CN107743105B - Crest factor reduction control method for multi-frequency digital communication system - Google Patents

Crest factor reduction control method for multi-frequency digital communication system Download PDF

Info

Publication number
CN107743105B
CN107743105B CN201710832119.9A CN201710832119A CN107743105B CN 107743105 B CN107743105 B CN 107743105B CN 201710832119 A CN201710832119 A CN 201710832119A CN 107743105 B CN107743105 B CN 107743105B
Authority
CN
China
Prior art keywords
signal
frequency
value
peak clipping
frequency band
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201710832119.9A
Other languages
Chinese (zh)
Other versions
CN107743105A (en
Inventor
王杰
赵书伦
门吉卓
管恩义
姚宜东
李文
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Aerospace Control Instrument Institute
Original Assignee
Beijing Aerospace Control Instrument Institute
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Aerospace Control Instrument Institute filed Critical Beijing Aerospace Control Instrument Institute
Priority to CN201710832119.9A priority Critical patent/CN107743105B/en
Publication of CN107743105A publication Critical patent/CN107743105A/en
Application granted granted Critical
Publication of CN107743105B publication Critical patent/CN107743105B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2614—Peak power aspects
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2614—Peak power aspects
    • H04L27/2623—Reduction thereof by clipping

Abstract

A Crest Factor Reduction (CFR) control method for a multi-frequency digital communication system is characterized in that multi-dimensional dynamic crest factor reduction factors are obtained by integrating the characteristics of all input frequency band signals, and each frequency band signal is subjected to signal crest reduction processing by independently using the multi-dimensional crest factor reduction factors. Calculating a dynamic root mean square value according to the input multi-frequency digital signal, and calculating a dynamic peak clipping threshold value by combining with amplitude limiting rate configuration; according to the dynamic threshold value and the multi-frequency signal modulus sum, performing peak value judgment and calculation to obtain a multi-dimensional hard peak clipping factor; the multi-dimensional hard peak clipping factor is subjected to a window function and an FIR filter to obtain a multi-dimensional smooth peak clipping factor; and multiplying the digital signals of each frequency band by the multi-dimensional smooth peak clipping factor to obtain signals after peak clipping, and obtaining multi-frequency combined signals after peak clipping through combining. The invention is used for solving the control requirement of reducing the crest factor of the multi-band branch communication system, can realize the peak clipping treatment of the multi-band signal, avoids the phenomenon of over-under peak clipping, and can solve the problems of linearity and efficiency of the radio frequency amplifier of the multi-band digital communication system to a certain extent.

