CN103312657A - Power line carrier data transmitting and processing method for improving channel throughput - Google Patents

Abstract

The invention discloses a power line carrier data transmitting and processing method for improving channel throughput. The power line carrier data transmitting and processing method includes: (1) subjecting data frames to detection, timing synchronization and FEQ (frequency domain equalization) initialization; (2) extracting phase differences among sub-bands among preamble word symbols; (3) calculating a unit sub-band phase shift of the preamble word symbols; (4) converting a unit sub-band phase shift of OFDM (orthogonal frequency division multiplexing) symbols; and (5) revising FEQ and restoring original data transmitted by a transmitting end. According to the power line carrier data transmitting and processing method which is applied to a receiving machine, an appropriate digital signal processing algorithm and technology is adopted, and on the basis that clock synchronization precision is not improved, continuous phase position deflection amount, of a final-to-be-judged signal, caused by clock synchronization is reduced; a large part of sub-bands are kept from performance reduction caused by poor clock synchronization, size of base band modulation is enlarged, and integral throughput of a system is increased.

Description

A kind of power carrier data transmission processing method for improving channel throughput

Technical field

The invention belongs to system clock and descend baseband modulation TDD-OFDM transmit-receive technology field synchronously, be specifically related to a kind of power carrier data transmission processing method for improving channel throughput.

Background technology

TDD-OFDM(time division duplex-orthogonal frequency division multiplexi) be a kind of several signals independent of each other are merged into one can be in the method for the composite signal of same channel, the basic principle of its transfer of data is that serial data stream is converted to the lower parallel data stream of N road speed, modulate respectively parallel transmission behind the N way carrier wave with them, the frequency domain sub-carriers is mutually orthogonal, overlapped in the time domain, thereby have very strong anti-channel fading ability and the higher availability of frequency spectrum, and can suppress well intersymbol interference.

Yet the factor of restricted T DD-OFDM communication system performance except uncontrollable channel performance itself, is unusual part and parcel synchronously; Generally include synchronously frame Timing Synchronization, carrier synchronization and clock frequency synchronous.Can not cause following consequence when clock frequency is good synchronously: the signal to noise ratio of input signal is descended, disturb the QAM(quadrature amplitude modulation of (ICI), baseband modulation between subcarrier) the planisphere phase place continues single direction rotation.Analyze and engineering practice shows, compare with the channel performance that usually runs into, the loss that the above two cause is very little, generally refuses special consideration and processing; And the asynchronous sampling location of causing of sample rate continues single-way moving, itself causes error except causing the signal phase rotation, and it continues unidirectional accumulation and can cause sampling window to overflow the symbol position, and then consists of subversive intersymbol interference (ISI).

In the relevant international industrial standard of up-to-date TDD-OFDM in recent years, for example the international standard of power line carrier communication G.9960 with the IEEE-1901 technology in, particularly in various Local Area Network and metropolitan area network (MAN), adopt the system clock simultaneous techniques, the clock synchronous of all transceivers in the system is arrived with the center relative frequency difference of main controller clock in certain limit, for example in the 1ppm.On this basis, guaranteed in the frame length scope that system needs, the above-mentioned subversive ISI that causes that moves because of sampling window can not occur.And planisphere is continued unidirectional phase rotating, because of frequency-domain equalizer (FEQ) by the symbol adaptive updates, be limited in limited range, for example within the phase range of the rotation in one to two symbolic range, be unlikely to continue accumulation and cause final unsentenced signal performance to decline to a great extent.Such arrangement has been saved and had been realized with the interpolation simultaneous techniques that needs larger system resource in the class standard in the past, replaces technology and the cost requirement higher to frequency synthesizer.

Through technical development for many years, adopt the self organizing maps technology of high quadrature amplitude modulation size to be tending towards ripe in the ofdm system, with the maximum channel handling capacity.Such as PLC in employed band limits, the signal to noise ratio of channel is in good, very most of frequency range can realize 4096-QAM and higher base band mapping size.But on above-mentioned clock synchronous quality (being in the 1ppm) basis, and only rely under the adaptation function condition of FEQ, very most of frequency range but can't make the performance that reaches like this match with higher channel performance, thereby can't make taking full advantage of of corresponding channel capacity.

