compression interval.

[0102] From the perspective of flexibility, the enhanced logarithmic piecewise compression function proposed in the present invention adopts the form of piecewise compression. The piecewise compression is performed by setting the piecewise compression threshold, which increases flexibility. Different threshold values ​​can be adjusted to control the balance between PAPR suppression performance and bit error rate to meet the performance requirements of the system. When α in the threshold value is 0, then G = 0. At this time, all signals are compressed without piecewise compression processing, which is consistent with the CNC method. When α≠0, then G≠0, then piecewise compression processing is performed, and signals greater than the threshold value are compressed, and signals less than the threshold value remain unchanged. If the set piecewise compression threshold value is high and the number of compressed signals is small, the bit error performance of the signal is good. Otherwise, the bit error performance of the signal is poor.

[0103] The MCNC method flow chart of the present invention is as follows Figure 5 As shown, the steps are as follows:

[0104] (1) The OFDM time domain signal x(n) after IFFT is limited according to the limiting ratio CR and the limiting threshold T to obtain a limited signal, and the limiting noise is obtained according to the difference between the original signal x(n) and the limited signal y(n).

[0105] (2) The limiting noise f(n) obtained after limiting is scaled and multiplied by the noise coefficient λ to obtain scaled limiting noise λf(n).

[0106] (3) Compress the noise λf(n) according to the compression parameters a, b, α, and obtain the compressed noise f c (n).

[0107] (4) The compressed noise f c (n) Correct the limited signal y(n) to obtain the final transmission signal

[0108]

[0109] The MCNC algorithm of the present invention is compared and simulated with other algorithms, including the CNC algorithm, the ICF algorithm, and the SCF algorithm. In order to maintain consistency with the CNC algorithm, it is first assumed that α=0 in the compression function of the MCNC algorithm of the present invention, at which time the compression threshold value G=0, and all noises are compressed without performing segmented compression processing. The CNC algorithm and the MCNC algorithm of the present invention each select two groups of different compression and expansion parameters for comparison. Figure 6 This is a comparison chart of PAPR suppression performance of different algorithms when the limiting ratio CR = 1.4. Figure 7 When the limit ratio CR = 1.4, the BER performance comparison of different algorithms under different signal-to-noise ratios through a Gaussian white noise channel. Other simulation parameter settings are shown in Table 1:

[0110] Table 1 Simulation experiment parameters

[0111]

[0112]

[0113] pass Figure 7 It can be obtained that the bit error rate curve of the CNC algorithm when μ=2 is basically consistent with the bit error rate curve of the MCNC algorithm of the present invention when a=1, b=1. -3 When μ=2, the PAPR value of the CNC algorithm is 4.87dB, while when a=1, b=1, the PAPR value of the MCNC algorithm of the present invention is 4.83dB, which is 0.05dB better than the CNC algorithm. Figure 7 The bit error rate curve of the CNC algorithm when μ=8 is substantially consistent with the bit error rate curve of the MCNC algorithm of the present invention when a=0.4 and b=2.5. Figure 6 In Pr(PAPR>PAPR0)=10 -3 When μ=8, the PAPR value of the CNC algorithm is 4.49dB, while when a=0.4, b=2.5, the PAPR value of the MCNC algorithm of the present invention is 4.30dB, which is 0.19dB better than the CNC algorithm. By comparing two groups of different compression parameters, under the condition of similar bit error rate, the PAPR performance of the MCNC algorithm of the present invention is better than that of the CNC algorithm.

[0114] Figure 6 In Pr(PAPR>PAPR0)=10 -3 When , the PAPR value of ICF is 5.23dB, the PAPR value of SCF is 5.83dB, and the PAPR value of the original OFDM signal is 10.82dB. The PAPR suppression capabilities of CNC and the MCNC method of the present invention are better than those of the SCF algorithm and the ICF algorithm. Figure 7 In the comparison of bit error rate performance, ICF has the worst bit error rate performance, SCF bit error rate is slightly better than ICF, and CNC and the MCNC of the present invention have better bit error rate performance than ICF and SCF. Because ICF and SCF algorithms directly cut off the peak value of the signal, and directly fix the signal exceeding the threshold value as the threshold value, while CNC and the MCNC algorithm of the present invention compress the limiting noise, and then use the obtained compressed limiting noise to correct the limiting signal. ICF and SCF methods have more signal loss and relatively high bit error rate, while CNC and the MCNC algorithm of the present invention have less signal loss, so their bit error rate is also relatively low.