Description

Crest factor reduction control method for multi-frequency digital communication system
Technical Field
The invention relates to a crest factor reduction control method for a multi-frequency digital communication system, belonging to the technical field of digital communication.
Background
With the continuous development of mobile communication technology, new standards are proposed and applied continuously, and multiple communication standards and communication systems coexist for a long time, so that an integrated communication system supported by multiple communication standards with the characteristics of fusion and flexibility becomes a research trend and a hotspot. The integrated communication system supported by multiple standards and multiple systems can realize shared co-construction of communication facilities and networks, shorten construction period, overall stage network facilities, and reduce Capital expenditure (CAPEX) and operation expenditure (OPEX) of the system to a certain extent. In a communication system supported by multiple service systems, the requirement of high data transmission rate needs to be met, and meanwhile, the requirements of stricter linearity and flexibility of radio frequency transmitters of various communication standards need to be met. The current 3G and 4G communication standards implement high-speed data transmission through Multiple carriers by using a complex modulation scheme, such as Wideband Code-division Multiple Access (WCDMA) and Long-Term Evolution (LTE) technologies, however, WCDMA and LTE signals are non-constant envelopes and have a large Peak-to-Average Power Ratio (PAPR), which aggravates the contradiction between the efficiency and linearity of the radio frequency PA. Research has shown that the efficiency of class AB amplifiers is around 5-10%, but reasonable linearity processing methods can increase the efficiency to around 60%. Mishandling of PA efficiency and linearity can be detrimental to the performance and cost of a communication system. In an integrated communication system supported by multiple service systems, the efficiency and linearity of the rf PA are more complex. Crest Factor Reduction (CFR) is a technique that effectively reduces the signal high PAPR, thereby reducing the backoff of the PA and improving the PA efficiency.
The current CFR control method mainly aims at single-frequency-band signals (1D-CFR for short), the direct effect of the 1D-CFR technology in a dual-band or multi-band transmitter is poor, and because the signal characteristics of all input frequency bands are not comprehensively considered, the problem that the single-frequency signals subjected to peak clipping by the 1D-CFR are easy to have insufficient peak clipping or excessive peak clipping after being combined is solved, and the problems of linearity and efficiency of a radio frequency amplifier of a multi-service frequency band supporting communication system can not be effectively solved. Currently, in the existing CFR control methods for multi-band signals or broadband signals, some control methods adopt a single-path peak clipping and re-combining method, peak clipping processing between signal frequency bands is independent, and comprehensive characteristics of signals of the frequency bands are not considered, so that the problem of over peak clipping or insufficient peak clipping exists in the combined signals; although the comprehensive characteristics of signals in each frequency band are considered, some methods usually estimate a given peak clipping threshold value through experience, and the problem of excessive peak clipping or insufficient peak clipping still exists without combining the real-time dynamic characteristics of input signals, so that performance parameters such as Vector Error (EVM) and Adjacent Channel Leakage Ratio (ACLR) of a communication system are deteriorated.
Disclosure of Invention
The technical problem solved by the invention is as follows: aiming at the problems of excessive peak clipping and insufficient peak clipping existing in the 1D-CFR and the existing multi-frequency CFR, the method for reducing the crest factor of the multi-frequency digital communication system is realized, the method adopts a dynamic peak clipping threshold value and a multi-dimensional peak clipping threshold value strategy, the problems of insufficient peak clipping or excessive peak clipping existing in the reduction control of the crest factor of the communication system under the support of multi-frequency band service are solved, and the realization process and the principle of the invention are shown in the attached figures 1 and 2.
The technical scheme of the invention is as follows: a crest factor reduction method for a multi-frequency digital communication system comprises the following steps:
(1) based on dynamic RMS value of multi-frequency digital signal inputted into multi-frequency digital communication system in a period of timeMNCalculating a dynamic peak clipping threshold value A according to the amplitude limiting rate CR configured according to the requirement of the multi-frequency digital communication system; the period of time is the duration of M signal sampling points, and N represents the number of input signal frequency segments;
(2) calculating power level modulus and P of multi-frequency digital signal inputted into multi-frequency digital communication system at mth sampling point momentsum(m) finding the maximum modulus x of the power levelmax(m) and a very small modulus value xmin(m) and calculating the maximum modulus and the absolute difference x of the minimum modulusd(m)；
(3) According to the dynamic peak clipping threshold value A obtained in the step (1) and the power P obtained in the step (2)sum(m) comparing peak values and obtaining a multi-dimensional hard peak clipping factor c (m, n) of each frequency band of the multi-frequency signal under the multi-frequency band condition, and using n to represent the nth frequency band signal in the input multi-frequency signal;