In sum, if according to traditional ofdm system design and receiver processing method, be difficult to realize very large baseband modulation size.One of its reason is the adaptive correction of FEQ, can't avoid the phase shift that produces every between the adjacent-symbol that imperfect because of clock synchronous (following we are referred to as " clock synchronous is remaining ") that symbol exists causes.Obviously, solving the simplest direct method of this contradiction, is exactly the higher quality requirement of frequency synthesizer proposition to clock generating.For example with the performance behind the initialization clock synchronous, bring up to 0.5ppm, even 0.2ppm.In such system receiver, the problems referred to above are just solved naturally, measure such as 0.5 in addition the frequency difference of 0.2ppm be not difficult.Difficulty is adjusted its frequency synthesis in the frequency difference that frequency synthesizer provides according to measurement, reach with the console for centralized control frequency difference 0.5 in addition the 0.2ppm scope in, although this is not to realize for frequency synthesizer, need to pays larger cost, power consumption and technology and realize cost.For the commercial market of application maximum, this technology and cost pressure consist of very large obstacle, therefore classify these high size base band mapping indexs as option in the generally held standard.

Summary of the invention

For the existing above-mentioned technical problem of prior art, the invention provides a kind of power carrier data transmission processing method for improving channel throughput, can in the restricted situation of system clock synchronization accuracy, improve its baseband modulation size, increase the throughput of entire system.

A kind of power carrier data transmission processing method for improving channel throughput comprises the steps:

(1) Frame that receiving terminal is received carries out frame detection, Timing Synchronization and FEQ initialization successively;

(2) appoint the introductory word before Frame and get two leading symbols, the phase place by between these two leading symbol corresponding subband signals that relatively receive extracts the phase difference that each subband of leading symbol causes because of clock synchronous remnants;

(3) according to the phase difference of each subband of leading symbol, calculate the unit subband phase shift of leading symbol;

(4) according to the unit subband phase shift of leading symbol, convert out the unit subband phase shift of OFDM symbol;

(5) according to the unit subband phase shift of OFDM symbol, before arriving, the OFDM symbol revises the FEQ of its each subband; And then according to revised FEQ and the OFDM symbol that receives, calculate and go back the initial data that original sender sends.

Comprise successively front introductory word, frame head and payload in the described Frame; Front introductory word, frame head and payload consist of by some symbols, and the symbol of introductory word is called leading symbol before forming, and the symbol that forms frame head or payload is called the OFDM symbol.

Preferably, in the described step (2), before Frame, get first leading symbol the introductory word and carry out mutually bit comparison of subband signal with last leading symbol.Its objective is the noise in order to suppress to introduce in the evaluation phase; Time gap between two leading word symbols that are used to estimate is larger, and the phase difference because of the remaining introducing of clock synchronous between them is also larger, and noise is constant.Signal to noise ratio in the estimation is also higher thus, and result's precision is also higher.

Preferably, in the described step (3), the method for calculating the subband phase shift of leading symbol unit is as follows: at first, from each subband phase difference of leading symbol, get the phase difference of interior each subband of institute's band limits that uses

Then, calculate the subband phase shift of leading symbol unit according to following formula

Wherein: k be frequency range by the sub-band serial number of first subband in the whole frequency range of leading symbol in the use band limits, m is that frequency range is the interior sub-band sum of institute's use band limits, i is natural number and k≤i≤k+m-1.

Only adopt the interior higher subband of centre frequency of institute's band limits that uses, can avoid Noise and Interference on the impact of measurement, precision is improved and make as a result.

In the described step (4), according to formula

Convert out the subband phase shift Δ θ of unit of OFDM symbol; Wherein,

Be the unit subband phase shift of leading symbol, K is given proportionality coefficient.

K is the practical experience value, and it utilizes actual measurement to determine according to system configuration.

In the described step (5), according to the unit subband phase shift of OFDM symbol, the process of the FEQ of each subband of correction OFDM symbol is as follows:

For arbitrary subband of OFDM symbol, according to the FEQ of following this subband of formula correction; And travel through according to this all subbands of OFDM symbol;

$F={\mathrm{Ae}}^{j({\mathrm{\Ψ}}_0+\mathrm{\θ})}$

Wherein: F is the filter coefficient of FEQ after revising, and the expression formula of FEQ filter coefficient is Aexp{j Ψ before revising ₀, A is the amplitude of FEQ, Ψ ₀For FEQ revises front phase place, j is imaginary unit, and θ is phase increment and θ=m* Δ θ, and Δ θ is the unit subband phase shift of OFDM symbol, and m is the sub-band serial number of this subband in the whole frequency range of OFDM symbol.

