CN105553909A - Iterative clipping and filtering method for reducing peak-to-average ratio of OFDM signals - Google Patents

Abstract

The invention provides an iterative clipping and filtering method for reducing the peak-to-average ratio of an OFDM signals. The method comprises the steps of: firstly, calculating OFDM time domain signal power, and carrying out clipping on the real parts and imaginary parts of the OFDMs exceeding a power threshold P0 according to a clipping factor [alpha]; secondly, converting the OFDM time domain signals after the clipping into OFDM frequency domain signals by means of FFT conversion, and carrying out frequency domain filtering; then converting the signals after the frequency domain filtering into OFDM time domain signals by means of IFFT conversion, and finishing one time of iterative clipping and filtering processing; and finally, according to a maximum iterative number, repeating the above iterative clipping and filtering process. In the method, an OFDM system does not need to transmit additional information, additional processing on the OFDM signals on a receiving end for compensating the influences of the clipping processing on a sending end is not needed, the clipping factor [alpha] is suitably selected, the clipping processing is completed only by means of additive operation, and frequency domain filtering can be directly carried out on the processing result by means of the FFT conversion; in addition, in the iterative clipping and filtering process, a work clock multiplexing FFT/IFFT conversion unit higher than a sampling rate is adopted, and realization resources inhibited by the peak-to-average ratio are saved.

Description

A kind of iteration limit filtration reduces the method for ofdm signal peak-to-average force ratio

Technical field

The invention belongs to digital wireless communication transmission technique field, relate to a kind of method that iteration limit filtration reduces ofdm signal peak-to-average force ratio.

Background technology

OFDM (OrthogonalFrequencyDivisionMultiplexing, OFDM) technology is a kind of multi-carrier modulation technology, by means of its advantage such as higher transmission rate, stronger ability of anti-multipath, being widely used in wireless communication field, is one of core technology of forth generation land mobile communication.

The general principle of OFDM technology is that the input traffic of high speed serialization is modulated to N _son the subcarrier of the low-speed parallel that road is mutually orthogonal, the information rate of every subcarriers data flow is made to be reduced to the 1/N of input traffic _sdoubly, data symbol duration expands to the N of input data _sdoubly.The modulating/demodulating process of OFDM adopts inverse Fourier transform/Fourier transform (InverseFastFourierTransform, IFFT/FFT) to realize usually, therefore baseband OFDM symbols is expressed as

$x\left(n\right)=\frac{1}{\sqrt{N_s}}\underset{k=0}{\overset{N_s-1}{\Σ}}X\left(k\right)e^{j2\mathrm{\π}kn/N_s},n=0,...,N_s-1$

Wherein, X (k) is the data on a kth subcarrier, N _sfor the size of IFFT.

As can be seen from the expression formula of OFDM baseband signal, ofdm modulation signal is by N _sindividual independent and orthogonal subcarrier is formed by stacking.Work as N _slarger peak value may be produced during individual subcarrier in-phase stacking, cause the peak-to-average power (Peak-to-AveragePowerRatio, PAPR) of ofdm system higher.PAPR is defined as ofdm signal peak signal peak power and the ratio of average power signal, is expressed as

$PAPR=\frac{\underset{0\≤n\≤N_s-1}{max}|x\left(n\right){|}^2}{E\[|x\left(n\right){|}^2\]}$

Wherein, x (n) is OFDM baseband signal, and E [] represents calculation expectation.Weighing the most frequently used method of PAPR performance is compensate cumulative distribution function (ComplementaryCumulativeDistributionFunction, CCDF), represent that the PAPR of data block exceedes the probability of given thresholding λ: CCDF=P{PAPR> λ }.

Because ofdm signal has larger PAPR, usually require that transmitter intermediate power amplifier must have larger linear dynamic range, especially when transmitter nonlinear power amplifier is operated in close to saturation point, signal can produce serious out of band spectrum and leak and inband distortion, cause monkey chatter and bit error rate (BitErrorRate, BER) hydraulic performance decline.Conventional PAPR suppression technology is mainly divided three classes, i.e. predistortion technology, signal coding technology, probability class technology.Wherein, signal coding technology and probability class technology need transmitting additional information for receiving terminal demodulation, and in predistortion technology, companding transform technology needs to carry out the process of solution companding at receiving terminal.Consider performance and the hardware implementing complexity of PAPR suppressing method, in predistortion technology, iteration limit filtration technology is comparatively simple and practical PAPR suppressing method.