(4) the multi-dimensional hard peak clipping factors c (m, n) obtained in the step (3) pass through a window function and an FIR filter to obtain smooth multi-dimensional smooth peak clipping factors b (m, n);
(5) multiplying each frequency band digital signal x (m, n) in the input multi-frequency signal by the multi-dimensional smooth peak clipping factor b (m, n) in the step (4) to obtain a smooth amplitude limiting signal, and combining all frequency band signals to obtain a multi-frequency amplitude limiting signal youiThereby realizing the control of the reduction of the crest factor of the multi-band signal.
Compared with the prior art, the invention has the advantages that:
(1) dynamic root mean square value RMS in the present inventionMNObtained by comprehensively calculating power level values of signals in all frequency bandsThe characteristics of the comprehensive signal can be more accurately reflected;
(2) the invention passes the dynamic root mean square value RMSMNCalculating a dynamic peak detection threshold A according to the amplitude limiting rate CR, and calculating peak clipping thresholds according to the real-time characteristics of the signal of the section by different signal sections, so that the peak clipping can be realized more accurately;
(3) the invention calculates the dynamic peak clipping threshold value according to the dynamic root mean square value of all frequency band signals in M sampling points, the peak clipping threshold value of each sampling point moment is determined by the comprehensive characteristics of each frequency band signal at the moment, thereby ensuring that each frequency band signal can avoid the phenomena of under clipping or over clipping of the combined signal while obtaining the proper peak clipping treatment
(4) The invention designs the frequency band power weight coefficient lambdanThe corresponding lambda can be configured according to the importance of the frequency band signalnTo implement weight configuration of signal power;
(5) the multi-dimensional peak clipping factor design scheme based on the dynamic peak clipping threshold and the comprehensive characteristics of the multi-frequency signal enables the peak factor reduction control method designed by the invention to achieve reasonable peak clipping of the multi-frequency signal and avoid the phenomena of excessive peak clipping and insufficient peak clipping of the combined signal.
Drawings
FIG. 1 is a flow chart of a method of the present invention;
FIG. 2 is a functional block diagram of an implementation of the present invention;
FIG. 3 is a schematic block diagram of simulation based on 3 different LTE signals according to the present invention;
fig. 4 is a diagram of simulation effect of the present invention on mixed downlink signals with LTE FDD TM2 and TM3 as input signals: wherein, FIG. 4(a) is the simulation effect diagram after the simulation signal passes through the 1D-CFR and the multi-frequency CFR designed by the present invention; FIG. 4(b) is a Distribution diagram of probability Distribution Function (CDF) after 1D-CFR and multi-frequency CFR designed by the present invention; fig. 4(c) is CDF distribution diagram of LTE FDD TM2 and TM3 hybrid downlink simulation signals under different C R.
FIG. 5 is a diagram illustrating simulation effect of the present invention on mixed downlink signals with input signals of LTE TDD 2 and TM 3: wherein, FIG. 5(a) is the simulation effect diagram after the simulation signal passes through the 1D-CFR and the multi-frequency CFR designed by the present invention; FIG. 5(b) is a CDF distribution graph of a simulation signal after being subjected to 1D-CFR and multi-frequency CFR designed by the present invention; fig. 5(c) is CDF distribution diagram of LTE TDD 2 and TM3 hybrid downlink simulation signals under different CRs.
Detailed Description
The basic idea of the invention is as follows: a crest factor reduction control method for a multi-frequency digital communication system is characterized in that the characteristics of signals of all input frequency bands are integrated to obtain multi-dimensional dynamic crest factor in a digital middle frequency domain, and the multi-dimensional crest factor is independently used for realizing signal crest reduction processing in each frequency band, and meanwhile, the phenomena of over crest reduction and under crest reduction of the signals after combination are avoided. Calculating a dynamic root mean square value according to the input multi-frequency digital signal, and calculating a dynamic peak clipping threshold value by combining an amplitude limiting rate configuration value; calculating the power level modulo sum of the multi-frequency signal according to the input multi-frequency digital signal; searching a maximum module value and a minimum module value of the power level of the multi-frequency signal according to the input multi-frequency digital signal; and detecting the peak value of the multi-frequency signal, and judging the peak value according to a dynamic threshold value and a multi-frequency signal modulus sum, so as to obtain a multi-dimensional hard peak clipping factor: if the peak value is less than or equal to the threshold value, directly obtaining a peak clipping factor under the condition, if the peak value is greater than the threshold value, determining and calculating the peak clipping factor according to the absolute difference value of the maximum value and the minimum value and whether the signal is an extreme value, and obtaining a multi-dimensional hard peak clipping factor according to the peak clipping factor under each condition; under the condition of obtaining the hard peak clipping factor, combining a window function and an FIR filter to obtain a smoothed multidimensional smooth peak clipping factor; multiplying the digital signals of each frequency band by a smooth peak clipping factor to obtain peak clipped signals; and finally, carrying out signal combination to obtain a multi-frequency combined signal after peak clipping. The invention is used for solving the control requirement of reducing the crest factor of a multi-frequency band service supporting communication system, adopts a dynamic peak clipping threshold value and a multi-dimensional signal peak value judgment method, can simultaneously realize the peak clipping processing of multi-frequency band signals, avoids the phenomena of over peak clipping and under peak clipping, and can solve the problems of linearity and efficiency of a radio frequency amplifier of the multi-frequency digital communication system to a certain extent.