[0115] Figure 8 is the PAPR suppression performance of different algorithms when CR=1.2, Fig. 9The BER performance comparison chart of different signal-to-noise ratios of Gaussian white noise channel when CR=1.2. At this time, two groups of μ=2 and μ=8 are set in CNC, and two groups of a=0.5, b=2 and a=1, b=1 are set in MCNC of the present invention. Other parameters are shown in Table 4-1. Figure 8 and Fig. 9 The comparison between Figure 6 and Figure 7 The effect is similar when CR=1.4. When the bit error rates of the MCNC and CNC of the present invention are similar, the PAPR suppression effect of the MCNC of the present invention is better than that of the CNC. For example, when the bit error rates of the CNC algorithm with μ=8 and the MCNC algorithm of the present invention with a=0.5 and b=2 are similar, when Pr(PAPR>PAPR0)=10 -3 The PAPR value of the CNC algorithm is 4.06dB, and the PAPR value of the MCNC algorithm of the present invention is 3.93dB. The PAPR suppression effect of the MCNC of the present invention is 0.13dB higher than that of the CNC.

[0116] Figure 8 The PAPR suppression capability of ICF when CR=1.2 is similar to that of CNC with μ=2 and MCNC with a=1, b=1. As the noise compression degree in CNC and MCNC algorithms increases, the PAPR suppression effect of CNC with μ=8 and MCNC with a=0.5, b=2 is better than that of ICF. Fig. 9 The bit error rate in the analysis is 10 -4 , the required signal-to-noise ratio of the original OFDM signal is 5.9dB. The required signal-to-noise ratio of CNC with μ=2 and MCNC with a=1, b=1 is about 6.5dB, the required signal-to-noise ratio of CNC with μ=8 and MCNC with a=0.5, b=2 is about 6.8dB, the required signal-to-noise ratio of the SCF algorithm is about 8.6dB, and the required signal-to-noise ratio of the ICF algorithm is about 10dB. It can be concluded that for CR=1.2, although the PAPR suppression effect of the ICF method is close to that of CNC with μ=2 and MCNC algorithm with a=1, b=1, the bit error rate performance of the two sets of CNC with different compression and expansion parameters and the MCNC algorithm of the present invention are better than that of ICF. Comparison Figure 7 and Fig. 9 The bit error rate performance under different limiting ratios is 10 -4When CR=1.4, the required SNR of ICF is 7.6dB, and the required SNR of CNC with μ=8 and MCNC with a=0.4, b=2.5 is about 6.4dB; the required SNR of ICF with CR=1.2 is 10dB, and the required SNR of CNC with μ=8 and MCNC with a=0.5, b=2 is about 6.8dB. Compared with CR=1.2, the bit error rate performance of ICF decreases by 2.4dB for CR=1.4. The bit error rate performance of CNC and MCNC decreases by 0.4dB. For different limiting ratios CR, the required SNR of ICF increases more under the same bit error rate. Therefore, the MCNC and CNC algorithms are superior to ICF in terms of PAPR suppression performance and bit error rate.

[0117] By analyzing the PAPR suppression performance and bit error rate performance of different algorithms at different limiting ratios, it is verified that the MCNC of the present invention has better PAPR suppression capability. In order to be consistent with the CNC algorithm, the above comparative experiment feeds all noise into the compression function and sets the segmentation threshold of noise compression to zero, i.e., no segmentation compression is performed. The segmentation characteristics of the MCNC of the present invention are simulated and analyzed below.

[0118] Fig.10 The PAPR performance comparison diagram of the MCNC algorithm of the present invention for different segment compression threshold parameters α. During simulation, the limit ratio CR = 1.2, a = 1, b = 1 in the enhanced logarithmic segment compression function. When Pr (PAPR > PAPR0) = 10 -3 When α=0, α=0.4 and α=0.7 are used, the PAPR is 4.51dB, 7.10dB and 9.08dB respectively, which are 6.37dB, 3.78dB and 1.80dB higher than the original signal of 10.88dB. α=0 is equivalent to the threshold value G=α·δ=0, δ is the maximum value of the limiting noise f(n), δ=max(f(n)), which is equivalent to no segmentation, and all the limiting noise signals are sent to the compression function. When α≠0, different degrees of segmented compression are performed. As α increases, the threshold value G increases, the compressed signal is small, and its PAPR value is relatively high.

[0119] Fig.11 The BER performance comparison diagram of the MCNC algorithm of the present invention with different compression threshold parameters α is shown in the figure. -4When the original signal, α=0.7, α=0.4, α=0, the required signal-to-noise ratio of the MCNC algorithm is 5.7dB, 6.1dB, 6.5dB, and 6.9dB. As α decreases, its bit error rate performance deteriorates. Because α increases. The segmentation threshold value G is high, more signals remain unchanged, the signal integrity is strong, and its bit error rate performance is good. The proposed enhanced logarithmic piecewise function has greater flexibility in noise compression. Setting different segmentation threshold parameters α can adjust the balance between PAPR suppression performance and bit error rate.

[0120] Fig.12 The power spectrum comparison diagram of OFDM signals after being processed by different PAPR suppression algorithms shows that the out-of-band spectrum of ICF and SCF methods is basically consistent with the original signal, because ICF and SCF are filtered to filter out the out-of-band noise. For the CNC method, two groups of μ=2 and μ=8 are set, and the MCNC improved method of the present invention is proposed, with segmented compression threshold parameter α=0, and two groups of a=0.5, b=2 and a=1, b=1 are set. The out-of-band spectrum of the MCNC method of the present invention is basically consistent with the out-of-band spectrum performance of CNC. The MCNC and CNC methods are not filtered, but only compress the noise, and the out-of-band spectrum is suppressed, unlike the filtering method that completely eliminates it, and the out-of-band spectrum performance is slightly worse than that of ICF and SCF. However, the out-of-band spectrum of the MCNC and CNC methods is acceptable, and the literature also verifies that the spectrum of the CNC method meets the 802.11ax signal and HE 20MHz spectrum. The ICF and SCF algorithms filter after limiting, resulting in peak recovery, poor PAPR suppression performance, and poor bit error rate performance.

[0121] ICF uses multiple iterations of limiting filtering, which has high computational complexity. On this basis, SCF reduces the complexity by scaling the limiting noise and only requires one iteration. The CNC method and the improved MCNC method only require one IFFT transformation, and the complexity of the CNC and MCNC methods is lower than that of ICF and SCF. In addition, the ICF and SCF algorithms are filtered, which leads to a peak rebound and a decrease in PAPR suppression performance, which is weaker than CNC and MCNC. At the same time, both ICF and SCF are directly limited, and the signal integrity is weak. The CNC and MCNC methods increase the signal integrity by compensating the limited signal with the compressed limiting noise, so the bit error rate performance of CNC and MCNC is better than that of ICF and SCF.

[0122] The CNC method uses typical μ-law compression, and the compression interval is x>1. Only {|f(n)| / E[|f(n)|]}>1 can be compressed. The MCNC method of the present invention uses a simpler enhanced logarithmic piecewise compression function, and the compression interval is x>0, and the compression condition is more relaxed. Through comparative analysis of the PAPR suppression performance and bit error rate performance of CNC and the MCNC of the present invention, when the bit error rates of the two methods are basically the same, the PAPR suppression performance of the MCNC method of the present invention is better. In addition, the MCNC method of the present invention can also flexibly adjust the PAPR suppression capability and bit error rate through piecewise compression. In summary, the comprehensive performance of the MCNC of the present invention is better.

[0123] System Example

[0124] The present invention provides a peak-to-average ratio suppression system for an OFDM system, comprising:

[0125] The signal preprocessing module is used to obtain the OFDM time domain signal after the inverse fast Fourier transform, and obtain the limited signal after performing a limiting operation on the time domain signal, and obtain the limited noise according to the difference between the time domain signal and the limited signal.

[0126] The noise compression module is used to scale the limiting noise and multiply it by the noise coefficient to obtain the scaled limiting noise; according to the piecewise compression function and the compression threshold value, the scaled limiting noise is segmentedly compressed while keeping the phase of the noise unchanged, the scaled limiting noise greater than the compression threshold value is subjected to a logarithmic operation compression, and the scaled limiting noise less than the compression threshold value remains unchanged to obtain the compressed noise.

[0127] The suppression module is used to obtain an OFDM time domain signal with a peak-to-average ratio suppressed by adding and correcting the compressed noise and the limiting signal.

[0128] Device Embodiment

[0129] The present invention provides a computer device, including a memory and a processor, wherein the memory stores computer execution instructions, and the processor executes the computer execution instructions stored in the memory to implement a peak-to-average ratio suppression method for an OFDM system as described above, wherein the method has been described in detail in the method embodiment and will not be repeated here.

[0130] Storage Medium Embodiments

[0131] The present invention provides a computer-readable storage medium, which is used to store computer execution instructions. When the computer execution instructions are executed by a processor, they are used to implement a peak-to-average ratio suppression method for an OFDM system as described above, wherein the method has been described in detail in the method embodiment and will not be repeated here.