Traditional iteration limit filtration method (IterativeClippingandFiltering, ICF) comprises amplitude limit and filtering two-step pretreatment.Amplitude limiting processing is by over-sampling J ofdm signal { x doubly _u(n), n=0 ..., JN _sthe amplitude of-1}, compared with thresholding A, is carried out amplitude limiting processing to the amplitude more than A, is expressed as

$x_{cu}\left(n\right)=\left\{\begin{array}{cc}\left|x_u\left(n\right)e^{{\mathrm{j\&theta;}}_{x_u\left(n\right)}}\right|,& \left|x_u\left(n\right)\right|\&le;A\\ {\mathrm{Ae}}^{{\mathrm{j\&theta;}}_{x_u\left(n\right)}},& \left|x_u\left(n\right)\right|<A\end{array}\right.,n=0,\mathrm{...},{\mathrm{JN}}_s-1$

Wherein, x _cun () is the ofdm signal after amplitude limiting processing, | x _u(n) | be ofdm signal amplitude before amplitude limiting processing, for ofdm signal phase place before amplitude limiting processing.By inserting N between centering _s(J-1) frequency-region signal of individual zero carry out JN _spoint IFFT conversion obtains over-sampling J ofdm signal doubly.

In order to reduce the out of band spectrum interference that amplitude limiting processing is introduced, usually frequency domain low-pass ripple is carried out to the signal after amplitude limit.First, by the signal { x after amplitude limit _cu(n), n=0 ..., JN _s-1} passes through JN _sthe FFT of point transforms to frequency domain and obtains frequency-region signal { X _cu(k), k=0 ..., JN _s-1}; Then, by { X _cu(k), k=0 ..., JN _sn in the middle of-1} _s(J-1) individual sampling zero setting.Because filtering process will cause ofdm signal peak regeneration, therefore need repeatedly to repeat above-mentioned limit filtration process to ofdm signal, obtain good PAPR rejection.

From the angle analysis that method realizes, traditional iteration limit filtration method needs signal x (n)=A (n) e before amplitude limiting processing ^{j θ (n)}amplitude A (n) polar coordinates and phase theta (n) is transformed to from real part and imaginary part, according to amplitude limit thresholding, amplitude limit is carried out to amplitude, phase place remains unchanged,, the amplitude A (n) of signal under polar coordinates and phase theta (n) are transformed to real part Re{x (n) of the ofdm signal under rectangular coordinate after amplitude limiting processing } and imaginary part Im{x (n).It can thus be appreciated that, just need to carry out 2 coordinate transforms in the amplitude limiting processing of 1 iterative process, so the iterative process of class ICF method realizes comparatively complicated.

Summary of the invention

The technical problem that the present invention solves is: overcome the deficiencies in the prior art, provides a kind of iteration limit filtration to reduce the method for ofdm signal peak-to-average force ratio.First the method carries out amplitude limiting processing according to amplitude limit factor-alpha according to limit ratio γ to the real part of OFDM time-domain signal and imaginary part respectively; Then, filtering and constellation extension are carried out to OFDM frequency-region signal, and transform to time domain continuation next iteration amplitude limiting processing.The amplitude limiting processing unit of the method carries out amplitude limit according to amplitude limit Factor minute safety pin to the real part of signal and imaginary part, by suitably selecting the amplitude limit factor, amplitude limiting processing only needs add operation, and interative computation unit adopts repeatedly reuse plan, can obtain good systematic function while reducing complexity.

The technical scheme that the present invention solves is: a kind of iteration limit filtration reduces the method for ofdm signal peak-to-average force ratio, and step is as follows:

(1) extract OFDM time domain baseband signals x (n), n=0 ..., N _sthe real part of-1} and imaginary part, calculating OFDM time domain baseband signals power P (n), n=0 ..., N _s-1}, is expressed as

P(n)＝(Re{x(n)}) ²+(Im{x(n)}) ²

In formula, Re{x (n) } and Im{x (n) be respectively real part and the imaginary part of the n-th sampling x (n) in OFDM time domain baseband signals; OFDM time domain baseband signals x (n), n=0 ..., N _sthe expression formula of-1} is

$x\left(n\right)=\frac{1}{\sqrt{N_s}}\underset{k=0}{\overset{N_s-1}{\&Sigma;}}X\left(k\right)e^{j2\mathrm{\&pi;}kn/N_s},n=0,...,N_s-1$