The invention is described in further detail below with reference to the figures and specific embodiments.
As shown in fig. 1 and fig. 2, the present invention realizes a crest factor reduction control method for a multi-frequency digital communication system, comprising the steps of:
(1) dynamic RMS value over a period of time based on input multi-frequency digital signalsMNAnd calculating a dynamic peak clipping threshold value A by the limiting rate CR required by the system, specifically: the power level sum of all frequency band signals at the mth sampling point time is taken as a comprehensive sampling point, the root mean square value of the M comprehensive sampling points is calculated as a calculation parameter of the dynamic peak clipping threshold, and the characteristics of all signals can be effectively introduced into the control method by adopting the method, so that the effective control of the multi-frequency signals is realized. The specific calculation expression is as follows,
Figure BDA0001408921420000051
Figure BDA0001408921420000052
wherein the content of the first and second substances,
m: participating in calculating the number of sampling points of the dynamic root mean square value; n: the frequency segment number of the input multi-frequency signal;
λn: the power weight coefficient of the nth service frequency band signal; CR: the amplitude limiting rate;
xm(n): the signal power level value corresponding to the nth signal frequency band at the mth sampling point time.
The values of the above-mentioned related parameters are explained as follows:
m is the number of signal samples participating in dynamic root mean square value calculation, and the value-taking principle is as follows: under the condition that the value of N is 3, different M values are taken to draw the RMS of the combined signal of the 3-frequency band signalMNAnd comparing the signal with the original 3-band combined signal to obtain the most appropriate value. When M takes the value of 50, RMSMNThe characteristics of the original combined signal cannot be accurately reflected, and when the value of M is 200, RMSMNThe reflected original combined signal characteristics are basically the same as those when the value of M is 100, but a larger system time delay is introduced, so that the value of M is set to be 100. Value of M is[95,120]The effect is basically the same between.
N is the number of service frequency bands input by the system and is determined by the number of the actually input service frequency bands; lambda [ alpha ]nAnd configuring a power value weight coefficient for the nth service frequency band according to the specific condition of the input frequency band.
CR is an amplitude limiting rate and can be configured according to the actual requirements of the system, and theoretically, the smaller CR is, the better the peak clipping effect is; experimental simulation shows that the effect of the multi-frequency CFR is better when the CR value is between [0.5 and 1.2], the peak clipping effect of the multi-frequency CFR is not as good as that of 1D-CFR when the CR value is more than 1.5, and the attached figures 4(c) and 5(c) are CDF test simulation distribution diagrams of the multi-frequency CFR under different CRs.
The dynamic peak clipping threshold value is calculated according to the dynamic root mean square values of all the frequency band signals in M sampling points, the peak clipping threshold value of each sampling point moment is determined by the comprehensive characteristics of each frequency band signal at the moment, and the phenomenon of under clipping or over clipping of the combined signal can be avoided while the signals of each frequency band are subjected to proper peak clipping treatment.
(2) Calculating power level modulus and P of input multi-frequency digital signal at m-th sampling point momentsum(m), maximum modulus x of power levelmax(m) (maximum modulus is maximum modulus), minimum modulus xmin(m) (minimum modulus is minimum modulus) and the difference x between the maximum and minimum valuesd(m) of the reaction mixture. The method specifically comprises the following steps:
Figure BDA0001408921420000061
Figure BDA0001408921420000062
wherein the content of the first and second substances,
Psum(m): the power level modulo sum of all frequency band signals at the mth sampling point moment;
xmax(m): the maximum module value of the power level in all the frequency band signals at the mth sampling point moment;
xmin(m): all frequencies at the m-th sampling point momentA power level minimum modulus value in the segment signal;
xd(m): and the absolute difference value of the maximum power level module value and the minimum power level module value at the mth sampling point moment.
λ: matrix of power weighting coefficients, λ ═ λ1,λ2,....λN]T；
X: the power level mode matrix of all frequency bands at the sampling point m moment, X [ | X [ ]m(1)|,|xm(2)|,....|xm(N)|]。
The multi-frequency CFR designed by the invention compares the power level modulus of all frequency band signals at the sampling point m moment with the dynamic threshold value during peak value detection and comparison, and comprehensively considers the characteristics of each frequency band signal, so that the peak clipping processing is more reasonable and accurate.