Preferably, in the described step (5), according to the unit subband phase shift of OFDM symbol, the process of the FEQ of each subband of correction OFDM symbol is as follows:

For arbitrary subband of OFDM symbol, according to the FEQ of following this subband of formula correction; And travel through according to this all subbands of OFDM symbol;

F＝[U-Vθ]+j[Uθ+V]

Wherein: F is the filter coefficient of FEQ after revising, and the expression formula of FEQ filter coefficient is Aexp{j Ψ before revising ₀, FEQ filter coefficient Aexp{j Ψ before U and V are respectively and revise ₀Real part and the imaginary part of plural expression-form U+jV, A is the amplitude of FEQ, Ψ ₀For FEQ revises front phase place, j is imaginary unit, and θ is phase increment and θ=m* Δ θ, and Δ θ is the unit subband phase shift of OFDM symbol, and m is the sub-band serial number of this subband in the whole frequency range of OFDM symbol.

Only have real multiplications in the following formula twice, when a symbol contains a large number of (thousands of even more) subband, therefore the priori phase place of FEQ is upgraded hardware or the software calculating operation saved will reduce by half.

In the described step (5), calculate the initial data of going back the original sender transmission according to formula Y=X*F; Wherein, X is that receiving terminal receives the arbitrary subband signal in the OFDM symbol, and Y is the subband signal of the corresponding X of transmitting terminal transmission, and F is the revised FEQ filter coefficient of subband under the X.

In the described step (1), Frame is carried out the initialized method of FEQ as follows:

A. all leading symbol additions in the introductory word before the Frame are averaging, and average result is carried out the FFT(fast Fourier transform);

B. for arbitrary subband Z of leading symbol _i, f=y/x calculates Z according to formula _iThe FEQ filter coefficient; Wherein, f represents Z _iThe FEQ filter coefficient, x is corresponding Z in the leading symbol that receives of receiving terminal _iInterior subband signal, y are the subband signal of the corresponding x of transmitting terminal transmission; Determine according to this Z _iNext subband Z _I+1The FEQ filter coefficient;

C. from the OFDM symbol, determine at Z _iSeveral subbands in the corresponding frequency band are with Z _iThe FEQ filter coefficient give these several subbands middle subband U1; From the OFDM symbol, determine at Z _I+1Several subbands in the corresponding frequency band are with Z _I+1The FEQ filter coefficient give these several subbands middle subband U2;

D. after determining to obtain the FEQ filter coefficient of subband U1 and subband U2, estimate the FEQ filter coefficient of OFDM symbol from subband U1 to each subband between subband U2 by the difference estimation technique;

E. according to step B～D, traversal is tried to achieve the FEQ filter coefficient of each subband of OFDM symbol.

Useful technique effect of the present invention is: the system clock of above-mentioned feature synchronously, in the TDD-OFDM digital communication system, the intersymbol signal phase movement deviation amount that the residual volume that causes when the synchronous imperfection of clock causes surpasses the high SNR(signal to noise ratio of high-quality that actual capabilities reach) during the physical channel ability, adopt that the present invention proposes to FEQ priori phase compensating method, can make transceiver adopt more modulation voltage near the channel capacity limit, thereby obtain more the Digital Transmission throughput near system.

Description of drawings

Fig. 1 is the steps flow chart schematic diagram of the inventive method.

Fig. 2 is the organigram of front introductory word.

Fig. 3 is for adopting the most front last two leading symbols to calculate the schematic flow sheet of the remaining unit of clock synchronous phase shift.

Fig. 4 is the measurement result schematic diagram of each subband phase difference of leading symbol.

Fig. 5 (a) be the time domain input signal SNR=45dB under, do not adopt the present invention to process later the OFDM symbol by subband performance schematic diagram.

Fig. 5 (b) be the time domain input signal SNR=45dB under, adopt the present invention to process later the OFDM symbol by subband performance schematic diagram.

Fig. 6 (a) is under the SNR=20dB for test environment, and frame signal arrives the output performance schematic diagram that detects.

Fig. 6 (b) is under the SNR=-3dB for test environment, and frame signal arrives the output performance schematic diagram that detects.

Fig. 7 (a) is under the SNR=20dB for test environment, the output performance schematic diagram of frame Timing Synchronization.

Fig. 7 (b) is under the SNR=-3dB for test environment, the output performance schematic diagram of frame Timing Synchronization.

Embodiment

In order more specifically to describe the present invention, below in conjunction with the drawings and the specific embodiments technical scheme of the present invention and relative theory thereof are elaborated.

As shown in Figure 1, a kind of power carrier data transmission processing method for improving channel throughput comprises the steps:

(1) to Frame detect, Timing Synchronization and FEQ initialization.

In the digital communication system that adopts TDD transmitting-receiving mode, frame is the basic group of receiving and sending messages.Although the frame structure of different industrial standards definition has such-and-such difference, general all such as the infrastructure mode of table 1:

Table 1

In the OFDM type systematic, above-mentioned three parts all are the symbols in several regular lengths of frequency domain definition, by IDFT(discrete fourier inverse transformation) be turned to that time domain realizes.In general, front introductory word all identical by several and the opposite leading symbol that repeats forms, and arrives detection, Timing Synchronization, equalizer initialization, local channel performance estimation etc. to carry out signal; Take the IEEE-1901 standard as example, each leading symbol is of a size of 512 time domain unit's digital signals.

Frame head contains the information such as the type of payload portions and structure, and payload contains real useful user and all kinds of control and management information.Frame head and payload are comprised of the OFDM symbol of equal length.Frame head only has the OFDM symbol usually one or two.And payload can be sky, also may be comprised of a plurality of OFDM symbols; The symbol dimension of introductory word is usually all short than OFDM symbol before forming, and take the IEEE-1901 standard as example, generally each frame head and payload symbol are of a size of 4096 time domain unit signals.

The Frame that present embodiment at first needs receiving terminal is received carries out frame detection, Timing Synchronization and FEQ initialization successively; Through development communication technologies for many years, large body technique is tending towards ripe, and the principle of execution is similar; Below sketch basic realization principle and the method for these several parts.

A. frame signal detects

Because the needed resource of frame Timing Synchronization that detects after the frame arrival is very large.Therefore in TDD transmitting-receiving process, after a frame sends, receive the beginning that signal will arrive through the system information prompting, receiver does not start Timing Synchronization, take the less frame signal detection of resource but start first.When acknowledgement frame signals enters receiver, restart and take the larger Timing Synchronization signal of resource and process.

Figure 2 shows that a kind of typical frame preamble word structure.In the standard of some system, its symbol is called as " short sequence ", because of its length short and small more a lot of than frame head and payload symbol lengths in most of frame structures.

Frame arrives in the various noises and interference that the purpose that detects is objective reality in physical channel, judges the arrival of the frame of native system definition.Frame signal detects, and normally carries out with auto-correlation computation.Namely distance is carried out auto-correlation computation take the length L of a leading word symbol (SP in this example) as auto-correlation.

$P\left(d\right)=\underset{m=0}{\overset{L-1}{\mathrm{\Σ}}}{r}_{d+m}^{*}{r}_{d+m+L}$

Wherein, P is the relative amplitude of auto-correlation output as a result, and r is the numeral sample sequence of input, and d is the numeral sample sequence number of this sequence, and m is the increment sequence number of autocorrelation calculation, and L is that the auto-correlation distance (is leading word symbol length, in the example of Fig. 2, L=512).P is a relative value, and when short sequence entered the frame arrival detection of following formula expression, the absolute value of output P presented very large numerical value.And noise, interference, or other irrelevant signals are when entering receiver, and because there is not periodicity shown in Figure 2 in it, autocorrelation is very weak, and the absolute value of output P is very little.

This primary expression directly uses in engineering and has such as inconvenience such as thresholding arrange.Real engineering reality has a lot of optimization methods as the case may be, improves the reliability of detection or the setting of thresholding.But fundamental starting point is all shown in following formula.Fig. 6 is the performance that following formula is used after certain normalized.Fig. 6 (a) is the Output rusults under the SNR=20dB.The curve left side is the situation of only having Noise and Interference to mid portion, and relative amplitude is very low.And the Output rusults of right side after to be signal arrive, because relatively very large output amplitude appears in the strong correlation of its contained signal component.

The algorithm of introducing after various engineering optimizations improve, has outstanding detectability under Strong Noise Jamming Environment.Fig. 6 (b) is the Output rusults under the SNR=-3dB.Can find out, select suitable thresholding, be lower than in signal power under the environment of noise jamming, still have correct reliably signal to arrive detectability.

B. frame Timing Synchronization

The structure of the leading word symbol of frame is the industrial standard regulation, all is identical to transmitter and receiver.Therefore, between the signal that receives and known leading word symbol (time domain), carry out following computing cross-correlation, can obtain extraordinary timing effect.

$P\left(d\right)=\underset{m=0}{\overset{L-1}{\mathrm{\Σ}}}{r}_{d+m}{R}_m^{*}$

Wherein, P is the relative amplitude of cross-correlation output, and r is the numeral sample sequence of input, and R is the time domain samples of the leading word symbol of standard code, and d is the numeral sample sequence number, and m is the increment sequence number of cross-correlation calculation, and L is that sequence length (is leading word symbol length.In the example of Fig. 2, L=512).

The reliability of Timing Synchronization and precision depend on that standard is to the design of sequence length and sequence itself.In example system, under the SNR=-3dB environment, still have reliable and accurately frame timing effect.Fig. 7 (a) is the Output rusults under the SNR=20dB.Fig. 7 (b) is the Output rusults under the SNR=-3dB.Notice, the amplitude of output can not only be concerned about in computing cross-correlation output as frame detects, also must keep its phase information.Figure 7 shows that the result of real part.In the deterministic process, the real part imaginary part all must be considered, only consider real part or only consider that imaginary part all is worthless, the result that usually can lead to errors.

There are several Output rusults can supply to judge, to strengthen its reliability and timing accuracy.The first, cross-correlation amplitude (absolute value), the second, the distance between the peak value that occurs when surpassing thresholding (should accurately be L), the 3rd, by anti-phase to SM of SP shown in Figure 2.Notice, computing cross-correlation, a sample enter into entering of next sample, do not have intermediate object program to share between two computings, and must keep relatively phase place of its real imaginary part.Therefore it is much larger that the operand that the frame Timing Synchronization relates to detects the operand that relates to than frame arrival.But still there is way to reduce by a relatively large margin whole operand.This has exceeded the technical scope of this paper concern, does not further give unnecessary details.

The C.FEQ initialization

C1. all leading symbols in the introductory word before the Frame (opposite symbol adds negative sign) addition is averaging, and average result is carried out FFT;

C2. for arbitrary subband Z of leading symbol _i, f=y/x calculates Z according to formula _iThe FEQ filter coefficient; Wherein, f represents Z _iThe FEQ filter coefficient, x is corresponding Z in the leading symbol that receives of receiving terminal _iInterior subband signal, y are the subband signal of the corresponding x of transmitting terminal transmission; Determine according to this Z _iNext subband Z _I+1The FEQ filter coefficient;

C3. from the OFDM symbol, determine at Z _iSeveral subbands in the corresponding frequency band are with Z _iThe FEQ filter coefficient give these several subbands middle subband U1; From the OFDM symbol, determine at Z _I+1Several subbands in the corresponding frequency band are with Z _I+1The FEQ filter coefficient give these several subbands middle subband U2;

C4. after determining to obtain the FEQ filter coefficient of subband U1 and subband U2, estimate the FEQ filter coefficient of OFDM symbol from subband U1 to each subband between subband U2 by the difference estimation technique;

C5. according to step C2～C4, traversal is tried to achieve the FEQ filter coefficient of each subband of OFDM symbol.

In some early stage industrial standard, one to two and the isometric known symbol of significant character after front introductory word, have been arranged.In such cases, can disposablely obtain the good FEQ for effective symbol of quality.Cost is that the reference symbol of foot length has taken effective frame length, has reduced entire throughput.

(2) extract each subband phase difference between leading word symbol.

If clock frequency is perfectly synchronously, then the next identical symbol phase place that corresponding subband presents in receiver in front introductory word Central Plains also is identical.Because synchronously can not be flawless, along with time duration, the phase place between the corresponding subband of the leading word symbol in back and front be not exclusively equal.This is that clock synchronous remnants cause, need to extract and manage to compensate.

We describe as an example of a representative leading word structure example, and this leading word structure as shown in Figure 2; SP is the time-domain signal of identical leading word symbol on the contrary with SM among Fig. 2.When Timing Synchronization was finished, digital system found the switching point (black thick line position) of SP and SM.Because SM=-SP (SM) is FFT simultaneously with the anti-phase of first SP of Far Left and a SM of rightmost.This does not need extra hardware resource, because the initialization of FEQ also needs leading word symbol is carried out the identical FFT conversion of length.As previously mentioned, because between the corresponding symbol of the frequency domain of these two time-domain symbol, there are certain phase difference in the local receiver of objective reality and the synchronous residual volume of sampling clock between the transmitter of a distant place.

Because synchronously the SP that causes of remnants and-phase difference and subband center frequency between each frequency domain subband of SM be directly proportional.As shown in Figure 3, first leading word symbol SP_1 of time domain and last leading word symbol SM_2(are anti-phase), be respectively FFT obtain two isometric as a result SFFTP of frequency domain and-SFFTM.Among Fig. 3, be the subband of the frequency range at signal place from m+1 to M, the subband phase place of corresponding sequence number is subtracted each other, and obtains interior each subband corresponding phase of band poor.This phase difference is exactly because the additive phase during this period of time that the incomplete residual volume of clock frequency synchronized result causes moves.

(3) the unit subband phase shift of calculating leading symbol.

Principle and simulation result all show, the additive phase of each subband moves and the height of subband center frequency (starting at from base band) is in direct ratio.A simulation result when Fig. 4 is signal to noise ratio snr=45dB, the result of test, each subband noise statistics is close, but the phase difference that the synchronous remnants of clock frequency cause presents linear change substantially, be each subband test result difference affected by noise, the subband test result that centre frequency is higher better (affected by noise less); Still can take full advantage of the phase place that shows in the above results with the relation of subband center frequency linear change.

In the present embodiment, we calculate the subband phase shift of leading symbol unit according to phase difference corresponding to each subband of leading symbol

Concrete methods of realizing is as follows:

At first, from each subband phase difference of leading symbol, get phase difference corresponding to each subband in the band limits that uses Then, calculate the subband phase shift of leading symbol unit according to following formula

Wherein: k by the sub-band serial number of first subband in the whole frequency range of leading symbol between in the use band limits, m is that frequency range is the interior sub-band sum of institute's use band limits, i is natural number and k≤i≤k+m-1.Also as aforementioned, present embodiment is abandoned and is not used in those subbands that carry effective information in the various situations.

(4) the unit subband phase shift of conversion OFDM symbol.

Obtaining the subband phase shift of leading symbol unit

After, we are according to formula

Calculate the subband phase shift Δ θ of unit of OFDM symbol; Wherein, K is given proportionality coefficient.

Although Proportional coefficient K can be obtained by approximate calculation, easier by measuring of reality, it is irrelevant with the phase difference of measuring.If the protection part that symbol relates to is different, can measure respectively.Different protection gap lengths is a limited number of some constants in the standard, can give Estimation and Measurement to K values different under several situations respectively in advance.

Present embodiment adopts protection to be spaced apart one of several protection gap lengths of the long 1/6(standard code of symbol body), K=0.25.According to the different parameters of different system, the K value is different in a big way.

In commission, if if phase increment is very little, the signal to noise ratio that runs in the measurement reduces a lot, and measurement result is unreliable.Set up following implementation strategy: thresholding 1 is set, is lower than 40dB when inputting average SNR, this moment, the clock frequency synchronously remaining (for example, by standard 1ppm) of maximum possible was not enough to affect the decision amount that base band is shone upon size.Do not adopt phase adjusting method, can make

When the phase increment on measurement result unit's subband is little when the certain numerical value, the relative effect that the result is disturbed noise becomes large, also can make

The input average SNR more than 36dB, not enough 45dB again, then measurement result is more coarse, but its direction (namely just or negative) can be wrong.Can use half of measurement result this moment, like this, equals calibration step less, compensation rate weak (being equivalent to conservative the improvement).When measuring and the result of estimation surpasses maximum possible (for example the 2ppm maximum intersymbol phase shift that may cause is The time, be forced to:

(5) revise FEQ, go back the initial data that original sender sends.