In formula, N _dfor the effective sub-carrier number of OFDM frequency domain, N _sfor the size of IFFT/FFT conversion, X (k) is that kth subcarrier raises the frequency domain data made, by being N by length _dmPSK/MQAM modulation multiple random sequence be mapped to the mutually orthogonal N of OFDM frequency domain _sthe N of individual subcarrier _don individual low frequency component, (N _s-N _d) individual high fdrequency component zero padding, obtain OFDM frequency domain sequence

(2) by calculate in step (1) OFDM time domain baseband signals power P (n), n=0 ..., N _s-1} and amplitude limit thresholding P ₀compare, according to the amplitude limit factor-alpha of setting, preferred 0< α <1, to more than P ₀ofdm signal sampling real part and imaginary part carry out amplitude limit, be expressed as

$\mathrm{Re}\left\{x_C\left(n\right)\right\}=\left\{\begin{array}{cc}\mathrm{Re}\left\{x\left(n\right)\right\},& P\left(n\right)\&le;P_0\\ \mathrm{\&alpha;}\&CenterDot;\mathrm{Re}\left\{x\left(n\right)\right\},& P\left(n\right)<P_0\end{array}\right.,\mathrm{Im}\left\{x_C\left(n\right)\right\}=\left\{\begin{array}{cc}\mathrm{Im}\left\{x\left(n\right)\right\},& P\left(n\right)\&le;P_0\\ \mathrm{\&alpha;}\&CenterDot;\mathrm{Im}\left\{x\left(n\right)\right\},& P\left(n\right)<P_0\end{array}\right.$

In formula, x _cn () is the OFDM time domain baseband signals after amplitude limit, amplitude limit thresholding P ₀by the limit ratio γ set and OFDM time domain baseband signals x (n), n=0 ..., N _sthe average power E of-1} [| x (n) | ²] determine, i.e. P ₀=10 ^γ/10e [| x (n) | ²].

(3) the OFDM time domain baseband signals { x after amplitude limit step (2) obtained _c(n), n=0 ..., N _s-1} passes through N _spoint FFT conversion obtains frequency-region signal { X _c(k), k=0 ..., N _s-1}, by { X _c(k), k=0 ..., N _sn in the middle of-1} _s-N _dindividual sampled value zero setting, obtains the signal { X after frequency domain filtering _cF(k), k=0 ..., N _s-1}, is expressed as

(4) by the signal { X after step (3) frequency domain filtering _cF(k), k=0 ..., N _s-1} passes through N _spoint IFFT conversion obtains the OFDM time-domain signal { x after 1 limit filtration _cF(n), n=0 ..., N _s-1};

(5) setting maximum iteration time is I, the OFDM time-domain signal { x after the limit filtration obtain step (4) _cF(n), n=0 ..., N _s-1} replace input in step (1) OFDM time domain baseband signals x (n), n=0 ..., N _s-1}, repeats step (1) ~ (4), until iterations meets the maximum iteration time I of setting, and the OFDM time-domain signal { x after the limit filtration that the I time step (4) is obtained _cF(n), n=0 ..., N _sthe value of-1}, as after iteration limit filtration OFDM time-domain signal y (n), n=0 ..., N _s-1}.

The peak-to-average power ratio (PAPR) of ofdm signal is expressed as OFDM time-domain signal peak signal peak power and the ratio of average power signal, is expressed as

$PAPR=\frac{\underset{0\&le;n\&le;N_s-1}{max}|x\left(n\right){|}^2}{E\&lsqb;|x\left(n\right){|}^2\&rsqb;}$

In formula, x (n) is OFDM baseband signal, and E [] represents calculation expectation.

The compensation cumulative distribution function (CCDF) weighing ofdm signal PAPR performance is expressed as

CCDF＝P{PAPR>λ}

In formula, λ is thresholding given when weighing PAPR performance.

Based on the OFDM time domain baseband signals of above-mentioned transmitting terminal, the implementation structure of the OFDM peak-to-average ratio suppressing method of iteration limit filtration proposed by the invention as shown in Figure 1, has following steps:

(1) signal power is calculated

OFDM baseband signal x (n), n=0 ..., N _s-1} is complex signal, according to its real part and imaginary part calculate OFDM baseband signal power P (n), n=0 ..., N _s-1}, is expressed as

P(n)＝(Re{x(n)}) ²+(Im{x(n)}) ²

In formula, Re{x (n) } and Im{x (n) be respectively real part and the imaginary part of the n-th sampling x (n) in OFDM time domain baseband signals.