(3) According to the dynamic peak clipping threshold value A obtained in the step (1) and the P obtained in the step (2)sum(m) comparing peak values and obtaining a multi-dimensional hard peak clipping factor (namely the hardware peak clipping factor determined by the multi-band signals together) of each frequency band signal under the comprehensive condition, and representing the nth frequency band signal in the multi-frequency input signal by n. The method specifically comprises the following steps:
if the power level modulo sum P is at the sampling point msum(m)≤AkThen, for all the peak-clipped signals in the service frequency band, the module value of the signal power level is the same as that of the original signal, expressed by the following formula,
Figure BDA0001408921420000071
namely, it is
Figure BDA0001408921420000073
If P issum(m)>AkThen, the power value of each service band is determined by designing a new judgment criterion based on the absolute difference of the extreme values. The new criteria are as follows:
if xd(m)≤AkThe difference value of the signal power corresponding to the maximum power module value and the minimum power module value at the sampling point m is shown in the dynamic threshold rangeIn this case, the peak clipping criteria of each service band is given by the following formula,
Figure BDA0001408921420000072
namely, it is
Figure BDA0001408921420000081
If xd(m)>AkDescription of xmin(m) corresponding to a weak service signal power, xmax(m) the corresponding service signal has strong power, and if peak clipping is performed according to a uniform standard, all frequency band signals cannot be subjected to proper peak clipping treatment, so that the phenomenon of insufficient peak clipping or excessive peak clipping is easy to occur. In this case n is definedmax、nminRespectively corresponding the serial numbers of the service frequency bands to the maximum power level modulus value and the minimum power level modulus value, and reasonably processing frequency band signals corresponding to the maximum power level modulus value and the minimum power level modulus value: the power value after the peak clipping of the signal frequency band corresponding to the minimum power level modulus is the same as that of the original signal, and the power value after the peak clipping of the signal frequency band corresponding to the maximum power level modulus is xd(m)<AkUnder the circumstances, the value is subtracted
Figure BDA0001408921420000082
The specific peak clipping criterion and expression are as follows:
Figure BDA0001408921420000083
based on the design, the peak clipping factor c of the kth section signal is obtainedk(m, n) is represented by the following formula,
Figure BDA0001408921420000084
wherein the content of the first and second substances,
Ak: the peak clipping threshold value of the signal segment of the k segment consisting of m sampling points is the same as the value of A, and the variable is proposed for facilitating each segment of the signalDescription of number processing;
nmax: the maximum power level module value at the mth sampling point corresponds to the serial number of the service frequency band;
nmin: the m-th sampling point moment minimum power level module value corresponds to the serial number of the service frequency band;
xd(m) is the absolute difference value of the maximum power level modulus signal and the minimum power level modulus signal at the mth sampling point moment, and is taken as P in the inventionsum(m)>AkNew criterion for peak clipping under the circumstances, and xdAnd (m) determining peak clipping processing of each frequency band signal according to the magnitude relation with A, wherein the comparison object is a dynamic threshold A. In practical applications, other contrast objects and contrast gradients can be set according to the specific needs of the system, such as
Figure BDA0001408921420000091
And the like.
(4) And (4) obtaining the multi-dimensional smooth peak clipping factor b (m, n) after smoothing by the multi-dimensional hard peak clipping factor c (m, n) obtained in the step (3) through a window function and an FIR filter, wherein the window function selected in the invention is a Hamming window, and the window length is set to 2000.
(5) Multiplying the input frequency band signals x (m, n) by the smooth peak clipping factor b (m, n) to obtain smooth amplitude limiting signals, and combining the signals to obtain a multi-frequency amplitude limiting signal youi. The method specifically comprises the following steps:
Figure BDA0001408921420000092
wherein the content of the first and second substances,
yout(m, n): and the power value of the nth service frequency band subjected to multi-frequency CFR amplitude limitation at the mth sampling point moment.
The multi-frequency digital communication system crest factor reduction control method designed by the invention takes subsequent practical system application into consideration during design, a simulation model can be built in a Simulink environment based on the model described by the invention, an HDL (high density description) coding module provided by Simulink can be used for converting the control method into Verilog HDL (hardware description language) codes, and the Verilog HDL codes are burnt after simple modificationAnd (5) verifying and using in the FPGA. Calculating the number M of sample points, the limiting rate CR and the secondary peak clipping judgment comparison object x in the dynamic root mean square valued(m), the number of input signal frequency bands N and the frequency band power value weight coefficient lambdanThe multi-frequency CFR control model designed by the invention is built and simulated in a Simulink environment under the configurations of 100, 1.2, A, 3 and 1 respectively. A high-intermediate-frequency scheme is adopted for simulation, a signal generator (SMW200A) of Rode and Swatt generates a simulated intermediate-frequency 4G LTE signal of a corresponding frequency band, the intermediate-frequency band is 115.2MHz, the signal bandwidth is 20MHz, a high-performance FPGA (EP4SGS530) of ALTERA controls a 14-bit ADC to acquire three frequency band signals for simulation analysis of a building model in Simulink, and 32768 sampling points are acquired under each power. The random combination of the three frequency band signals under different powers is used as the input of the control model for simulation, the simulation principle block diagram is shown in figure 3, and the simulation result is shown in figures 4 and 5.