After obtaining the subband phase shift Δ θ of unit of OFDM symbol, we are for arbitrary subband of OFDM symbol, according to the FEQ of following this subband of formula correction, and travel through according to this all subbands of OFDM symbol;

$F={\mathrm{Ae}}^{j({\mathrm{\Ψ}}_0+\mathrm{\θ})}$

Wherein: F is the filter coefficient of FEQ after revising, and the expression formula of FEQ filter coefficient is Aexp{j Ψ before revising ₀, A is the amplitude of FEQ, j is imaginary unit, Ψ ₀For FEQ revises front phase place, θ is phase increment and θ=m* Δ θ, and m is the sub-band serial number of this subband in the whole frequency range of OFDM symbol.

In real process, because receiver need to have the long period to wait for the arrival of next symbol after FEQ revises computing.To the phase compensation that needs in next symbol to carry out, can carry out phase place correction to FEQ in this section stand-by period.

Present embodiment is utilized DSP that the FEQ coefficient is carried out and is revised, and following formula can be decomposed into:

$F={\mathrm{Ae}}^{j({\mathrm{\Ψ}}_0+\mathrm{\θ})}={\mathrm{Ae}}^{{\mathrm{j\Ψ}}_0}{e}^{\mathrm{j\θ}}$

Wherein: first half Aexp{j Ψ ₀Be equivalent to revise front FEQ filter coefficient, it multiply by latter half e ^{J θ}And e ^{J θ}Be that the phase place correction is done balanced the use for successive character, be equivalent to the phase place correction to FEQ.Two numbers all are plural numbers.In the engineering, plural number all is that the form with the real part imaginary part exists.In other words, this coefficient makeover process is two complex multiplication.Complex multiplication needs four real arithmetics in the ordinary course of things.But under the requirement of the clock frequency coincident indicator of standard code, θ herein is in a small amount, and the phase shift every between the symbol corresponding subband that causes thus is very little (see figure 4) also; Therefore have following:

e ^jθ＝cosθ+jsinθ≈1+jθ

And then have:

$F={\mathrm{Ae}}^{j{\mathrm{\Ψ}}_0}[\cos \mathrm{\θ}+j\sin \mathrm{\θ}]$

$=[U+\mathrm{jV}][\cos \mathrm{\θ}+j\sin \mathrm{\θ}]$

$=[U\cos \mathrm{\θ}-V\sin \mathrm{\θ}]+j[U\sin \mathrm{\θ}+V\cos \mathrm{\θ}]$

$\≈[U-\mathrm{V\θ}]+j[\mathrm{U\θ}+V]$

Only have real multiplications in the following formula twice, when a symbol contains a large number of (thousands of even more) subband, therefore the priori phase place of FEQ is upgraded hardware or the software operation of saving will reduce by half; According to carrying out accordingly and directly carrying out several indifferences of performance according to general complex multiplication.

At last, we calculate according to formula Y=X*F and go back the data modulated signal that original sender sends; Wherein, X is that receiving terminal receives the arbitrary subband signal in the OFDM symbol, and Y is the subband signal of the corresponding X of transmitting terminal transmission, and F is the revised FEQ filter coefficient of subband under the X.

We carry out simulation comparison to present embodiment, as shown in Figure 5; Wherein Fig. 5 (a) is that by the signal performance of subband, obviously, the residual error of clock synchronous causes the hydraulic performance decline of high band under the synchronously rear maximum relative residual frequency difference 2ppm of system clock.The DSP Phase Tracking adjustment that Fig. 5 (b) present embodiment is carried out, the improvement that obtains is obvious.

Claims (8)

1. a power carrier data transmission processing method that is used for improving channel throughput comprises the steps:

(1) Frame that receiving terminal is received carries out frame detection, Timing Synchronization and FEQ initialization successively;

(2) appoint the introductory word before Frame and get two leading symbols, the phase place by between these two leading symbol corresponding subband signals that relatively receive extracts the phase difference that each subband of leading symbol causes because of clock synchronous remnants;

(3) according to the phase difference of each subband of leading symbol, calculate the unit subband phase shift of leading symbol;

(4) according to the unit subband phase shift of leading symbol, convert out the unit subband phase shift of OFDM symbol;

(5) according to the unit subband phase shift of OFDM symbol, before arriving, the OFDM symbol revises the FEQ of its each subband; And then according to revised FEQ and the OFDM symbol that receives, calculate and go back the initial data that original sender sends.

2. power carrier data transmission processing method according to claim 1 is characterized in that: in the described step (2), get first leading symbol the introductory word and carry out mutually bit comparison of subband signal with last leading symbol before Frame.

3. power carrier data transmission processing method according to claim 1, it is characterized in that: in the described step (3), the method of calculating the subband phase shift of leading symbol unit is as follows: at first, from each subband phase difference of leading symbol, get the phase difference of interior each subband of institute's band limits that uses

Then, calculate the subband phase shift of leading symbol unit according to following formula

Wherein: k be frequency range by the sub-band serial number of first subband in the whole frequency range of leading symbol in the use band limits, m is that frequency range is the interior sub-band sum of institute's use band limits, i is natural number and k≤i≤k+m-1.

4. power carrier data transmission processing method according to claim 1 is characterized in that: in the described step (4), according to formula

Convert out the subband phase shift Δ θ of unit of OFDM symbol; Wherein,

Be the unit subband phase shift of leading symbol, K is given proportionality coefficient.

5. power carrier data transmission processing method according to claim 1 is characterized in that: in the described step (5), according to the unit subband phase shift of OFDM symbol, the process of FEQ of revising each subband of OFDM symbol is as follows:

For arbitrary subband of OFDM symbol, according to the FEQ of following this subband of formula correction; And travel through according to this all subbands of OFDM symbol;

$F={\mathrm{Ae}}^{j({\mathrm{\Ψ}}_0+\mathrm{\θ})}$

Wherein: F is the filter coefficient of FEQ after revising, and the expression formula of FEQ filter coefficient is Aexp{j Ψ before revising ₀, A is the amplitude of FEQ, Ψ ₀For FEQ revises front phase place, j is imaginary unit, and θ is phase increment and θ=m* Δ θ, and Δ θ is the unit subband phase shift of OFDM symbol, and m is the sub-band serial number of this subband in the whole frequency range of OFDM symbol.

6. power carrier data transmission processing method according to claim 1 is characterized in that: in the described step (5), according to the unit subband phase shift of OFDM symbol, the process of FEQ of revising each subband of OFDM symbol is as follows:

For arbitrary subband of OFDM symbol, according to the FEQ of following this subband of formula correction; And travel through according to this all subbands of OFDM symbol;

F＝[U-Vθ]+j[Uθ+V]

Wherein: F is the filter coefficient of FEQ after revising, and the expression formula of FEQ filter coefficient is Aexp{j Ψ before revising ₀, FEQ filter coefficient Aexp{j Ψ before U and V are respectively and revise ₀Real part and the imaginary part of plural expression-form U+jV, A is the amplitude of FEQ, Ψ ₀For FEQ revises front phase place, j is imaginary unit, and θ is phase increment and θ=m* Δ θ, and Δ θ is the unit subband phase shift of OFDM symbol, and m is the sub-band serial number of this subband in the whole frequency range of OFDM symbol.

7. power carrier data transmission processing method according to claim 1 is characterized in that: in the described step (5), calculate the initial data of going back the original sender transmission according to formula Y=X*F; Wherein, X is that receiving terminal receives the arbitrary subband signal in the OFDM symbol, and Y is the subband signal of the corresponding X of transmitting terminal transmission, and F is the revised FEQ filter coefficient of subband under the X.

8. power carrier data transmission processing method according to claim 1 is characterized in that: in the described step (1), Frame is carried out the initialized method of FEQ as follows:

A. all leading symbol additions in the introductory word before the Frame are averaging, and average result is carried out FFT;

B. for arbitrary subband Z of leading symbol _i, f=y/x calculates Z according to formula _iThe FEQ filter coefficient; Wherein, f represents Z _iThe FEQ filter coefficient, x is corresponding Z in the leading symbol that receives of receiving terminal _iInterior subband signal, y are the subband signal of the corresponding x of transmitting terminal transmission; Determine according to this Z _iNext subband Z _I+1The FEQ filter coefficient;

C. from the OFDM symbol, determine at Z _iSeveral subbands in the corresponding frequency band are with Z _iThe FEQ filter coefficient give these several subbands middle subband U1; From the OFDM symbol, determine at Z _I+1Several subbands in the corresponding frequency band are with Z _I+1The FEQ filter coefficient give these several subbands middle subband U2;

D. after determining to obtain the FEQ filter coefficient of subband U1 and subband U2, estimate the FEQ filter coefficient of OFDM symbol from subband U1 to each subband between subband U2 by the difference estimation technique;

E. according to step B～D, traversal is tried to achieve the FEQ filter coefficient of each subband of OFDM symbol.

Images