(2) amplitude limiting processing of real part and imaginary part

By calculate in step (1) OFDM baseband signal power P (n), n=0 ..., N _s-1} and amplitude limit thresholding P ₀compare, according to amplitude limit factor-alpha (usually getting 0< α <1) to more than P ₀ofdm signal sampling real part and imaginary part carry out amplitude limit, be expressed as

$\mathrm{Re}\left\{x_C\left(n\right)\right\}=\left\{\begin{array}{cc}\mathrm{Re}\left\{x\left(n\right)\right\},& P\left(n\right)\&le;P_0\\ \mathrm{\&alpha;}\&CenterDot;\mathrm{Re}\left\{x\left(n\right)\right\},& P\left(n\right)<P_0\end{array}\right.,\mathrm{Im}\left\{x_C\left(n\right)\right\}=\left\{\begin{array}{cc}\mathrm{Im}\left\{x\left(n\right)\right\},& P\left(n\right)\&le;P_0\\ \mathrm{\&alpha;}\&CenterDot;\mathrm{Im}\left\{x\left(n\right)\right\},& P\left(n\right)<P_0\end{array}\right.$

In formula, x _cn () is the ofdm signal after amplitude limiting processing.Amplitude limit thresholding P ₀by limit ratio γ and average power signal E [| x (n) | ²] determine, γ=10log ₁₀(E [| x (n) | ²]/(α ²p ₀)), when E [| x (n) | ²during]=1, P ₀=10 ^γ/10/ α ².

(3) FFT conversion and frequency domain filtering

Signal { x after the amplitude limit that step (2) is obtained _c(n), n=0 ..., N _s-1} passes through N _spoint FFT conversion obtains frequency-region signal { X _c(k), k=0 ..., N _s-1}.By { X _c(k), k=0 ..., N _sn in the middle of-1} _s-N _dindividual sampled value zero setting, obtains the signal { X after frequency domain filtering _cF(k), k=0 ..., N _s-1}, is expressed as

$X_{CF}\left(k\right)=\left\{\begin{array}{cc}0,& N_d/2\&le;k\&le;(N_s-N_d/2-1)\\ X_C\left(k\right),& else\end{array}\right.$

In formula, N _dfor effective sub-carrier number of ofdm signal, N _sfor IFFT/FFT size.

(4) IFFT conversion

By the signal { X after frequency domain filtering _cF(k), k=0 ..., N _s-1} passes through N _spoint IFFT conversion obtains the OFDM time-domain signal { x after 1 limit filtration _cF(n), n=0 ..., N _s-1}.

If maximum iteration time is I, to the signal { x that step (4) obtains _cF(n), n=0 ..., N _s-1} repeats step (1) ~ (4), until iterations meets the maximum iteration time I of setting, obtain the OFDM time-domain signal after iteration limit filtration y (n), n=0 ..., N _s-1}.Wherein, the selection of maximum iteration time I needs the peak-to-average force ratio considering system to suppress the implementation complexity required and peak-to-average force ratio suppresses.

In the iteration limit filtration method that proposes of Main Analysis, amplitude limit calculates the implementation structure with reuse plan below.

1. amplitude limit calculates: when being the integral number power superposition of 2 by selecting the value of amplitude limit factor-alpha, the multiplying in above-mentioned steps (2) can be completed by addition.For the real part clipping operation of signal, limit width factor-alpha=α ₁2 ^-1+ α ₂2 ^-2+ α _l2 ^-ll is had in the binary number represented ₀position is 1, l-l ₀position is 0, and data sampling is all quantified as M bit in the implementation.The amplitude limiting processing step simplified is as follows:

A () will exceed thresholding P ₀the M bit of ofdm signal real part move to right respectively and obtain l ₀individual binary number, and by l ₀individual binary number is added;

B () can obtain M bit by the result be added in brachymemma step (a), be the ofdm signal real part numerical value after 1 iteration amplitude limiting processing.

2. reuse plan: in iteration limit filtration method, each iteration needs to calculate 1 FFT conversion respectively and converts with 1 IFFT.In the implementation, by higher than sample rate L ₀repeat under clock doubly to call the mode that 1 FFT unit completes FFT/IFFT conversion in successive ignition computing, FFT computing unit required in ICF method implementation procedure can be reduced, save and realize resource.