Fig. 4 is a simulation result diagram of 3 signals of which input signals are composed of LTE FDD TM2 and TM3, and the 3 signal characteristics used for simulation are as follows:
signal 1 (S1-1): signal template FDD TM2 Signal Strength 0dBm
Signal 2 (S1-2): signal template FDD TM3 Signal Strength-20 dBm
Signal 3 (S1-3): signal template FDD TM2 Signal Strength-10 dBm
Wherein, fig. 4(a) is a diagram of the effect of combining 3 input signals after being processed by 1D-CFR and multi-frequency CFR designed by the present invention, in which the abscissa is the sampling time, the range is [0,32768], and the ordinate is the power level value. The first action is 1D-CFR combined signal, the range of the ordinate power level value is [ -5000,5000], the second action is multi-frequency CFR combined signal, the range of the ordinate power level value is [ -4000,4000], the multi-frequency CFR peak clipping effect designed by the invention is obviously better than 1D-CFR from the simulation effect diagram, and the peak value of the combined signal is obviously lower than that of the 1D-CFR combined signal. Fig. 4(b) is a probability distribution function (CDF) distribution diagram of a combined signal after a simulated signal is subjected to 1D-CFR and multi-frequency CFR designed by the present invention, wherein the ordinate is probability, the range is [0,1], and the abscissa is power level value, the range is [ -10000,10000 ]. In the simulation result graph, a red zone curve is the CDF distribution of the original signal combined signals, a blue implementation curve is the CDF distribution of the combined signals after passing through 1D-CFR, a green zone curve is the CDF distribution of the multi-frequency CFR combined signals designed by the invention, and the simulation effect graph proves that the large peak probability of the signals after peak clipping by the multi-frequency CFR is obviously reduced from the peak probability distribution condition of the signals after amplitude clipping. Fig. 4(c) is a CDF distribution diagram of the simulation signal under the conditions that CR values are 0.5,1, 1.5, and 2, the ordinate range [0,1], and the abscissa is the power level value, the simulation result shows that different amplitude limiting rate CR values have different peak clipping effects, and the smaller the CR value is, the better the peak clipping effect is; when the CR value is 1.5, the peak clipping effect of the multi-frequency CFR is lower than that of 1D-CFR; the optimal CR value range of the multi-frequency CFR designed by the invention is [0.5,1.2 ]. Fig. 4 shows 3 simulation effect diagrams illustrating the correctness of the multi-frequency CFR control method designed by the present invention from 3 different planes, and the multi-frequency CFR control method has a better control effect
To further verify the design of the present invention, a similar simulation test experiment was performed on another set of simulation signals. FIG. 5 is a graph of the simulation effect of 3 signals with input signals consisting of LTE TDD 2 and TM3, the characteristics of the 3 simulated input signals are as follows:
signal 1 (S2-1): signal template TDD TM3 Signal Strength-25 dBm
Signal 2 (S2-2): signal template TDD TM2 Signal Strength-5 dBm
Signal 3 (S2-3): signal template TDD TM3 Signal Strength-10 dBm
Wherein, fig. 5(a) is a multi-frequency CFR combined signal effect diagram after 1D-CFR and multi-frequency CFR designed by the invention are performed on 3 kinds of input signals, the first action is a 1D-CFR combined signal, the range of ordinate [ -5000,5000], the second action is a multi-frequency CFR combined signal, the range of ordinate [ -4000,4000], the simulation effect diagram also proves that the multi-frequency CFR peak clipping effect is better than that of 1D-CFR, and the peak value is basically within 2000 after multi-frequency CFR. Fig. 5(b) is a CDF distribution diagram of a combined signal after a simulation signal passes through a 1D-CFR and a multi-frequency CFR designed by the present invention, the corresponding relationship between the vertical coordinate range and the color of the CDF distribution diagram and the signal is the same as that in fig. 4(b), and the simulation effect diagram shows that the probability of a large peak value of the signal after the peak clipping by the multi-frequency CFR is obviously reduced, but the difference between the peak clipping effects of the 1D-CFR and the multi-frequency CFR is smaller than that of the previous group of simulation signals, because the simulation signals have different signal characteristics. Fig. 5(c) is a CDF distribution diagram of the simulation signal under different CR values, where the vertical coordinate range and the CR value are the same as those in fig. 4(c), and the simulation result also shows that the peak clipping effect is different under different CR values, and the smaller the CR value is, the better the peak clipping effect is; when the value of CR is 1.5, the peak clipping effect of the multi-frequency CFR is also lower than that of 1D-CFR, but the difference is smaller compared with that of the figure 4 (c).
Two sets of simulation signals obtain consistent simulation results: for simulation signals of different combinations, the effect of the multi-frequency CFR control method designed by the invention is obviously better than that of 1D-CFR, and the adopted design scheme of dynamic root-mean-square and dynamic peak clipping threshold value achieves good effect; different CR values have different peak clipping effects, theoretically, the smaller the CR value is, the better the peak clipping effect is, but the CR has an optimal value range, and when the CR value is 1.5, the peak clipping effect of the multi-frequency CFR is basically lower than that of 1D-CFR.