[0132] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0133] The present invention is described with reference to flowcharts and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as a combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0136] It should be noted that the above-described specific implementation methods can enable those skilled in the art to more fully understand the invention, but do not limit the invention in any way. Therefore, although the present specification and embodiments have described the invention in detail, those skilled in the art should understand that the invention can still be modified or replaced by equivalents; and all technical solutions and improvements that do not deviate from the spirit and scope of the invention are included in the protection scope of the patent for the invention. Any figure mark in the claims should not be regarded as limiting the claims involved.

Claims

1. A peak-to-average ratio suppression method for an OFDM system, characterized in that: The steps include: Obtaining the OFDM time domain signal after the inverse fast Fourier transform, and obtaining the limited signal after performing a limiting operation on the time domain signal, and obtaining the limited noise according to the difference between the time domain signal and the limited signal; After normalizing and denormalizing the limiting noise by the expected value of the limiting noise, the limiting noise is scaled and multiplied by the noise coefficient to obtain the scaled limiting noise; according to the piecewise compression function and the compression threshold value, the scaled limiting noise is piecewise compressed under the condition of keeping the phase of the noise unchanged, the scaled limiting noise greater than the compression threshold value is subjected to a logarithmic operation compression, and the scaled limiting noise less than the compression threshold value remains unchanged to obtain the compressed noise; After adding the compressed noise to the limited signal for correction, the OFDM time domain signal with peak-to-average ratio suppressed is obtained; Among them, the limiting noise is scaled, and the specific formula is as follows: Where, f(n) is the clipping noise, E[] is the expectation, and x represents the value after the clipping noise f(n) is normalized; The expression of the piecewise compression function is: G=α·δ δ=max(f(n)) Among them, a and b are compression coefficients, G is the compression threshold value of segmented compression, δ is the maximum value of the limiting noise amplitude, α is the threshold parameter, and φ is the phase; φ f (n) represents the phase of the limiting noise f(n).

2. The peak-to-average ratio suppression method of an OFDM system according to claim 1, characterized in that: The step of obtaining a limited signal after performing a limiting operation on the time domain signal, and obtaining the limited noise according to the difference between the time domain signal and the limited signal comprises: Set a threshold value T. For signals whose amplitude exceeds the threshold value, fix its amplitude to the threshold value T. For signals whose amplitude is less than or equal to the threshold value, keep it unchanged. The specific formula is as follows: Where x(n) is the OFDM time domain signal, y(n) represents the limiting signal, φ(x) is the phase of the OFDM time domain signal, and j is a complex unit; Subtracting the time domain signal from the limiting signal yields the limiting noise. The specific formula is as follows: f(n)=x(n)-y(n).

3. The peak-to-average ratio suppression method of an OFDM system according to claim 1, characterized in that: The relevant formula for calculating the OFDM time domain signal after the peak-to-average ratio is: Among them, f c (n) is the compressed noise, and y(n) is the limited signal.

4. A peak-to-average ratio suppression system for an OFDM system, characterized in that: include: The signal preprocessing module is used to obtain the OFDM time domain signal after the inverse fast Fourier transform, and obtain the limited signal after performing a limiting operation on the time domain signal, and obtain the limited noise according to the difference between the time domain signal and the limited signal; The noise compression module is used to scale the clipping noise and multiply it by the noise coefficient to obtain the scaled clipping noise; according to the piecewise compression function and the compression threshold value, the scaled clipping noise is segmentedly compressed while keeping the phase of the noise unchanged, the scaled clipping noise greater than the compression threshold value is subjected to a logarithmic operation compression, and the scaled clipping noise less than the compression threshold value remains unchanged to obtain the compressed noise; wherein, the specific formula for scaling the clipping noise is as follows: Where, f(n) is the clipping noise, E[] is the expectation, and x represents the value after the clipping noise f(n) is normalized; The expression of the piecewise compression function is: G=α·δ δ=max(|f(n)|) Among them, a and b are compression coefficients, G is the compression threshold value of segmented compression, δ is the maximum value of the limiting noise amplitude, α is the threshold parameter, and φ is the phase; φ f (n) represents the phase of the limiting noise f(n); The suppression module is used to obtain an OFDM time domain signal with a peak-to-average ratio suppressed by adding and correcting the compressed noise and the limiting signal.

5. A computer device comprising a memory and a processor, characterized in that: The memory stores computer-executable instructions, and the processor executes the computer-executable instructions stored in the memory to implement a peak-to-average ratio suppression method for an OFDM system according to any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement a peak-to-average ratio suppression method for an OFDM system as described in any one of claims 1-3.

Images