The present invention's advantage is compared with prior art:

(1) iteration limit filtration method of the present invention does not need to transmit extra information for receiving terminal demodulation, do not need to carry out extra operation at receiving terminal and compensate method for suppressing peak to average ratio impact, only need to carry out simple process to the real part of signal and imaginary part according to the amplitude limit factor according to limit ratio at transmitting terminal yet;

(2) iteration limit filtration method of the present invention is in step (2), multiply operation can be avoided by suitably selecting the amplitude limit factor, complete amplitude limiting processing by means of only add operation, and amplitude limiting processing result directly can transform to frequency domain by FFT carries out frequency domain filtering process;

(3) iteration limit filtration method of the present invention is in step (3) with step (4), repeating to call higher than the work clock of ofdm signal sample rate the IFFT/FFT operation that FFT unit realizes in successive ignition process by adopting, having saved the hardware implementing resource that in OFDM transmitter, peak-to-average force ratio suppresses;

(4) iteration limit filtration method of the present invention has lower implementation complexity, under identical iterations and limit ratio condition, by suitably selecting the amplitude limit factor, the PAPR being better than conventional iterative limit filtration can be obtained when performance of BER is close.

Accompanying drawing explanation

Fig. 1 is the implementation structure figure that iteration limit filtration proposed by the invention reduces Peak-to-Average Power Ratio method;

Fig. 2 is the method flow diagram that iteration limit filtration proposed by the invention reduces Peak-to-Average Power Ratio method;

Fig. 3 is for adopting the CCDF curve of iteration limit filtration method of the present invention and conventional method (ICF);

Fig. 4 is for adopting the BER performance schematic diagram of iteration limit filtration method of the present invention and conventional method (ICF).

Embodiment

Basic ideas of the present invention are: a kind of iteration limit filtration reduces the method for ofdm signal peak-to-average force ratio, first calculate OFDM time-domain signal power, will exceed power threshold P according to amplitude limit factor-alpha ₀oFDM real part and imaginary part carry out amplitude limiting processing respectively; Secondly the OFDM time-domain signal after amplitude limiting processing is transformed to OFDM frequency-region signal by FFT carries out frequency domain filtering; Then the signal after frequency domain filtering is transformed to OFDM time-domain signal by IFFT and completes the process of 1 iteration limit filtration; Finally repeat above-mentioned iteration limit filtration processing procedure according to maximum iteration time, method in the present invention does not need ofdm system to transmit extra information, do not need to carry out on ofdm signal the impact that extra process compensates amplitude limiting processing of making a start at receiving terminal yet, by suitably selecting amplitude limit factor-alpha, only need to complete amplitude limiting processing by add operation, and result directly can convert laggard row frequency domain filtering by FFT; Can adopt the multiplexing FFT/IFFT converter unit of work clock higher than sample rate in the computing of iteration limit filtration, that saves peak-to-average force ratio suppression realizes resource.

Below in conjunction with accompanying drawing and concrete embodiment, the present invention is described in further detail.Following instance for illustration of the present invention, but is not used for limiting the scope of application of the present invention.

Main thought of the present invention is: provide a kind of iteration limit filtration to reduce the method for ofdm signal peak-to-average force ratio.The iteration limit filtration method proposed does not need transmitting additional information or carries out extra process at receiving terminal, only needs to carry out amplitude limiting processing to the real part of ofdm signal and imaginary part according to amplitude limit factor-alpha according to limit ratio γ at transmitting terminal.As shown in Figure 1 first iteration limit filtration of the present invention reduces the implementation structure of the method for ofdm signal peak-to-average force ratio, according to ofdm signal x (n), n=0 ..., N _sthe real part of-1} and imaginary part calculate signal power P (n), n=0 ..., N _s-1}; Secondly, according to amplitude limit factor-alpha to exceeding power threshold P ₀the real part of the signal of (being calculated by limit ratio γ) and imaginary part carry out amplitude limit, and be the integral number power superposition of 2 by selection amplitude limit factor-alpha, amplitude limiting processing can avoid multiply operation, realizes by means of only add operation; Then, by the signal { x after amplitude limit _c(n), n=0 ..., N _s-1} passes through N _sthe FFT of point transforms to frequency domain, and by N in the middle of frequency-region signal _s-N _dindividual sampled value zero setting, completes frequency domain filtering; Finally, the signal after frequency domain filtering is passed through N _spoint IFFT transform to time domain obtain 1 iteration limit filtration after ofdm signal.Repeat above-mentioned steps, until iterations equals the maximum iteration time I arranged, complete OFDM peak-to-average ratio and suppress.