Claims (6)

1. A crest factor reduction control method for a multi-frequency digital communication system, comprising the steps of:
(1) based on dynamic RMS value of multi-frequency digital signal inputted into multi-frequency digital communication system in a period of timeMNCalculating a dynamic peak clipping threshold value A according to the amplitude limiting rate CR configured according to the requirement of the multi-frequency digital communication system; the period of time is the duration of M signal sampling points, and N represents the number of input signal frequency segments;
the specific method for calculating the dynamic peak clipping threshold value A in the step (1) comprises the following steps:
Figure FDA0002669779960000011
Figure FDA0002669779960000012
wherein the content of the first and second substances,
m: participating in calculating the number of sampling points of the dynamic root mean square value; n: the frequency segment number of the input multi-frequency signal;
λn: the power weight coefficient of the nth service frequency band signal; c R: the amplitude limiting rate;
xm(n): the signal power level value corresponding to the nth signal frequency band at the mth sampling point moment;
(2) calculating power level modulus and P of multi-frequency digital signal inputted into multi-frequency digital communication system at mth sampling point momentsum(m) finding the maximum modulus x of the power levelmax(m) and a very small modulus value xmin(m) and calculating the maximum modulus and the absolute difference x of the minimum modulusd(m)；
(3) According to the dynamic peak clipping threshold value A obtained in the step (1) and the power P obtained in the step (2)sum(m) comparing peak values and obtaining a multi-dimensional hard peak clipping factor c (m, n) of each frequency band of the multi-frequency signal under the multi-frequency band condition, and using n to represent the nth frequency band signal in the input multi-frequency signal;
(4) the multi-dimensional hard peak clipping factors c (m, n) obtained in the step (3) pass through a window function and an FIR filter to obtain smooth multi-dimensional smooth peak clipping factors b (m, n);
(5) multiplying each frequency band digital signal x (m, n) in the input multi-frequency signal by the multi-dimensional smooth peak clipping factor b (m, n) in the step (4) to obtain a smooth amplitude limiting signal, and combining all frequency band signals to obtain a multi-frequency amplitude limiting signal youiThereby realizing the control of the reduction of the crest factor of the multi-band signal.
2. The crest factor reduction control method for a multi-frequency digital communication system according to claim 1, wherein: step (2) calculating the power level modulus and P of the input multi-frequency digital signal at the mth sampling point momentsum(m), maximum modulus x of power levelmax(m) very small modulus value xmin(m), maximum and minimum difference xd(m), the specific method is as follows:
Psum(m) is calculated according to the following formula,
Figure FDA0002669779960000021
the matrix lambda is a power weight coefficient matrix, lambda is [ lambda ]1,λ2,....λN]T，λnCorresponding to the power weight coefficient of the nth frequency band signal, and
Figure FDA0002669779960000022
the matrix X is a matrix formed by corresponding power level module values of all frequency band signals at the sampling point m moment, and X [ | X [ ]m(1)|,|xm(2)|,....|xm(N)|]，|xm(N) | is a power level module value of the nth service frequency band at the sampling point m moment, and N is 1,2, … and N;
maximum modulus x of power levelmax(m) very small modulus value xmin(m), absolute difference x of maximum and minimum modulusd(m) is calculated according to the following formula,
Figure FDA0002669779960000023
wherein the content of the first and second substances,
Psum(m): the power level modulo sum of all frequency band signals at the mth sampling point moment;
xmax(m): the maximum module value of the power level in all the frequency band signals at the mth sampling point moment;
xmin(m): the minimum module value of power level in all frequency band signals at the mth sampling point moment;
xd(m): and the absolute difference value of the maximum power level module value and the minimum power level module value at the mth sampling point moment.
3. The crest factor reduction control method for a multi-frequency digital communication system according to claim 1, wherein: the step (3) of calculating the multi-dimensional hard peak clipping factor c (m, n) specifically comprises the following steps:
if the power level modulo sum P of all frequency band signals at the sampling point m timesum(m)≤AkThen, for all the service frequency bands, there is,
Figure FDA0002669779960000031
i.e. | xm(n)_CFR|＝|xm(n)|,
Figure FDA0002669779960000032
If P issum(m)>AkThen, based on the absolute difference of the extreme values, a new judgment criterion is designed to determine the peak clipping value of each service frequency band, and the judgment criterion is as follows: if xd(m)≤AkThe peak clipping rule of each service frequency band is given by the following formula,
Figure FDA0002669779960000033
in the formula, xm(1)_CFR: the power level of the 1 st frequency band signal after peak clipping processing at the mth sampling point moment;
xm(2)_CFR: the power level of the 2 nd frequency band signal after peak clipping processing at the mth sampling point moment;
xm(N)_CFR: the power level of the Nth frequency band signal at the mth sampling point moment after peak clipping processing;
namely, it is
Figure FDA0002669779960000034
If xd(m)>AkThen, the peak clipping is carried out according to the following criteria,
Figure FDA0002669779960000041
obtaining a multi-dimensional hard peak clipping factor c of the kth segment signal according to the processingk(m, n) is represented by the following formula,
Figure FDA0002669779960000042
wherein the content of the first and second substances,
Ak: the kth section is a signal section peak clipping threshold value consisting of m sampling points;
nmax: the maximum power level module value at the mth sampling point corresponds to the serial number of the service frequency band;
nmin: and the minimum power level module value at the mth sampling point corresponds to the serial number of the service frequency band.
4. The crest factor reduction control method for a multi-frequency digital communication system according to claim 1, wherein: c calculated in step (4)k(m, n) obtaining a smooth multi-dimensional peak clipping factor b by a window function and a FIR filterk(m, n), each input band signal and bkMultiplying (m, n) to obtain an amplitude-limited signal, and combining to obtain a multi-band combined signal, which specifically comprises:
Figure FDA0002669779960000051
wherein, yin(m,1) inputting a signal of the 1 st frequency band at the mth sampling point moment;
yin(m,2) inputting a signal of the 2 nd frequency band at the mth sampling point moment;
yin(m, N) inputting a signal of an Nth frequency band signal at the mth sampling point moment;
youtand (m, n) the power value of the nth frequency band signal after the multi-frequency CFR amplitude limitation at the mth sampling point moment.
5. The crest factor reduction control method for a multi-frequency digital communication system according to claim 1, wherein: the sampling point value M participating in the dynamic root mean square value calculation is [95,120 ]](ii) a The amplitude limiting rate CR has different peak clipping effects when being different in value, and the value of CR is [0.5,1.2]](ii) a Weight coefficient lambda of each frequency band signalnValue is according to systemIs set to the actual requirements.
6. The crest factor reduction control method for a multi-frequency digital communication system according to claim 3, wherein: at Psum(m)>AkUnder the condition, the power value of each service frequency band is determined based on the size relation between the absolute difference value of the extreme value and the dynamic threshold, the comparison object of the absolute difference value of the extreme value is the cutting threshold, and a proper comparison object and a comparison gradient can be set according to the actual condition of the system.
CN201710832119.9A 2017-09-15 2017-09-15 Crest factor reduction control method for multi-frequency digital communication system Active CN107743105B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710832119.9A CN107743105B (en) 2017-09-15 2017-09-15 Crest factor reduction control method for multi-frequency digital communication system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710832119.9A CN107743105B (en) 2017-09-15 2017-09-15 Crest factor reduction control method for multi-frequency digital communication system

