阅读说明：本技术 基于加速近似梯度算法优化tr方法抑制papr过程的方法 (Method for optimizing TR (transmitter-receiver) method to inhibit PAPR (peak-to-average power ratio) process based on accelerated approximation gradient algorithm ) 是由 杨国 李杨 吴文 钱玉文 刘奇 于 2019-11-14 设计创作，主要内容包括：本发明公开了一种基于加速近似梯度算法优化TR方法抑制PAPR过程的方法,属于宽带通信领域,包括以下步骤：利用傅里叶变换将正交频分复用OFDM系统的时域信号转换为频域信号；获取时域信号与频域信号之间的映射关系F；利用载波预留TR方法产生抑制PAPR的信号,并结合加速近似梯度算法APGM抑制峰均比PAPR。本发明通过加速近似梯度算法APGM优化TR方法抑制PAPR的过程,不需要切削与滤波步骤,简单有效,此外无需计算最优的切削阈值,且不同的切削阈值不会影响最终的抑制效果。本发明不仅能够降低抑制PAPR的复杂度,实现有效抑制PAPR,且还能够取得较好的BER效果。(The invention discloses a method for inhibiting PAPR process based on accelerated approximate gradient algorithm optimization TR method, belonging to the field of broadband communication and comprising the following steps: converting a time domain signal of the orthogonal frequency division multiplexing OFDM system into a frequency domain signal by utilizing Fourier transform; acquiring a mapping relation F between a time domain signal and a frequency domain signal; and generating a signal for inhibiting the PAPR by utilizing a carrier reservation TR method, and inhibiting the PAPR by combining an accelerating approximate gradient algorithm APGM. The method for restraining the PAPR through the accelerated approximate gradient algorithm APGM optimization TR method is simple and effective, cutting and filtering steps are not needed, in addition, the optimal cutting threshold value does not need to be calculated, and the final restraining effect cannot be influenced by different cutting threshold values. The invention can not only reduce the complexity of PAPR suppression and realize effective PAPR suppression, but also can obtain better BER effect.)

1. A method for optimizing a TR method to inhibit a PAPR process based on an accelerated approximation gradient algorithm is characterized by comprising the following steps:

step 1, converting a time domain signal X of an Orthogonal Frequency Division Multiplexing (OFDM) system into a frequency domain signal X by utilizing Fourier transform;

step 2, obtaining a mapping relation F between the time domain signal X and the frequency domain signal X;

step 3, generating PAPR restraining signal by using carrier wave reserving TR method

2. The method for optimizing TR method based on accelerated approximation gradient algorithm to suppress PAPR process according to claim 1, wherein step 2 obtains the mapping relationship F between the time domain signal X and the frequency domain signal X by the following formula:

x＝FX

in the formula, F represents a mapping relation of an inverse fast fourier transform matrix with the size of JN × N, and elements in the mth row and the jth column in the matrix are:

3. The method for optimizing TR method based on accelerated approximation gradient algorithm to suppress PAPR process as claimed in claim 1, wherein step 3 utilizes the TR method of carrier reservation to generate PAPR suppressing signal

step 3-1, generating PAPR inhibiting signal by using carrier wave reserved TR method

where D is the peak reduction signal in the frequency domain: d ═ D₀,D₁,D₂,...,D_N-1]^T，D_iFor the peak reduction signal on the ith subcarrier frequency domain, the value of i is 0 to N-1;

step 3-2, PAPR based on time domain signal x, using signal

the PAPR of the original time domain signal x is:

wherein | × | non-conducting phosphor² _∞Representing the peak power of the signal "+", E [ + ]² ₂]Represents the average power of the signal ";

step 3-3, optimizing the inhibition process of the step 3-2:

PAPR suppression on demand acquisitionMinimum value of (d):

obtaining

the constraints for adding D are:

||D||² ₂≤β

wherein β represents the amplitude of the maximum allowed subcarrier signal;

thereby converting the suppression process of the optimization step 3-2 into an optimized quadratic constraint quadratic programming problem, the formula of which is:

s.t.||D||² ₂≤β

step 3-4, converting the optimized quadratic constraint quadratic programming problem into an optimized following formula:

in the formula, lambda is a scale factor;

3-5, replacing x + FD with z, and carrying out the step 3-4

step 3-6, ordering:

h(z)＝λ||z||² _∞

g(z)＝||F^Hz-X||² ₂

step 3-7, optimizing h (λ | | | z | |) through APGM algorithm² _∞The process of (1);

step 3-8, optimizing g (z) | | | F by using APGM algorithm^Hz-X||² ₂The obtaining process of (1).

4. The method for optimizing TR method based on accelerated approximation gradient algorithm to suppress PAPR process according to claim 3, wherein step 3-7 optimizes h (z) λ | | | z | | tormentia using APGM algorithm² _∞The obtaining process specifically comprises the following steps:

calculating h (z) | | | z | | non-conducting phosphor by using APGM algorithm² _∞Is close to the optimal solution of the factor u, the formula used is:

wherein gamma is a step factor;

the specific process comprises the following steps:

step 3-7-1, initializing lambda and gamma;

3-7-2, performing descending order arrangement on the z values to generate a vector s;

step 3-7-3, enabling i to be 1;

3-7-4, calculating:

in the formula, s_π(k)Is the kth element in the vector s, s_π(max)Is the largest element in the vector s;

and 3-7-5, judging the magnitude relation of i and j, and if i is not more than j, outputting the optimal solution of u as follows:

otherwise, judging whether i is less than or equal to JN, if so, increasing i by 1, and executing the steps from 3-7-4 to 3-7-5; otherwise, the obtaining process is finished.

5. The method for optimizing TR method based on accelerated approximation gradient algorithm to suppress PAPR process according to claim 3, wherein step 3-8 utilizes APGM algorithm to optimize g (z) | | F^Hz-X||² ₂The obtaining process specifically comprises the following steps:

using APGM algorithm to obtain g (z) | | F^Hz-X||² ₂The specific process comprises the following steps:

step 3-8-1, let j equal to 1, initialize x^j、y^j、γ^j-1、β，β∈(0,1)；

Step 3-8-2, calculating z:

step 3-8-3, based on the z value calculated in step 3-8-2, obtaining g (z):

g(z)＝||F^Hz-X||² ₂

obtaining

Step 3-8-4, judging g (z) and

otherwise, judging whether j is larger than JN, if so, ending the solving process; otherwise, let j increment by 1 and update x^j、y^jThe formulas are respectively as follows:

y^j＝x^j-1+ω^j(x^j-x^j-1)

in the formula, ω^jE [0,1) represents the jth acceleration element;

then, step 3-8-2 to step 3-8-4 are performed.

Technical Field

The invention belongs to the field of broadband communication, and particularly relates to a method for inhibiting PAPR (peak-to-average power ratio) process based on an accelerated approximation gradient algorithm optimized TR (transmitter-receiver) method.

Background

Wireless communication technology has developed rapidly in recent years. In the 80 s of the 20 th century, the first generation of wireless communication technology (1G) centered on FDMA and analog technologies, providing only voice services, appeared. By the beginning of the 90 s, a second generation mobile communication system (2G) using CDMA technology and TDMA technology as core technologies appeared. In the late 90 s, 2.5G communication systems represented by packet radio service (GPRS) and high speed circuit data switching service (HSCSD) were developed, which increased the transmission rate of wireless data and network capacity. The third generation mobile communication system (3G) still uses the network structure of the second generation mobile communication system. CDMA is used as a core technology and supports information services such as voice, image, file transmission, web browsing and the like. Since the 21 st century, with the further development of technology and the great demand of users for wireless access to the internet and wireless multimedia data services, 4G systems have been promoted in various countries. The Orthogonal Frequency Division Multiplexing (OFDM) technology is a novel digital modulation technology, and can solve the problems of multipath fading and the like in a wireless mobile environment. Meanwhile, the method has the defects of sensitivity to phase noise, large peak-to-average ratio and the like.

The accumulation of multiple subcarriers in OFDM results in a large peak signal, and therefore requires a large linear dynamic range for the power amplifier, which increases the cost of the high power amplifier and reduces the efficiency of the power amplifier. If the linear dynamic range of the high power amplifier is exceeded, in-band distortion and out-of-band dispersion will result. The suppression of the PAPR is a key technology of an OFDM system and has important significance.

At present, three algorithms for restraining PAPR exist, the first algorithm is the simplest and easily-realized amplitude limiting algorithm, but the nonlinear processing of signals reduces the error rate performance of the system, so the research of the algorithms focuses on the error rate performance of the system after amplitude limiting. In order to reduce noise and prevent peak regeneration, an algorithm combining repeated amplitude limiting and filtering is generally adopted, but the algorithm is complex. The second encoding algorithm is a technique for suppressing PAPR of a signal without distortion, such as block coding and golay complementary sequence. The signal transmission with the smallest PAPR value among them is selected by coding the combination of data sequences that can be used for transmission. However, such algorithms are only suitable for systems with a small number of carriers and are therefore rarely used. The third is a probabilistic class of algorithms, usually TR, SLM and PTS. The PTS algorithm divides an input data vector into a plurality of mutually non-overlapping sub-vectors, multiplies sub-carriers in the sub-vectors by a same rotation factor, and reduces the PAPR by selecting the rotation factor. But the computational complexity is high.

Disclosure of Invention

The invention aims to provide a method for reducing the computational complexity while realizing effective PAPR suppression.

The technical solution for realizing the purpose of the invention is as follows: a method for optimizing a TR method to inhibit a PAPR process based on an accelerated approximation gradient algorithm comprises the following steps:

step 1, converting a time domain signal X of an Orthogonal Frequency Division Multiplexing (OFDM) system into a frequency domain signal X by utilizing Fourier transform;

step 2, obtaining a mapping relation F between the time domain signal X and the frequency domain signal X;

step 3, generating PAPR restraining signal by using carrier wave reserving TR method

And inhibiting the PAPR by combining an APGM (approximate acceleration gradient algorithm).

Further, in step 2, obtaining a mapping relationship F between the time domain signal X and the frequency domain signal X, the formula is:

x＝FX

in the formula, F represents a mapping relation of an inverse fast fourier transform matrix with the size of JN × N, and elements in the mth row and the jth column in the matrix are:

where J is the oversampling factor and N is the number of subcarriers in the OFDM system.

Further, step 3, generating PAPR-suppressing signal by using carrier reservation TR methodAnd inhibiting the peak-to-average power ratio (PAPR) by combining an accelerated approximate gradient Algorithm (APGM), specifically comprising the following steps:

step 3-1, generating PAPR inhibiting signal by using carrier wave reserved TR method

Comprises the following steps:

where D is the peak reduction signal in the frequency domain: d ═ D₀,D₁,D₂,...,D_N-1]^T，D_iFor the peak reduction signal on the ith subcarrier frequency domain, the value of i is 0 to N-1;

step 3-2, PAPR based on time domain signal x, using signal

The suppression PAPR is as follows:

the PAPR of the original time domain signal x is:

wherein | × | non-conducting phosphor² _∞Representing the peak power of the signal "+", E [ + ]² ₂]Represents the average power of the signal ";

step 3-3, optimizing the inhibition process of the step 3-2:

PAPR suppression on demand acquisition

Minimum value of (d):

obtaining

Obtaining the minimum value of D:

the constraints for adding D are:

||D||² ₂≤β

wherein β represents the amplitude of the maximum allowed subcarrier signal;

thereby converting the suppression process of the optimization step 3-2 into an optimized quadratic constraint quadratic programming problem, the formula of which is:

m_Din||x+FD||² _∞

s.t.||D||² ₂≤β

step 3-4, converting the optimized quadratic constraint quadratic programming problem into an optimized following formula:

in the formula, lambda is a scale factor;

3-5, replacing x + FD with z, and carrying out the step 3-4

To convert to:

step 3-6, ordering:

h(z)＝λ||z||² _∞

g(z)＝||F^Hz-X||² ₂

step 3-7, optimizing h (λ | | | z | |) through APGM algorithm² _∞The process of (1);

step 3-8, optimizing g (z) | | | F by using APGM algorithm^Hz-X||² ₂The obtaining process of (1).

Further, the optimization of h (z) | | | z | | non-calculation using APGM algorithm as described in steps 3-7² _∞The obtaining process specifically comprises the following steps:

calculating h (z) | | | z | | non-conducting phosphor by using APGM algorithm² _∞Is close to the optimal solution of the factor u, the formula used is:

wherein gamma is a step factor;

the specific process comprises the following steps:

step 3-7-1, initializing lambda and gamma;

3-7-2, performing descending order arrangement on the z values to generate a vector s;

step 3-7-3, enabling i to be 1;

3-7-4, calculating:

in the formula, s_π(k)Is the kth element in the vector s, s_π(max)Is the largest element in the vector s;

and 3-7-5, judging the magnitude relation of i and j, and if i is not more than j, outputting the optimal solution of u as follows:

otherwise, judging whether i is less than or equal to JN, if so, increasing i by 1, and executing the steps from 3-7-4 to 3-7-5; otherwise, the obtaining process is finished.

Further, the optimization of g (z) | | | F using APGM algorithm as described in steps 3-8^Hz-X||² ₂The obtaining process specifically comprises the following steps:

using APGM algorithm to obtain g (z) | | F^Hz-X||² ₂The specific process comprises the following steps:

step 3-8-1, let j equal to 1, initialize x^j、y^j、γ^j-1、β，β∈(0,1)；

Step 3-8-2, calculating z:

step 3-8-3, based on the z value calculated in step 3-8-2, obtaining g (z):

g(z)＝||F^Hz-X||² ₂

obtaining

Step 3-8-4, judging g (z) and

the magnitude relationship of (1), if

The optimal solution γ of the step factor γ is output as β γ;

otherwise, judging whether j is larger than JN, if so, ending the solving process; otherwise, let j increment by 1 and update x^j、y^jThe formulas are respectively as follows:

y^j＝x^j-1+ω^j(x^j-x^j-1)

in the formula, ω^j∈[01) represents the jth acceleration element;

then, step 3-8-2 to step 3-8-4 are performed.

Compared with the prior art, the invention has the following remarkable advantages: 1) the method can count the function mu | z | non-woven phosphor² _∞Provides a closed solution to the approximate operation factor, while other algorithms cannot render the function μ | | | z | | calculation_∞The approach operation factor of (a) provides a closed solution; 2) the evaluation measurement of the step factor gamma can be solved by only using a simple linear search, the calculated amount is small, and different from the calculation of the step factor gamma by the FISTA algorithm, the gradient needs to be calculated

The minimum Lipschitz constant L, and the calculation amount of the constant L is large; 3) the APGM is used without cutting and filtering steps, so that the method is simple and effective, the optimal cutting threshold value is not required to be calculated, and the effect of the algorithm cannot be influenced by different cutting threshold values; 4) the APGM can not only effectively inhibit PAPR, but also obtain better BER effect.

The present invention is described in further detail below with reference to the attached drawing figures.

Drawings

FIG. 1 is a flowchart of a PAPR restraining process based on an accelerated approximation gradient algorithm TR optimization method according to the present invention.

Fig. 2 is a schematic diagram comparing the CCDF after signal processing according to the methods in an embodiment of the present invention.

Fig. 3 is a graph comparing the decrease curves of the average PAPR of the signal processed by each method with the number of iterations according to an embodiment of the present invention.

Fig. 4 is a graph showing BER curves after the signals processed by the methods pass through an additive white gaussian noise channel according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

With reference to fig. 1, the present invention provides a method for optimizing TR based on accelerated approximation gradient algorithm to suppress PAPR process, comprising the following steps:

step 1, converting a time domain signal X of an Orthogonal Frequency Division Multiplexing (OFDM) system into a frequency domain signal X by utilizing Fourier transform;

step 2, obtaining a mapping relation F between the time domain signal X and the frequency domain signal X;

step 3, generating PAPR restraining signal by using carrier wave reserving TR method

And inhibiting the PAPR by combining an APGM (approximate acceleration gradient algorithm).

Further, in one embodiment, the mapping relationship F between the time-domain signal X and the frequency-domain signal X is obtained in step 2, and the formula is as follows:

x＝FX

in the formula, F represents a mapping relation of an inverse fast fourier transform matrix with the size of JN × N, and elements in the mth row and the jth column in the matrix are:where J is the oversampling factor and N is the number of subcarriers in the OFDM system.

Further, in one embodiment, the PAPR suppressing signal is generated in step 3 by using a carrier reservation TR method

And inhibiting the peak-to-average power ratio (PAPR) by combining an accelerated approximate gradient Algorithm (APGM), specifically comprising the following steps:

step 3-1, generating PAPR inhibiting signal by using carrier wave reserved TR methodComprises the following steps:

where D is the peak reduction signal in the frequency domain: d ═ D₀,D₁,D₂,...,D_N-1]^T，D_iFor the peak reduction signal on the ith subcarrier frequency domain, the value of i is 0 to N-1;

step 3-2, PAPR based on time domain signal x, using signalThe suppression PAPR is as follows:

the PAPR of the original time domain signal x is:

wherein | × | non-conducting phosphor² _∞Representing the peak power of the signal "+", E [ + ]² ₂]Represents the average power of the signal ";

step 3-3, optimizing the inhibition process of the step 3-2:

PAPR suppression on demand acquisition

Minimum value of (d):

obtaining

Obtaining the minimum value of D:

the constraints for adding D are:

||D||² ₂≤β

wherein β represents the amplitude of the maximum allowed subcarrier signal;

therefore, the inhibition process of the optimization step 3-2 is converted into an optimization quadratic constraint quadratic programming problem, and the formula of the quadratic constraint quadratic programming problem is as follows:

m_Din||x+FD||² _∞

s.t.||D||² ₂≤β

step 3-4, converting the optimized quadratic constraint quadratic programming problem into the optimized following formula:

in the formula, lambda is a scale factor;

3-5, replacing x + FD with z, and carrying out the step 3-4

To convert to:

step 3-6, ordering:

h(z)＝λ||z||² _∞

g(z)＝||F^Hz-X||² ₂

step 3-7, optimizing h (λ | | | z | |) through APGM algorithm² _∞The process of (1);

step 3-8, optimizing g (z) | | | F by using APGM algorithm^Hz-X||² ₂The obtaining process of (1).

Further, in one embodiment, h (z) | | z | | y is optimized using APGM algorithm in the above-mentioned steps 3-7² _∞The obtaining process specifically comprises the following steps:

calculating h (z) | | | z | | non-conducting phosphor by using APGM algorithm² _∞Is close to the optimal solution of the factor u, the formula used is:

wherein gamma is a step factor;

the specific process comprises the following steps:

step 3-7-1, initializing lambda and gamma;

3-7-2, performing descending order arrangement on the z values to generate a vector s;

step 3-7-3, enabling i to be 1;

3-7-4, calculating:

in the formula, s_π(k)Is the kth element in the vector s, s_π(max)Is the largest element in the vector s;

and 3-7-5, judging the magnitude relation of i and j, and if i is not more than j, outputting the optimal solution of u as follows:

otherwise, judging whether i is less than or equal to JN, if so, increasing i by 1, and executing the steps from 3-7-4 to 3-7-5; otherwise, the obtaining process is finished.

Further, in one embodiment, the above step 3-8 optimizes g (z) | | F using APGM algorithm^Hz-X||² ₂The obtaining process specifically comprises the following steps:

using APGM algorithm to obtain g (z) | | F^Hz-X||² ₂The specific process comprises the following steps:

step 3-8-1, let j equal to 1, initialize x^j、y^j、γ^j-1、β，β∈(0,1)；

Step 3-8-2, calculating z:

step 3-8-3, based on the z value calculated in step 3-8-2, obtaining g (z):

g(z)＝||F^Hz-X||² ₂

obtaining

Step 3-8-4, judging g (z) and

the magnitude relationship of (1), ifThe optimal solution γ of the step factor γ is output as β γ;

otherwise, judging whether j is larger than JN, if so, ending the solving process; otherwise, let j increment by 1 and update x^j、y^jThe formulas are respectively as follows:

y^j＝x^j-1+ω^j(x^j-x^j-1)

in the formula, ω^jE [0,1) represents the jth acceleration element, and is generally set

Then, step 3-8-2 to step 3-8-4 are performed.

Illustratively, the same signal is processed by the method of the present invention (simulation parameters include λ 10, β 0.5, γ 0.5) and the adaptive scaling algorithm, the adaptive amplitude clipping algorithm and the fast iterative threshold shrinking algorithm, respectively, the obtained CCDF is as shown in fig. 2, the decreasing curve of the average PAPR with the number of iterations is as shown in fig. 3, the BER curve of the processed signal after passing through the additive white gaussian noise channel is as shown in fig. 4, and as can be seen from fig. 2 to 4, the method of the present invention not only can effectively suppress the PAPR, but also can obtain a better BER effect.

The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.