In iteration limit filtration method of the present invention, ofdm signal has following characteristics:

(1) in OFDM transmitting terminal, if the effective sub-carrier number of OFDM frequency domain is N _d, the size of IFFT/FFT conversion is N _s.Be N by length _dmPSK/MQAM modulation multiple random sequence be mapped to the mutually orthogonal N of OFDM frequency domain _sthe N of individual subcarrier _don individual low frequency component, (N _s-N _d) individual high fdrequency component zero padding, obtain OFDM frequency domain sequence by OFDM frequency domain sequence X (k), k=0 ..., N _s-1} carries out N _sthe IFFT conversion of point obtains OFDM time domain baseband signals, is expressed as

$x\left(n\right)=\frac{1}{\sqrt{N_s}}\underset{k=0}{\overset{N_s-1}{\&Sigma;}}X\left(k\right)e^{j2\mathrm{\&pi;}kn/N_s},n=0,...,N_s-1$

In formula, X (k) is the frequency domain data of a kth subcarrier rise system.

(2) peak-to-average power ratio (PAPR) of ofdm signal is expressed as OFDM time-domain signal peak signal peak power and the ratio of average power signal, is expressed as

$PAPR=\frac{\underset{0\&le;n\&le;N_s-1}{max}|x\left(n\right){|}^2}{E\&lsqb;|x\left(n\right){|}^2\&rsqb;}$

In formula, x (n) is OFDM baseband signal, and E [] represents calculation expectation.

(3) the compensation cumulative distribution function (CCDF) weighing ofdm signal PAPR performance is expressed as

CCDF＝P{PAPR>λ}

In formula, λ is thresholding given when weighing PAPR performance.

Based on the OFDM time domain baseband signals of above-mentioned transmitting terminal, the flow chart of the OFDM peak-to-average ratio suppressing method of iteration limit filtration proposed by the invention as shown in Figure 2, has following steps:

(1) signal power is calculated

OFDM baseband signal x (n), n=0 ..., N _s-1} is complex signal, according to its real part and imaginary part calculate OFDM baseband signal power P (n), n=0 ..., N _s-1}, is expressed as

P(n)＝(Re{x(n)}) ²+(Im{x(n)}) ²

In formula, Re{x (n) } and Im{x (n) be respectively real part and the imaginary part of the n-th sampling x (n) in OFDM time domain baseband signals.

(2) amplitude limiting processing of real part and imaginary part

By calculate in step (1) OFDM baseband signal power P (n), n=0 ..., N _s-1} and amplitude limit thresholding P ₀compare, according to amplitude limit factor-alpha (usually getting 0< α <1) to more than P ₀ofdm signal sampling real part and imaginary part carry out amplitude limit, be expressed as

$\mathrm{Re}\left\{x_C\left(n\right)\right\}=\left\{\begin{array}{cc}\mathrm{Re}\left\{x\left(n\right)\right\},& P\left(n\right)\&le;P_0\\ \mathrm{\&alpha;}\&CenterDot;\mathrm{Re}\left\{x\left(n\right)\right\},& P\left(n\right)<P_0\end{array}\right.,\mathrm{Im}\left\{x_C\left(n\right)\right\}=\left\{\begin{array}{cc}\mathrm{Im}\left\{x\left(n\right)\right\},& P\left(n\right)\&le;P_0\\ \mathrm{\&alpha;}\&CenterDot;\mathrm{Im}\left\{x\left(n\right)\right\},& P\left(n\right)<P_0\end{array}\right.$

In formula, x _cn () is the ofdm signal after amplitude limiting processing, the value of amplitude limit factor-alpha can be determined by the PAPR of diplomatic copy inventive method and BER performance.Amplitude limit factor-alpha value is too small, and after suppressing causing peak-to-average force ratio, the BER performance of signal is poorer; Amplitude limit factor-alpha value is excessive, be deteriorated by the PAPR caused under identical iterations, therefore the preferable range of amplitude limit factor-alpha is 0.8≤α <1.Amplitude limit thresholding P ₀by the limit ratio γ set and average power signal E [| x (n) | ²] determine, P ₀=10 ^γ/10e [| x (n) | ²], when E [| x (n) | ²during]=1, P ₀=10 ^γ/10.Adopt in the system of OFDM system in WLAN (wireless local area network) (WLAN), forth generation land mobile communication etc. based on IEEE802.11 standard, the usual value of limit ratio γ is 3≤γ <8 (unit: dB).

(3) FFT conversion and frequency domain filtering

Signal { x after the amplitude limit that step (2) is obtained _c(n), n=0 ..., N _s-1} passes through N _spoint FFT conversion obtains frequency-region signal { X _c(k), k=0 ..., N _s-1}.By { X _c(k), k=0 ..., N _sn in the middle of-1} _s-N _dindividual sampled value zero setting, obtains the signal { X after frequency domain filtering _cF(k), k=0 ..., N _s-1}, is expressed as

$X_{CF}\left(k\right)=\left\{\begin{array}{cc}0,& N_d/2\&le;k\&le;(N_s-N_d/2-1)\\ X_C\left(k\right),& else\end{array}\right.$

In formula, N _dfor effective sub-carrier number of ofdm signal, N _sfor IFFT/FFT size.

(4) by the signal { X after frequency domain filtering _cF(k), k=0 ..., N _s-1} passes through N _spoint IFFT conversion obtains the OFDM time-domain signal { x after 1 limit filtration _cF(n), n=0 ..., N _s-1}.

If maximum iteration time is I, when iterations is less than I, by the signal { x that step (4) obtains _cF(n), n=0 ..., N _s-1}, as the input signal of step (1), repeats step (1)-(4); When iterations equals I, by the signal { x that step (4) obtains _cF(n), n=0 ..., N _s-1} as after iteration limit filtration OFDM time-domain signal y (n), n=0 ..., N _s-1}.Wherein, the selection of maximum iteration time I needs the peak-to-average force ratio considering system to suppress the implementation complexity required and peak-to-average force ratio suppresses, and namely maximum iteration time I is larger, and the PAPR of the method is better, but implementation procedure is more complicated, the hardware implementing resource of needs is more.The usual span of maximum iteration time I is 1≤I≤20.

In the iteration limit filtration method that proposes of Main Analysis, amplitude limit calculates the implementation structure with reuse plan below.

1. amplitude limit calculates: when being the integral number power superposition of 2 by selecting the value of amplitude limit factor-alpha, the multiplying in above-mentioned steps (2) can be completed by addition.For the real part clipping operation of signal, limit width factor-alpha=α ₁2 ^-1+ α ₂2 ^-2+ α _l2 ^-ll is had in the binary number represented ₀position is 1, l-l ₀position is 0, wherein l to be amplitude limit factor representation be 2 integral number power superposition time sum term sum, and in realizing, data sampling is all quantified as M bit.The amplitude limiting processing step simplified in iteration amplitude limit method of the present invention is as follows:

A () will exceed thresholding P ₀the M bit of ofdm signal real part move to right respectively and obtain l ₀individual binary number, and by l ₀individual binary number is added;

B () can obtain M bit by the result be added in brachymemma step (a), be the ofdm signal real part numerical value after 1 iteration amplitude limiting processing.

2. reuse plan: in iteration limit filtration method, each iteration needs to calculate 1 FFT conversion respectively and converts with 1 IFFT.In the implementation, by higher than sample rate L ₀repeat under clock doubly to call the mode that 1 FFT unit completes FFT/IFFT conversion in successive ignition computing, FFT computing unit required in ICF method implementation procedure can be reduced, save and realize resource.

Effect of the present invention can be illustrated further by following emulation.

In ofdm communication system, if sub-carrier number N _s=256, the effective sub-carrier number N of user _u=180, signal bandwidth is B _w=3MHz, subcarrier spacing is Δ f=15kHz, up-sampling multiple L=19, amplitude limit factor-alpha=2 ^-1+ 2 ^-2+ 2 ^-3+ 2 ^-4=0.9375, limit ratio γ=5dB, maximum iteration time I=12.

Fig. 3, Fig. 4 are the CCDF curve and the BER performance that adopt iteration limit filtration method of the present invention and traditional IC F method.Can find out, when limit ratio γ, maximum iteration time I are identical, when amplitude limit factor-alpha=0.9375, the PAPR rejection of the inventive method is better than traditional IC F method, and BER performance is close with traditional IC F method.As CCDF=0.1%, the PAPR value of OFDM original signal, more than 11.33dB, adopts traditional IC F method the PAPR value of ofdm signal can be reduced to 8.30dB, adopts iteration limit filtration method proposed by the invention to be reduced to 8.18dB.

Non-elaborated part of the present invention belongs to techniques well known.

Claims (3)

1. iteration limit filtration reduces a method for ofdm signal peak-to-average force ratio, it is characterized in that step is as follows:

(1) extract OFDM time domain baseband signals x (n), n=0 ..., N _sthe real part of-1} and imaginary part, calculating OFDM time domain baseband signals power P (n), n=0 ..., N _s-1}, is expressed as

P(n)＝(Re{x(n)}) ²+(Im{x(n)}) ²

In formula, Re{x (n) } and Im{x (n) be respectively real part and the imaginary part of the n-th sampling x (n) in OFDM time domain baseband signals; OFDM time domain baseband signals x (n), n=0 ..., N _sthe expression formula of-1} is

$x\left(n\right)=\frac{1}{\sqrt{N_s}}\underset{k=0}{\overset{N_s-1}{\&Sigma;}}X\left(k\right)e^{j2\mathrm{\&pi;}kn/N_s},n=0,...,N_s-1$

In formula, N _dfor the effective sub-carrier number of OFDM frequency domain, N _sfor the size of IFFT/FFT conversion, X (k) is that kth subcarrier raises the frequency domain data made, by being N by length _dmPSK/MQAM modulation multiple random sequence be mapped to the mutually orthogonal N of OFDM frequency domain _sthe N of individual subcarrier _don individual low frequency component, (N _s-N _d) individual high fdrequency component zero padding, obtain OFDM frequency domain sequence

(2) by calculate in step (1) OFDM time domain baseband signals power P (n), n=0 ..., N _s-1} and amplitude limit thresholding P ₀compare, according to the amplitude limit factor-alpha of setting to more than P ₀ofdm signal sampling real part and imaginary part carry out amplitude limit, be expressed as

$\mathrm{Re}\left\{x_C\left(n\right)\right\}=\left\{\begin{array}{cc}\mathrm{Re}\left\{x\left(n\right)\right\},& P\left(n\right)\&le;P_0\\ \mathrm{\&alpha;}\&CenterDot;\mathrm{Re}\left\{x\left(n\right)\right\},& P\left(n\right)>P_0\end{array}\right.,\mathrm{Im}\left\{x_C\left(n\right)\right\}=\left\{\begin{array}{cc}\mathrm{Im}\left\{x\left(n\right)\right\},& P\left(n\right)\&le;P_0\\ \mathrm{\&alpha;}\&CenterDot;\mathrm{Im}\left\{x\left(n\right)\right\},& P\left(n\right)>P_0\end{array}\right.$

In formula, x _cn () is the OFDM time domain baseband signals after amplitude limit, amplitude limit thresholding P ₀by the limit ratio γ set and OFDM time domain baseband signals x (n), n=0 ..., N _sthe average power E of-1} [| x (n) | ²] determine, i.e. P ₀=10 ^γ/10e [| x (n) | ²];

(3) the OFDM time domain baseband signals { x after amplitude limit step (2) obtained _c(n), n=0 ..., N _s-1} passes through N _spoint FFT conversion obtains frequency-region signal { X _c(k), k=0 ..., N _s-1}, by { X _c(k), k=0 ..., N _sn in the middle of-1} _s-N _dindividual sampled value zero setting, obtains the signal { X after frequency domain filtering _cF(k), k=0 ..., N _s-1}, is expressed as

(4) by the signal { X after step (3) frequency domain filtering _cF(k), k=0 ..., N _s-1} passes through N _spoint IFFT conversion obtains the OFDM time-domain signal { x after 1 limit filtration _cF(n), n=0 ..., N _s-1};

(5) setting maximum iteration time is I, the OFDM time-domain signal { x after the limit filtration obtain step (4) _cF(n), n=0 ..., N _s-1} replace input in step (1) OFDM time domain baseband signals x (n), n=0 ..., N _s-1}, repeats step (1) ~ (4), until iterations meets the maximum iteration time I of setting, and the OFDM time-domain signal { x after the iteration limit filtration that the I time step (4) is obtained _cF(n), n=0 ..., N _sthe value of-1}, as after iteration limit filtration OFDM time-domain signal y (n), n=0 ..., N _s-1}.

2. a kind of iteration limit filtration according to claim 1 reduces the method for ofdm signal peak-to-average force ratio, it is characterized in that: the peak-to-average power ratio (PAPR) of described OFDM time domain baseband signals is expressed as OFDM time-domain signal peak signal peak power and the ratio of average power signal, is expressed as

$PAPR=\frac{\underset{0\&le;n\&le;N_s-1}{max}|x\left(n\right){|}^2}{E\&lsqb;|x\left(n\right){|}^2\&rsqb;}$

In formula, x (n) is OFDM baseband signal, and E [] represents calculation expectation.

3. a kind of iteration limit filtration according to claim 1 reduces the method for ofdm signal peak-to-average force ratio, it is characterized in that: the compensation cumulative distribution function (CCDF) weighing described OFDM time domain baseband signals PAPR performance is expressed as

CCDF＝P{PAPR>λ}

In formula, λ is thresholding given when weighing PAPR performance.