Publications (2)

Publication Number Publication Date
CN107743105A CN107743105A (en) 2018-02-27
CN107743105B true CN107743105B (en) 2020-11-20

Family

ID=61235993

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710832119.9A Active CN107743105B (en) 2017-09-15 2017-09-15 Crest factor reduction control method for multi-frequency digital communication system

Country Status (1)

Country Link
CN (1) CN107743105B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20210012235A (en) 2019-07-24 2021-02-03 삼성전자주식회사 Electronic device and wireless communication system thereof

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101714784B1 (en) * 2010-01-25 2017-03-10 삼성전자주식회사 Method and apparatus for reducing peak to average power ratio by using peak windowing
CN102223338B (en) * 2010-04-15 2013-10-16 电信科学技术研究院 Self-adaption peak clipping method and device of multi-carrier system
CN103428139B (en) * 2012-05-15 2016-08-17 京信通信系统（中国）有限公司 A kind of method and device suppressing peak-to-average force ratio
CN102957659B (en) * 2012-11-14 2014-12-31 西南石油大学 Self-adaptive amplitude limiting method for reducing peak-to-average power ratio of OFDM (orthogonal frequency division multiplexing) system
US10044543B2 (en) * 2015-07-30 2018-08-07 Futurewei Technologies, Inc. Reducing crest factors
CN105357749B (en) * 2015-10-16 2018-08-31 北京航天万达高科技有限公司 A kind of new type auto gain control method for digital communication

Also Published As

Publication number Publication date
CN107743105A (en) 2018-02-27

Similar Documents

Publication Publication Date Title
CN105656830B (en) Ofdm signal method for inhibiting peak-to-average ratio based on distributed implementation
US20090285194A1 (en) Efficient Peak Cancellation Method for Reducing the Peak-To-Average Power Ratio in Wideband Communication Systems
US9160378B2 (en) Control method and equipment for radio-frequency signal
CN101150357A (en) Method for eliminating peak power
US20080285673A1 (en) Apparatus and method for reducing peak to average power ratio based on tile structure in broadband wireless communication system
CN102970266A (en) Multi-band signal peak clipping method and device
CN105763495A (en) Digital pre-distortion method and device
CN103812817A (en) Peak-to-average power ratio inhibition method for orthogonal frequency division multiplexing (OFDM) signal
CN107743105B (en) Crest factor reduction control method for multi-frequency digital communication system
JP7079566B2 (en) Signal processing circuit
Nader et al. Peak-to-average power reduction of OFDM signals by convex optimization: Experimental validation and performance optimization
US9042434B2 (en) Filter
US20120057650A1 (en) Transmitting a signal from a power amplifier
US7236533B2 (en) Method and apparatus for reducing ratio of peak power to average power of multi-carrier signals
EP2839619B1 (en) Universal peak power reduction in communication radios
Zhao et al. A novel two-dimensional crest factor reduction for performance improvement of RF power amplifiers
CN104994051B (en) Based on instantaneously the signal transmitting with the linear-scale conversion of the raising mixed carrier system PA efficiency of average power ratio and method of reseptance
Clark et al. Fixed-point modeling in an ultra wideband (UWB) wireless communication system
US11139781B2 (en) Power amplifiers
CN104980385B (en) A kind of signal despicking method and equipment
CN109714288B (en) Amplitude limiting filtering method for reducing cubic metric of OFDM (orthogonal frequency division multiplexing) signal
CN102833012A (en) Downlink and uplink digital intermediate-frequency self-check method in multi-carrier transmitting system
Mohammady et al. Comparison of Piece-Wise and full CFR for OFDM (LTE, 5G beyond), WCDMA, and DVB-S2X signals
CN111262808B (en) Peak clipping method based on distribution function in wireless base station
Liu et al. Impact of carrier frequency dependent power amplifier behavior on 802.11 a WLAN system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant