CN103220252A - Coding orthogonal frequency division multiplexing wireless signal receiving and processing device and processing method thereof - Google Patents

Abstract

The invention discloses a coding orthogonal frequency division multiplexing wireless signal receiving and processing device and a processing method thereof. The device comprises a receiving synchronizer, a channel equalizer and a decoder. The receiving synchronizer comprises a frame detection module, a time domain decimal frequency offset rectifying module, a sign timing rectifying module, a frequency rectifying module, a Fourier transform module, an integer frequency offset estimation module, a pilot frequency extraction module, a frequency domain decimal frequency offset estimation module, a sign timing offset estimation module, an inherent spurious frequency offset tracking module and a sampling frequency offset rectifying module. The receiving synchronizer comprehensively takes the effects of carrier synchronization offset, carrier phase offset, sampling interval offset, sampling timing offset and orthogonal frequency division multiplexing (OFDM) sign timing offset on OFDM signal demodulation into consideration to design a receiving synchronization algorithm from the three aspects of carrier synchronization, sign timing synchronization and sampling synchronization. The coding orthogonal frequency division multiplexing wireless signal receiving and processing device further has the advantages of being high in video definition, small in transmission delay, strong in interference resistance, high in reliability and the like. The invention further discloses the processing method of the device.

Description

Reception of wireless signals processing unit and the processing method thereof of Coded Orthogonal Frequency Division Multiplexing (COFDM)

Technical field

The present invention relates to data processing technique, relate in particular to a kind of reception of wireless signals processing unit and processing method thereof of Coded Orthogonal Frequency Division Multiplexing (COFDM).

Background technology

In recent years, along with GPRS, CDMA, 3G(TD-SCDMA, WCDMA, CDMA2000), the rise of the cordless communication network such as OFDM, wireless video transmission system is with its maneuverability, visual pattern, the advantage such as easy to operate, be widely used in the every field such as army individual-soldier operation, public security remote monitoring evidence obtaining, TV programme relay, the monitoring of mine oilfield safety, the environmental monitoring of underground garage supermarket, cell environment monitoring, and become that people realize at a distance, a kind of important means of on-site supervision on a large scale.The COFDM technology is high with its transmission rate, antijamming capability is strong, and this makes at " in high-speed motion " and " non-looking under the poker part " and realizes that high-quality realtime graphic and transfer of data become possibility.The mobile multimedia transmission system of developing based on the COFDM modulation technique adopts advanced COFDM modulation-demodulation technique, channel coding/decoding technology, and, in conjunction with technology such as digital image compression technology, can under high-speed mobile environment, realize real-time, the synchronous transmission of the broadband multimedia services such as video, voice, data.The remarkable advantage such as have wide coverage, highly sensitive, mobility good, anti-interference and anti-fading ability is strong, transmission data rate is high, stability and reliability are outstanding, for commanding, speedily carry out rescue work, the emergency communications such as scouting, Field Operational provide at a distance, the ideal solution of high-quality, two-forty, wireless real-time transmission.

At present have many companies to carry out research and the sale of wireless video monitoring field product both at home and abroad, from the transmission of video product based on OFDM of external main flow, its transmitting image resolution is at D1(720 * 480p), portioned product is embargoed China; And there are the problems such as core technology (baseband coding, modulation, decoding deck) dependence on import, the application of expensive, product are single mostly in home products, and Introduced From Abroad advanced person's video monitoring system exists expensive, problems such as key technology is under one's control, poor expandability.Therefore design realize high definition wireless video transmission system based on COFDM for breaking international blockade, break-through skill bottleneck, to meet specific industry significant to the demand of HD video.Band receiver of base is based on one of the core component of the high-resolution video transfer system of COFDM, directly determined the system core indexs such as transmission delay, the error rate, anti-interference, unfailing performance, the appropriate design band receiver of base is vital for optimization system performance, reduction implementation complexity.

Summary of the invention

For the existing problems and shortcomings of video wireless transmission product, the purpose of this invention is to provide a kind ofly have that video definition is high, propagation delay time is little, anti-interference strong, the high-definition video wireless transmission system band receiver of base algorithm design structure based on COFDM that reliability is high.The present invention is suitable for the field application such as high definition audio-video monitoring, broadband wireless communications.

The present invention is achieved in that the reception of wireless signals processing unit of Coded Orthogonal Frequency Division Multiplexing (COFDM), and this reception of wireless signals processing unit comprises:

Receive synchronizer, it comprises:

The frame detection module, it is for to receiving signal r _kFirst after the frame Check processing, obtain thick frame synchronizing signal y _k

Time domain decimal frequency bias correction module, it is for to thick frame synchronizing signal y _kAdvanced after the time domain decimal frequency bias is proofreaied and correct and obtained signal y;

The Symbol Timing correction module, it is for to thick frame synchronizing signal y _kCarry out Symbol Timing and proofread and correct the signal obtained through the timing offset correction

Frequency correction module, it is for territory decimal frequency bias correction signal

Obtain signal

${\hat{Y}}_k=Y_k{e}^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}k/N};$

The Fourier transform module, it obtains frequency-region signal Y for the signal y to after the time domain decimal frequency bias is proofreaied and correct through fast Fourier transform;

The integer frequency bias estimation module, it relevant cumulative searched peaked method and realizes that integer frequency bias estimates and compensation for what select the adjacent OFDM symbol continuous pilot, and the estimated value of normalization integer frequency bias meets following formula:

$\hat{m}=\underset{m\&Element;(-C,C)}{\arg \max \left(\right|W\left(m\right)\left|\right)}=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{Y}_i(k+m)Y_{(i+1)}^{*}(k+m)\left|\right)$

$=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{S}_i(k+m)e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;\&epsiv;}{N}_i/N}{S}_{i+1}^{*}(k+m){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">e</figure-callout>}}^j\left|\right),$ In formula, the subcarrier sequence number collection that P is continuous pilot place in the OFDM symbol, (C, C) is the scope of displacement search;

Pilot extraction module, it is for obtaining the pilot frequency information R of OFDM symbol _Sp

Frequency domain decimal frequency bias estimation module, it is for estimating remaining decimal frequency bias;

The symbol time offset estimation module, k in its same OFDM symbol received for employing ₁, k ₂Locational pilot tone y _K1, y _K2Cross-correlation, asked the phase angle computing and be averaging cross correlation value, obtains timing error and estimate

$\widehat{\mathrm{\&sigma;}}=\underset{l=1}{\overset{1-\mathrm{NP}/2}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_l=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{\mathrm{N\&Delta;}{\mathrm{\&phi;}}_{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">l</figure-callout>}}}{2\mathrm{\&pi;\&Delta;k}}=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{N(\&angle;[y_{k1}{\mathrm{<figure-callout\; id="2"\; label="y\; k"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">y</figure-callout>}}_k^2\left]\right)}{2\mathrm{\&pi;\&Delta;k}},$ Δ k=k wherein ₂-k ₁, NP is continuous pilot number of sub carrier wave in a symbol, the carrier number that N is each OFDM symbol, σ _lFor the estimated value of utilizing l to use this algorithm to obtain to pilot tone;

The inherent spurious frequency deviation tracking module, it is for utilizing the continuous pilot subcarrier to carry out estimation and the compensation of excess phase; And

The sampling frequency offset correction module, it,, for utilizing pilot signal sampling frequency offset is estimated and proofreaied and correct at frequency domain, at first estimates the caused phase rotating of sampling frequency offset, and then accordingly each sampled value is compensated;

Channel equalizer, it is for adopting the channel estimation and equalization algorithm based on scattered pilot to realize channel estimating and channel equalization; And

Decoder, it is for carrying out the complex data symbol after chnnel coding obtains normalization according to input code flow.

As the further improvement of such scheme, in the frame detection module, Frame Detection Algorithm is: detection statistic M (n) can be expressed as

$M\left(n\right)=\frac{{\left|c\left(n\right)\right|}^2}{{\left|p\left(n\right)\right|}^2}=\frac{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(k+n)r^{*}(n+k+N)\right|}^2}{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(n+k)\right|}^2},$ If it is output signal y that M (n)>TH detects frame start position _kOtherwise, do not detect, continue to detect; Wherein, TH is the judgement threshold, and c (n) is for receiving signal and the cross correlation value that receives signal lag, and p (n) means the energy of corresponding reception signal, and N means the carrier number of OFDM symbol.

As the further improvement of such scheme, in frequency domain decimal frequency bias estimation module, for the continuous pilot in the OFDM symbol, when k ∈ P, meet S _I+1(k)=S _i(k); If only there is decimal frequency bias in the reception signal, draw thus the estimated value of decimal frequency bias Wherein, what N was FFT counts, N _GFor the carrier number at protection interval, ∠ is the tangent of negating.

Preferably, in order to increase the resistivity to noise, the accuracy of improve calculating, average the estimated value of a plurality of pilot tones of OFDM symbol inside, obtains corresponding decimal frequency bias:

As the further improvement of such scheme, in the inherent spurious frequency deviation tracking module,

The excess phase estimation

Meet following formula:

${\widehat{\mathrm{\&phi;}}}_m=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{\hat{R}}_{m,n}{\left(P_{m,n}\right)}^{*}\right]$

$=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{P}_{m,\mathrm{<figure-callout\; id="2"\; label="n\; e\; j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">n</figure-callout>}}{e}^j{2\mathrm{\&pi;m\&Delta;f}}^{}{\left(P_{m,n}\right)}^{*}\right]$

Wherein,

N pilot tone of m the OFDM symbol that expression receives, P _m,nFor the pilot tone of transmitting terminal, i.e. local known pilot, P _m,nFor+4/3 or-4/3;

Calculate compensating factor Meet following formula:

$e^{-j{\mathrm{\&phi;}}_m}=e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}$

$=\frac{1}{64}\&times;e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}\&times;\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})$

$=\frac{1}{64}\&times;{\left[\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})\right]}^{*};$

Finally, according to following formula to receiving data

Carry out phase compensation:

As the further improvement of such scheme, in the sampling frequency offset correction module,

Sampling frequency offset estimated value s _lBe shown below:

Wherein, P _kPilot sub-carrier, wherein k _jFor the subcarrier label, j=0,1,2 ..., M;

The sampling frequency offset compensation

${\hat{R}}_k={\hat{Y}}_k{e}^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}{s}_{l}k/N}.$

The present invention also provides a kind of reception of wireless signals processing method of Coded Orthogonal Frequency Division Multiplexing (COFDM), and this reception of wireless signals processing method comprises the following steps:

To receiving signal r _kFirst after the frame Check processing, obtain thick frame synchronizing signal y _k

To thick frame synchronizing signal y _kAdvanced after the time domain decimal frequency bias is proofreaied and correct and obtained signal y;

To thick frame synchronizing signal y _kCarry out Symbol Timing and proofread and correct the signal obtained through the timing offset correction

${\hat{y}}_k=y_k{e}^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;k}\widehat{\mathrm{\&sigma;}}};$

Territory decimal frequency bias correction signal

Obtain signal

Signal y after the time domain decimal frequency bias is proofreaied and correct is obtained to frequency-region signal Y through fast Fourier transform;

That selects continuous pilot in adjacent OFDM symbol relevant cumulatively searches peaked method and realizes that integer frequency bias estimates and compensation, and the estimated value of normalization integer frequency bias meets following formula:

$\hat{m}=\underset{m\&Element;(-C,C)}{\arg \max \left(\right|W\left(m\right)\left|\right)}=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{Y}_i(k+m)Y_{(i+1)}^{*}(k+m)\left|\right)$

$=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{S}_i(k+m)e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;\&epsiv;}{N}_i/N}{S}_{i+1}^{*}(k+m){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">e</figure-callout>}}^j\left|\right),$ In formula, the subcarrier sequence number collection that P is continuous pilot place in the OFDM symbol, (C, C) is the scope of displacement search;

Obtain the pilot frequency information R in the OFDM symbol _Sp

Estimate remaining decimal frequency bias;

K in the same OFDM symbol that employing receives ₁, k ₂Locational pilot tone y _K1, y _K2Cross-correlation, asked the phase angle computing and be averaging cross correlation value, obtains timing error and estimate

, $\widehat{\mathrm{\&sigma;}}=\underset{l=1}{\overset{1-\mathrm{NP}/2}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_l=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{\mathrm{N\&Delta;}{\mathrm{\&phi;}}_{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">l</figure-callout>}}}{2\mathrm{\&pi;\&Delta;k}}=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{N(\&angle;[y_{k1}{\mathrm{<figure-callout\; id="2"\; label="y\; k"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">y</figure-callout>}}_k^2\left]\right)}{2\mathrm{\&pi;\&Delta;k}},$ , Δ k=k wherein ₂-k ₁, NP is continuous pilot number of sub carrier wave in a symbol, the carrier number that N is each OFDM symbol, σ _lFor the estimated value of utilizing l to use this algorithm to obtain to pilot tone;

Utilize the continuous pilot subcarrier to carry out estimation and the compensation of excess phase;

Utilize pilot signal sampling frequency offset is estimated and proofreaied and correct at frequency domain, at first estimate the caused phase rotating of sampling frequency offset, and then accordingly each sampled value is compensated;

The channel estimation and equalization algorithm of employing based on scattered pilot realized channel estimating and channel equalization; And

Carry out the complex data symbol after chnnel coding obtains normalization according to input code flow.

As the further improvement of such scheme, in the frame detecting step, Frame Detection Algorithm is: detection statistic M (n) can be expressed as

$M\left(n\right)=\frac{{\left|c\left(n\right)\right|}^2}{{\left|p\left(n\right)\right|}^2}=\frac{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(k+n)r^{*}(n+k+N)\right|}^2}{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(n+k)\right|}^2},$ If it is output signal y that M (n)>TH detects frame start position _kOtherwise, do not detect, continue to detect; Wherein, TH is the judgement threshold, and c (n) is for receiving signal and the cross correlation value that receives signal lag, and p (n) means the energy of corresponding reception signal, and N means the carrier number of OFDM symbol.

As the further improvement of such scheme, in frequency domain decimal frequency bias estimating step, for the continuous pilot in the OFDM symbol, when k ∈ P, meet S _I+1(k)=S _i(k); If only there is decimal frequency bias in the reception signal, draw thus the estimated value of decimal frequency bias

, wherein, what N was FFT counts, N _GFor the carrier number at protection interval, ∠ is the tangent of negating.

Preferably, in order to increase the resistivity to noise, the accuracy of improve calculating, average the estimated value of a plurality of pilot tones of OFDM symbol inside, obtains corresponding decimal frequency bias:

As the further improvement of such scheme, in the inherent spurious frequency deviation tracking step, the excess phase estimation

Meet following formula:

${\widehat{\mathrm{\&phi;}}}_m=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{\hat{R}}_{m,n}{\left(P_{m,n}\right)}^{*}\right]$

$=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{P}_{m,n}{e}^{\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}{\left(P_{m,n}\right)}^{*}\right]$

Wherein,

N pilot tone of m the OFDM symbol that expression receives, P _m,nFor the pilot tone of transmitting terminal, i.e. local known pilot, P _m,nFor+4/3 or-4/3;

Calculate compensating factor

Meet following formula:

$e^{-j{\mathrm{\&phi;}}_m}=e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}$

$=\frac{1}{64}\&times;e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}\&times;\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})$

$=\frac{1}{64}\&times;{\left[\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})\right]}^{*};$

Finally, according to following formula to receiving data

Carry out phase compensation:

.

As the further improvement of such scheme, in the sampling frequency offset aligning step, sampling frequency offset estimated value s _lBe shown below:

, wherein, P _kPilot sub-carrier, wherein k _jFor the subcarrier label, j=0,1,2 ..., M;

The sampling frequency offset compensation

, ${\hat{R}}_k={\hat{Y}}_k{e}^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}{s}_{l}k/N}.$

Coded Orthogonal Frequency Division Multiplexing (COFDM) (Code Orthogonal Frequency Division Multiplexing, COFDM) be a kind of novel data processing technique, rely on its high availability of frequency spectrum, high transmission rate and strong anti-multipath interference performance, arouse widespread concern in the wireless image transmission field.COFDM produces and develops on the basis of OFDM, it adopts data encoding technique, solve well Digital Image Transmission route multipath reflection and transmission route barrier and blocked brought puzzlement, no matter under the complicated special non line of sight transmission environment such as indoor, outdoor, ground, down-hole, can both be on-the-spot real time information, be sent in real time destination with high-quality data image signal, really realized " resistance gear ", " non line of sight " and " communication in moving " of image transmitting.The present invention is based on the important component part of the high-resolution video transfer system of COFDM, is applicable to foundation and the enforcement of the high-resolution video transfer system based on COFDM.

The accompanying drawing explanation

The project organization block diagram of the reception of wireless signals processing unit of the Coded Orthogonal Frequency Division Multiplexing (COFDM) that Fig. 1 provides for better embodiment of the present invention.

The frame that Fig. 2 is the frame detection module in Fig. 1 detects the implementation structure block diagram.

The integer frequency offset estimation algorithm implementation structure block diagram that Fig. 3 is the integer frequency bias estimation module in Fig. 1.

The hardware implementation structure block diagram of the excess phase tracking module that Fig. 4 is the inherent spurious frequency deviation tracking module in Fig. 1.

The sample frequency synchronization hardware implementation structure block diagram that Fig. 5 is the sampling frequency offset correction module in Fig. 1.

The channel estimation model based on pilot tone or training symbol that Fig. 6 is channel equalizer 2 in Fig. 1.

The band receiver of base implementation structure block diagram of the reception of wireless signals processing unit that Fig. 7 is Coded Orthogonal Frequency Division Multiplexing (COFDM) in application drawing 1.

The transmitting-receiving baseband system Modelsim analogous diagram that Fig. 8 is band receiver of base in Fig. 7.

Fig. 9 is band receiver of base FPGA occupation condition in Fig. 7.

Embodiment

In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.

As shown in Figure 1, the reception of wireless signals processing unit of the Coded Orthogonal Frequency Division Multiplexing (COFDM) that better embodiment of the present invention provides is three large parts of band receiver of base, comprise and receive synchronizer 1, channel equalizer 2, decoder 3, wherein receive synchronizer 1 and mainly contain frame detection module 11, time domain decimal frequency bias correction module 12, Symbol Timing correction module 13, frequency correction module 14, Fourier transform module 15, integer frequency bias estimation module 16, pilot extraction module 17, frequency domain decimal frequency bias estimation module 18, symbol time offset estimation module 19, inherent spurious frequency deviation tracking module 110, sampling frequency offset correction module 111.Channel equalizer 2 comprises channel estimation module 21, channel equalization module 22; Decoder 3 mainly is comprised of modules such as frame decomposer, constellation demapping, solution interleaver, interior coding decoder, solution external interleaver, outer coding decoder, descrambler.

Step (one) receives synchronously and processes

Orthogonality between subcarrier is vital for the COFDM system, be the basis that guarantees accurately reception and demodulation, so Synchronization Design just seems particularly important for the COFDM system.

Considered carrier synchronization deviation, carrier phase deviation, sampling interval deviation, sample timing error and the impact that demodulation brings on ofdm signal of OFDM symbol time offset, be shown below:

$R_i\left(k\right)=\underset{m=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{S}_i\left(m\right){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">e</figure-callout>}}^j{e}^{-\mathrm{j\&pi;}(N-1)[(m+\mathrm{m\&delta;}-\mathrm{\&epsiv;}-k]/N-\mathrm{j\&phi;}}\frac{\sin c[(m-\mathrm{\&epsiv;})(1+\mathrm{\&delta;})-k]}{\sin c\left[\frac{(m-\mathrm{\&epsiv;})(1+\mathrm{\&delta;})-k}{N}\right]}+{\mathrm{\&eta;}}^{\&prime;\&prime;}$

$S_i\left(k\right)={\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">e</figure-callout>}}^j{e}^{\mathrm{j\&pi;}(N-1)(\mathrm{k\&delta;}-\mathrm{\&epsiv;})/N-\mathrm{j\&Delta;\&phi;}}\frac{\sin c[(m-\mathrm{\&epsiv;})(1+\mathrm{\&delta;})-k]}{\sin c\left[\frac{(m-\mathrm{\&epsiv;})(1+\mathrm{\&delta;})-k}{N}\right]}$

$+\underset{m\&NotEqual;k}{\overset{N-1}{\underset{m=0}{\mathrm{\&Sigma;}}}}{S}_i\left(m\right){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">e</figure-callout>}}^j{e}^{-\mathrm{j\&pi;}(N-1)[(m+\mathrm{m\&delta;}-\mathrm{\&epsiv;}-k]/N-\mathrm{j\&phi;}}\frac{\sin c[(m-\mathrm{\&epsiv;})(1+\mathrm{\&delta;})-k]}{\sin c\left[\frac{(m-\mathrm{\&epsiv;})(1+\mathrm{\&delta;})-k}{N}\right]}+{\mathrm{\&eta;}}^{\&prime;\&prime;}$

$=S_i\left(k\right)I_{k,k}+\underset{m\&NotEqual;k}{\overset{N-1}{\underset{m=0}{\mathrm{\&Sigma;}}}}{S}_i\left(m\right)+I_{k,m}+{\mathrm{\&eta;}}^{\&prime;\&prime;},k=\mathrm{0,1},...,N-1---(2-1)$

Δ f wherein _cFor the receiving terminal carrier frequency offset, Δ φ is the carrier phase deviation, Δ T _sFor the sampling interval deviation, Δ t _SpFor sample timing error, Δ t _SbFor the OFDM symbol time offset, Δ T _sFor the sampling interval deviation.Can be found out carrier phase error Δ φ, sample-timing error Δ t by (2-1) formula _SpWith OFDM symbol timing error Δ t _SbOnly produce to received signal phase rotating in various degree; And carrier frequency error Δ f _cWith sampling interval error Δ T _sCan destroy the orthogonality between subcarrier in the OFDM symbol, produce ICI, the loss of bringing signal to noise ratio to the OFDM restituted signal.Design the reception synchronized algorithm from carrier synchronization, timing synchronization and three aspects of sample-synchronous.

There are the deviations such as carrier synchronization deviation, carrier phase deviation, sampling interval deviation, sample timing error and OFDM symbol time offset in the signal of receiver after the ADC sampling, and carrier phase error, sample-timing error and OFDM symbol timing error only produce phase rotating in various degree to received signal; Carrier frequency error and sampling interval error can destroy the orthogonality between subcarrier in the OFDM symbol, produce ICI, the loss of bringing signal to noise ratio to the OFDM restituted signal.The synchronous processing of reception that receives synchronizer 1 is exactly in order to compensate or eliminate above-mentioned error effect.Concrete steps are as follows:

Step 1. frame detects

In this device, it is to depend on hard-wiredly that frame detects, although existing some algorithm can be obtained performance preferably, computational complexity is large, is unfavorable for realizing.Therefore should take into full account computational complexity during the conducting frame detection design when considering algorithm performance, select economic algorithm; When conducting frame detects simultaneously, not yet carry out Frequency Synchronization, may there be larger frequency departure in the complex baseband signal received, therefore needs to select a kind of frequency departure that is subject to affect less algorithm.In view of top analysis, this is device Z _cDetection statistic is adjudicated, below to Z _cSimply change.

Order

$c\left(n\right)=\underset{k=M}{\overset{L-1}{\mathrm{\&Sigma;}}}r(k+n)r^{*}(k+M+n)---(2-2)$

$p\left(n\right)=\underset{k=M}{\overset{L-1}{\mathrm{\&Sigma;}}}{\left|r(n+k)\right|}^2---(2-3)$

For the ease of calculating, to above-mentioned two formula deliverys square, detection statistic Z _cCan be expressed as:

$M\left(n\right)=\frac{{\left|c\left(n\right)\right|}^2}{{\left|p\left(n\right)\right|}^2}---(2-4)$

Also with regard to book, say, frame detection module 11 receives signal r _kFirst after the frame Check processing, obtain thick frame synchronizing signal y _k, Frame Detection Algorithm is as follows:

Detection statistic M (n) can be expressed as:

$M\left(n\right)=\frac{{\left|c\left(n\right)\right|}^2}{{\left|p\left(n\right)\right|}^2}=\frac{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(k+n)r^{*}(n+k+N)\right|}^2}{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(n+k)\right|}^2}---(2-5)$

If it is output signal y that M (n)>TH detects frame start position _kOtherwise, do not detect, continue to detect.

Wherein, TH is the judgement threshold, and c (n) is for receiving signal and the cross correlation value that receives signal lag, and p (n) means the energy of corresponding reception signal, and N means the carrier number of OFDM symbol, obtains thus frame detection architecture block diagram as shown in Figure 2.

Step 2. time domain decimal carrier wave frequency deviation is proofreaied and correct

The signal y proofreaied and correct through timing offset _kCause receiving even losing lock of distorted signals owing to affected by decimal frequency bias, therefore need 12 pairs of decimal frequency bias of time domain decimal frequency bias correction module estimate and proofread and correct, in order to estimate more accurately and to proofread and correct decimal frequency bias, the decimal frequency bias backoff algorithm that time domain decimal frequency bias correction module 12 of the present invention adopts time-domain and frequency-domain to combine, time domain decimal frequency bias correcting algorithm and frequency domain decimal frequency bias backoff algorithm are similar, after the time domain decimal frequency bias is proofreaied and correct, obtain signal y.

Step 3. Symbol Timing is proofreaied and correct

Obtain through Symbol Timing correction module 13 signal that timing offset is proofreaied and correct

, can be expressed as:

${\hat{y}}_k=y_k{e}^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;k}\widehat{\mathrm{\&sigma;}}}---(2-6).$

Step 4. frequency domain decimal frequency bias is estimated

The signal obtained after the frequency domain decimal frequency bias is proofreaied and correct through frequency correction module 14 is:

${\hat{Y}}_k=Y_k{e}^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}k/N}---(2-7)$

The step 5. signal y after the time domain decimal frequency bias is proofreaied and correct obtains frequency-region signal Y through the fast Fourier transform (FFT) of Fourier transform module 15;

Step 6. integer frequency bias is estimated

As shown in Figure 3, integer frequency bias estimation module 16 of the present invention is selected continuous pilot in adjacent OFDM symbol relevant cumulatively searches peaked method and realizes that integer frequency bias estimates and compensation, and algorithm is as follows:

The estimated value of normalization integer frequency bias can be obtained by following formula:

$\hat{m}=\underset{m\&Element;(-C,C)}{\arg \max \left(\right|W\left(m\right)\left|\right)}=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{Y}_i(k+m)Y_{(i+1)}^{*}(k+m)\left|\right)$

$=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{S}_i(k+m)e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;\&epsiv;}{N}_i/N}{S}_{i+1}^{*}(k+m){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">e</figure-callout>}}^j\left|\right)---(2-8)$

In formula, the subcarrier sequence number collection that P is continuous pilot place in the OFDM symbol, (C, C) is the scope of displacement search.

Step 7. pilot frequency information is extracted

After above-mentioned steps is processed, signal carries out the pilot frequency information extraction of pilot extraction module 17, obtains the pilot frequency information R in the OFDM symbol _Sp.

Step 8. frequency domain decimal frequency bias is estimated

Frequency domain decimal frequency bias estimation module 18 frequency domains adopt the continuous pilot in the two adjacent OFDM symbol to estimate remaining decimal frequency bias (phase error) with higher precision in more among a small circle, and specific algorithm is as follows:

For the continuous pilot in the OFDM symbol, when k ∈ P, meet S _I+1(k)=S _i(k).If only there is decimal frequency bias in the reception signal, draw thus the estimated value of decimal frequency bias

${\widehat{\mathrm{\&epsiv;}}}_i==\frac{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">N</figure-callout>}}{2\mathrm{\&pi;}(N+N_G)}\&angle;\left[Y_i\left(k\right)Y_{i+1}^{*}\left(k\right)\right]---(2-9)$

What wherein N was FFT counts, N _GFor the carrier number at protection interval, ∠ is the tangent of negating.

In order to increase the resistivity to noise, the accuracy of improve calculating, can average the estimated value of a plurality of pilot tones of OFDM symbol inside, obtains corresponding decimal frequency bias:

$\widehat{\mathrm{\&epsiv;}}=\frac{1}{N_p}\underset{i=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{\widehat{\mathrm{\&epsiv;}}}_i---(2-10)$

Step 9. symbol time offset is estimated

The given FFT the window's position of Coarse Symbol Synchronization Algorithms guarantees the uncontaminated zone at Cyclic Prefix, is not generally the accurate starting position of actual OFDM symbol, but leading several data acquisition sampling point.So, be necessary to carry out thin timing synchronization at frequency domain after the FFT demodulation.Because the timing synchronization error of time domain can cause different sub carrier at frequency domain with different phase rotatings, and the size of rotation rotation is directly proportional to sub-carrier indices to the Timing Synchronization error.

The present invention adopts pilot signal to estimate the rotation due to signal phase that timing offset causes, according to estimated result, at frequency domain, signal is compensated, and specific algorithm is as follows:

If system is through after thick sign synchronization, remaining normalized symbol timing error is Δ n, and supposes that normalized sample timing error is Δ t, and the pilot tone frequency domain restituted signal that is subject to symbol timing error and sample timing error impact can be expressed as:

Y _k=X _ke ^{j2πk(Δn+Δt)/N} （2-11）

K in same OFDM symbol ₁And k ₂The phase difference of pilot sub-carrier be:

Δφ=2πΔk(Δn+Δt)/N=2Δkσ/N （2-12）

Δ k=k wherein ₂-k ₁, σ=Δ n+ Δ t.

Can find out k in same OFDM symbol from formula (2-12) ₁With k ₂The phase difference of pilot sub-carrier except relevant with pilot sub-carrier index interval, only relevant with sample-timing error sum σ with the timing synchronization error, and be the linear function of σ, and can not show that the estimated value of σ is:

$\widehat{\mathrm{\&sigma;}}=\frac{\mathrm{N\&Delta;\&phi;}}{2\mathrm{\&pi;\&Delta;k}}---(2-13)$

In order to improve anti-white Gaussian noise ability, improve arithmetic accuracy, can calculate and be averaging respectively multipair pilot tone, have:

$\widehat{\mathrm{\&sigma;}}=\underset{l=1}{\overset{l-\mathrm{NP}/2}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_l---(2-14)$

σ wherein _lFor the estimated value of utilizing l to use this algorithm to obtain to pilot tone.Will

Be decomposed into integer part and fractional part, the estimated value that can obtain symbol timing error and sample timing error is respectively:

$\widehat{\mathrm{\&Delta;n}}=\mathrm{integer}\left(\widehat{\mathrm{\&sigma;}}\right)---(2-15)$

$\widehat{\mathrm{\&Delta;t}}=\mathrm{fraction}\left(\widehat{\mathrm{\&sigma;}}\right)---(2-16)$

That is to say, symbol time offset estimation module 19 of the present invention adopts k in the same OFDM symbol received ₁, k ₂Locational pilot tone y _K1, y _K2Cross-correlation, asked the phase angle computing and be averaging cross correlation value, obtains timing error and estimate, is shown below:

$\widehat{\mathrm{\&sigma;}}=\underset{l=1}{\overset{l-\mathrm{NP}/2}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_l=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{\mathrm{N\&Delta;}{\mathrm{\&phi;}}_{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">l</figure-callout>}}}{2\mathrm{\&pi;\&Delta;k}}=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{N(\&angle;[y_{k1}{\mathrm{<figure-callout\; id="2"\; label="y\; k"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">y</figure-callout>}}_k^2\left]\right)}{2\mathrm{\&pi;\&Delta;k}}---(2-17)$

Δ k=k wherein ₂-k ₁, NP is continuous pilot number of sub carrier wave in a symbol, the carrier number that N is each OFDM symbol, σ _lFor the estimated value of utilizing l to use this algorithm to obtain to pilot tone.

Step 10. excess phase is followed the tracks of

Although the ofdm system receiver has Carrier Synchronization Algorithm, but this frequence estimation is not perfect in every way, data after carrier wave frequency deviation is proofreaied and correct still leave certain remaining deviation, and these remaining deviations will cause the phase deviation of data, and it shows as the rotation of planisphere.

The algorithm that excess phase is followed the tracks of mainly contains data auxiliary law and the large class of non-data auxiliary law two.In native system, each symbol has comprised 45 continuous pilot subcarriers, adopts data-aided excess phase tracking in design.

If R _m,nN pilot tone of m the OFDM symbol that expression receives, P _m,nFor the pilot tone of transmitting terminal, i.e. local known pilot.After m receiving symbol DFT processes, pilot sub-carrier R _m,nEqual domain channel response H _nWith known pilot symbols P _m,nProduct and the later result of residual frequency error turning effort:

R _m,n=H _nP _m,ne ^j2πm·Δf （2-18）

Wherein, Δ f is frequency departure remaining after carrier wave frequency deviation is proofreaied and correct.

Suppose the domain channel response estimated value

Be available, and be accurately reliably, can ignore the impact of domain channel response,

R _m,n=P _m,ne ^j2πm·Δf （2-19）

Phase estimation is:

${\widehat{\mathrm{\&phi;}}}_m=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{R}_{m,n}{\left(P_{m,n}\right)}^{*}\right]$

$\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{P}_{m,n}{e}^{\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}{\left(P_{m,n}\right)}^{*}\right]---(2-20)$

For DVB-T system, P _m,nFor+4/3 or-4/3, so above formula can be reduced to

${\widehat{\mathrm{\&phi;}}}_m=\&angle;\left[e^{j2m\&CenterDot;\mathrm{\&Delta;f}}\underset{1}{\overset{45}{\mathrm{\&Sigma;}}}(4/3)\right]$

$=\&angle;[64\&times;{\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">e</figure-callout>}}^j]=2\mathrm{\&pi;m}\&CenterDot;\mathrm{\&Delta;f}---(2-21)$

Note, in reality, channel estimating is not strictly accurately, and therefore estimation is also affected.

In the Phase Tracking stage, compensating factor should be:

$e^{-j\mathrm{\&phi;}}=e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}$

$=\frac{1}{64}\&times;e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}\&times;\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}(R_{m,n}\&times;P_{m,n})$

$=\frac{1}{64}\&times;{\left[\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}(R_{m,n}\&times;P_{m,n})\right]}^{*}---(2-22)$

Finally, carry out phase compensation according to following formula to receiving data:

${\hat{r}}_m=r_m\&times;e^{-\mathrm{j\&phi;}}---(2-23)$

That is to say, the signal after above-mentioned treatment step still leaves certain remaining deviation, and these remaining deviations will cause the phase deviation of data, and it shows as the rotation of planisphere.As shown in Figure 4, inherent spurious frequency deviation tracking module 110 of the present invention utilizes the continuous pilot subcarrier to carry out estimation and the compensation of excess phase.

(1) excess phase estimation

${\widehat{\mathrm{\&phi;}}}_m=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{\hat{R}}_{m,n}{\left(P_{m,n}\right)}^{*}\right]$

$\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{P}_{m,n}{e}^{\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}{\left(P_{m,n}\right)}^{*}\right]$

$=\&angle;\left[e^{j2m\&CenterDot;\mathrm{\&Delta;f}}\underset{1}{\overset{45}{\mathrm{\&Sigma;}}}(4/3)\right]=2\mathrm{\&pi;m}\&CenterDot;\mathrm{\&Delta;f}---(2-24)$

Wherein,

N pilot tone of m the OFDM symbol that expression receives, P _m,nFor the pilot tone of transmitting terminal, i.e. local known pilot, P _m,nFor+4/3 or-4/3;

(2) calculate compensating factor

$e^{-j{\mathrm{\&phi;}}_m}=e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}$

$=\frac{1}{64}\&times;e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;m\&Delta;f}}\&times;\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})$

$=\frac{1}{64}\&times;{\left[\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})\right]}^{*}---(2-25)$

Finally, carry out phase compensation according to following formula to receiving data:

$R_m={\hat{R}}_m\&times;e^{-\mathrm{j\&phi;}}---(2-26)$

Step (11) sampling frequency offset is proofreaied and correct

The method of revising sampling frequency deviation in ofdm system mainly contains two kinds, and a kind of is to carry out in time domain, by processed to realize the synchronous of sample frequency before the fast Fourier transform of receiver; Another kind is through after FFT, in frequency, is proofreaied and correct.For the former usually the most direct method be to adjust the ADC sampling clock, but this needs ADC to adopt voltage controlled oscillator, will certainly increase the cost of receiver like this, therefore, comparatively reasonable approach is to adopt interpolation method, be that clock is still produced by fixing crystal oscillator, remove or insert a sample value when sampling error is added to a sampling clock from data sample.The most frequently used for the method for being estimated at frequency domain and compensate is to utilize pilot signal, at first estimates the caused phase rotating of sampling frequency offset, and then accordingly each sampled value is compensated.

Suppose that pilot sub-carrier is

, k wherein _jFor the subcarrier label, j=0,1,2 ..., M, when only considering the affecting of signal and sampling frequency offset, the pilot signal received can be expressed as

$Y_{{l,k}_j}=H_{k_j}{P}_{k_j}{e}^{{\mathrm{\&theta;}}_{{l,k}_j}}---(2-27)$

Wherein,

Mean the frequency signal response of this pilot signal, carry out to received signal channel estimating, obtain the channel estimation value of each subcarrier

, then the pilot signal received is carried out to channel equalization, then samples with known pilot tone Carrying out related operation can obtain:

$Y_{{l,k}_j}\left({{\hat{H}}_{k_j}}^{*}{P_{k_j}}^{*}\right)=H_{k_j}{P}_{k_j}{e}^{{\mathrm{\&theta;}}_{{l,k}_k}}\left({{\hat{H}}_{k_j}}^{*}{P_{k_j}}^{*}\right)={\left|H_{k_j}\right|}^2{\left|P_{k_j}\right|}^2{e}^{{\mathrm{\&theta;}}_{{l,k}_j}}---(2-28)$

Above formula is carried out to phase estimation and just can obtain phase rotating information

,

${\mathrm{\&theta;}}_{{l,k}_j}=s_l{k}_j---(2-29)$

The present invention has provided a kind of known each pilot tone

Ask s _lMethod, because the phase rotating of each pilot signal all meets above formula, they have formed about s together _lSystem of linear equations, with vector, can be expressed as:

θ=ks _l （2-30）

Wherein, $\mathrm{\&theta;}={({\mathrm{\&theta;}}_{{l,k}_1},{\mathrm{\&theta;}}_{{l,\mathrm{<figure-callout\; id="2"\; label="k"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">k</figure-callout>}}_2},...,{\mathrm{\&theta;}}_{{l,k}_M})}^T,k={(k_1,{\mathrm{<figure-callout\; id="2"\; label="k"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">k</figure-callout>}}_2,...,k_M)}^T,$ This is the overdetermination system of linear equations of an equation number more than the unknown number number, and its optimal solution can obtain by least square (LS) method:

$s_l={\left(k^{T}k\right)}^{-1}{k}^T\mathrm{\&theta;}=\frac{\underset{j=1}{\overset{M}{\mathrm{\&Sigma;}}}{k}_j{\mathrm{\&theta;}}_{{l,k}_j}}{\underset{j=1}{\overset{M}{\mathrm{\&Sigma;}}}{k}_j^2}---(2-31)$

Above formula is further simplified, and the method usually directly be averaging is asked for s _l,

$s_l=\frac{\underset{j=1}{\overset{M}{\mathrm{\&Sigma;}}}{\mathrm{\&theta;}}_{{l,k}_j}}{\underset{j=1}{\overset{M}{\mathrm{\&Sigma;}}}{k}_j}---(2-32)$

That is to say, as shown in Figure 5, sampling frequency offset correction module 111 of the present invention utilizes pilot signal sampling frequency offset is estimated and proofreaied and correct at frequency domain, at first estimates the caused phase rotating of sampling frequency offset, and then accordingly each sampled value is compensated.

(1) the sampling frequency offset estimated value is shown below:

$s_l=\frac{\underset{j=1}{\overset{M}{\mathrm{\&Sigma;}}}{\mathrm{\&theta;}}_{k_j}}{\underset{j=1}{\overset{M}{\mathrm{\&Sigma;}}}{k}_j}=\frac{\underset{j=1}{\overset{M}{\mathrm{\&Sigma;}}}\&angle;\left[P_k{{\hat{Y}}_k}^{*}\right]}{\underset{j=1}{\overset{M}{\mathrm{\&Sigma;}}}{k}_j}---(2-33)$

Wherein, P _kPilot sub-carrier, wherein k _jFor the subcarrier label, j=0,1,2 ..., M.

(2) sampling frequency offset compensation

${\hat{R}}_k={\hat{Y}}_k{e}^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003032314500011.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}{s}_{l}k/N}---(2-34)$

Step is channel equalizer 2 (ii)

The channel estimation and equalization algorithm of the present invention's sampling based on scattered pilot, the algorithm structure block diagram as shown in Figure 6, mainly comprises several modules such as pilot extraction, channel rough estimate, interpolation.

Step is channel estimating 1.

The channel estimating of the scattered pilot point of channel estimation module 21

$\widehat{H_{\mathrm{sp}}}=R_{\mathrm{sp}}/L_{\mathrm{sp}}---(2-35)$

Because local scattered pilot is L _SpReal number value for ± 4/3,

Be easy to obtain, utilize linear interpolation to obtain the frequency response of whole channel

.

Step is channel equalization 2.

22 pairs of data that receive of channel equalization module compensate and can obtain:

$\hat{R}=R/\hat{H}=(R\&times;{\hat{H}}^{*})/{\left|\hat{H}\right|}^2---(2-36)$

For avoiding division arithmetic, will

Send into the constellation demodulation module, only need to calculate

Get final product.

Step is decoder 3 (iii)

Signal after receiving synchronizer, channel equalizer processing

Successively through the frame decomposer 31 of decoder 3, constellation demapping 32, separate interleaver 33, interior coding decoder 34, export base-band information after separating external interleaver 35, outer coding decoder 36, descrambler 37.Decoder 3 belongs to general design, in the present embodiment, no longer describes in detail.

When hardware is realized, significant advantage based on fpga chip at aspects such as processing speed, logical resource, flexible configuration, cost power consumptions, the present invention adopts take the hardware implementation scheme that FPGA is main body, peripheral circuit comprises received RF circuit, ADC sampling, DSP(DaVinci DM6467) subsequent treatment etc., the specific implementation structure is as shown in Figure 7.For further reasonability and the function realizability of verifying the band receiver of base design of the reception of wireless signals processing unit with Coded Orthogonal Frequency Division Multiplexing (COFDM) of the present invention, algorithm to above-mentioned introduction has carried out semi-physical simulation, Fig. 8 has provided band receiver of base Modelsim simulation result, the Counter Value that wherein in figure, the input data are-127～127, pass through baseband modulation, demodulation, channel, receive synchronous and balanced after, demodulation, reception data after decoding etc. are processed are as shown in " dout ", output " dout " is still-127～127 Counter Value, thereby reasonability and the accuracy of band receiver of base algorithm design have been verified.Fig. 9 is that the band receiver of base algorithm takies the fpga chip resource situation, and as can be seen from Figure, designed band receiver of base takies less FPGA resource, and selected chip can meet the requirement of band receiver of base resource requirement fully.

In sum, technical problem solved by the invention mainly comprises: fast signal detection technique, Carrier Synchronization, sane sample-synchronous technology, problems such as channel estimation and equalization technology efficiently accurately.The present invention is suitable for the application of the fields such as high definition audio-video monitoring, broadband wireless communications, has that video definition is high, propagation delay time is little, anti-interference strong, high reliability.

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.

Claims (12)

1. the reception of wireless signals processing unit of Coded Orthogonal Frequency Division Multiplexing (COFDM), is characterized in that, this reception of wireless signals processing unit comprises:

Receive synchronizer, it comprises:

The frame detection module, it is for to receiving signal r _kFirst after the frame Check processing, obtain thick frame synchronizing signal y _k

Time domain decimal frequency bias correction module, it is for to thick frame synchronizing signal y _kAdvanced after the time domain decimal frequency bias is proofreaied and correct and obtained signal y;

The Symbol Timing correction module, it is for to thick frame synchronizing signal y _kCarry out Symbol Timing and proofread and correct the signal obtained through the timing offset correction

Frequency correction module, it is for territory decimal frequency bias correction signal

, obtain signal

: ${\hat{Y}}_k=Y_k{e}^{-j2\mathrm{\&pi;}\widehat{\mathrm{\&epsiv;}}k/N};$

The Fourier transform module, it obtains frequency-region signal Y for the signal y to after the time domain decimal frequency bias is proofreaied and correct through fast Fourier transform;

The integer frequency bias estimation module, it relevant cumulative searched peaked method and realizes that integer frequency bias estimates and compensation for what select the adjacent OFDM symbol continuous pilot, and the estimated value of normalization integer frequency bias meets following formula:

$\hat{m}=\underset{m\&Element;(-C,C)}{\arg \max \left(\right|W\left(m\right)\left|\right)}=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{Y}_i(k+m)Y_{(i+1)}^{*}(k+m)\left|\right)$

$=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{S}_i(k+m)e^{-j2\mathrm{\&pi;\&epsiv;}{N}_i/N}{S}_{i+1}^{*}(k+m)e^{j2\mathrm{\&pi;\&epsiv;}{N}_{(i+1)}/N}\left|\right),$ In formula, the subcarrier sequence number collection that P is continuous pilot place in the OFDM symbol, (C, C) is the scope of displacement search;

Pilot extraction module, it is for obtaining the pilot frequency information R of OFDM symbol _Sp

Frequency domain decimal frequency bias estimation module, it is for estimating remaining decimal frequency bias;

The symbol time offset estimation module, k in its same OFDM symbol received for employing ₁, k ₂Locational pilot tone y _K1, y _K2Cross-correlation, asked the phase angle computing and be averaging cross correlation value, obtains timing error and estimate

, $\widehat{\mathrm{\&sigma;}}=\underset{l=1}{\overset{1-\mathrm{NP}/2}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_l=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{\mathrm{N\&Delta;}{\mathrm{\&phi;}}_l}{2\mathrm{\&pi;\&Delta;k}}=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{N(\&angle;[y_{k1}{y}_{k2}^{*}\left]\right)}{2\mathrm{\&pi;\&Delta;k}},$ Δ k=k wherein ₂-k ₁, NP is continuous pilot number of sub carrier wave in a symbol, the carrier number that N is each OFDM symbol, σ _lFor the estimated value of utilizing l to use this algorithm to obtain to pilot tone;

The inherent spurious frequency deviation tracking module, it is for utilizing the continuous pilot subcarrier to carry out estimation and the compensation of excess phase; And

The sampling frequency offset correction module, it,, for utilizing pilot signal sampling frequency offset is estimated and proofreaied and correct at frequency domain, at first estimates the caused phase rotating of sampling frequency offset, and then accordingly each sampled value is compensated;

Channel equalizer, it is for adopting the channel estimation and equalization algorithm based on scattered pilot to realize channel estimating and channel equalization; And

Decoder, it is for carrying out the complex data symbol after chnnel coding obtains normalization according to input code flow.

2. the reception of wireless signals processing unit of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 1, is characterized in that, in the frame detection module, Frame Detection Algorithm is: detection statistic M (n) can be expressed as

$M\left(n\right)=\frac{{\left|c\left(n\right)\right|}^2}{{\left|p\left(n\right)\right|}^2}=\frac{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(k+n)r^{*}(n+k+N)\right|}^2}{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(n+k)\right|}^2},$ If it is output signal y that M (n)>TH detects frame start position _kOtherwise, do not detect, continue to detect; Wherein, TH is the judgement threshold, and c (n) is for receiving signal and the cross correlation value that receives signal lag, and p (n) means the energy of corresponding reception signal, and N means the carrier number of OFDM symbol.

3. the reception of wireless signals processing unit of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 1, is characterized in that, in frequency domain decimal frequency bias estimation module, for the continuous pilot in the OFDM symbol, when k ∈ P, meets S _I+1(k)=S _i(k); If only there is decimal frequency bias in the reception signal, draw thus the estimated value of decimal frequency bias

, wherein, what N was FFT counts, N _GFor the carrier number at protection interval, ∠ is the tangent of negating.

4. the reception of wireless signals processing unit of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 3, it is characterized in that, in order to increase the resistivity to noise, improve the accuracy of calculating, estimated value to a plurality of pilot tones of OFDM symbol inside averages, and obtains corresponding decimal frequency bias:

.

5. the reception of wireless signals processing unit of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 1, is characterized in that, in the inherent spurious frequency deviation tracking module,

The excess phase estimation

Meet following formula:

${\widehat{\mathrm{\&phi;}}}_m=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{\hat{R}}_{m,n}{\left(P_{m,n}\right)}^{*}\right]$

$=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{P}_{m,n}{e}^{j2\mathrm{\&pi;m\&Delta;f}}{\left(P_{m,n}\right)}^{*}\right]$

Wherein,

N pilot tone of m the OFDM symbol that expression receives, P _m,nFor the pilot tone of transmitting terminal, i.e. local known pilot, P _m,nFor+4/3 or-4/3;

Calculate compensating factor

Meet following formula:

$e^{-j{\mathrm{\&phi;}}_m}=e^{-j2\mathrm{\&pi;m\&Delta;f}}$

$=\frac{1}{64}\&times;e^{-j2\mathrm{\&pi;m\&Delta;f}}\&times;\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})$

$=\frac{1}{64}\&times;{\left[\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})\right]}^{*};$

Finally, according to following formula to receiving data

Carry out phase compensation:

.

6. the reception of wireless signals processing unit of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 1, is characterized in that, in the sampling frequency offset correction module,

Sampling frequency offset estimated value s _lBe shown below:

, wherein, P _kPilot sub-carrier, wherein k _jFor the subcarrier label, j=0,1,2 ..., M;

The sampling frequency offset compensation

, ${\hat{R}}_k={\hat{Y}}_k{e}^{-j2\mathrm{\&pi;}{s}_{l}k/N}.$

7. the reception of wireless signals processing method of Coded Orthogonal Frequency Division Multiplexing (COFDM), is characterized in that, this reception of wireless signals processing method comprises the following steps:

To receiving signal r _kFirst after the frame Check processing, obtain thick frame synchronizing signal y _k

To thick frame synchronizing signal y _kAdvanced after the time domain decimal frequency bias is proofreaied and correct and obtained signal y;

To thick frame synchronizing signal y _kCarry out Symbol Timing and proofread and correct the signal obtained through the timing offset correction

: ${\hat{y}}_k=y_k{e}^{-j2\mathrm{\&pi;k}\widehat{\mathrm{\&sigma;}}};$

Territory decimal frequency bias correction signal

, obtain signal

:

Signal y after the time domain decimal frequency bias is proofreaied and correct is obtained to frequency-region signal Y through fast Fourier transform;

That selects continuous pilot in adjacent OFDM symbol relevant cumulatively searches peaked method and realizes that integer frequency bias estimates and compensation, and the estimated value of normalization integer frequency bias meets following formula:

$\hat{m}=\underset{m\&Element;(-C,C)}{\arg \max \left(\right|W\left(m\right)\left|\right)}=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{Y}_i(k+m)Y_{(i+1)}^{*}(k+m)\left|\right)$

$=\underset{m\&Element;(-C,C)}{\arg \max }\left(\right|\underset{k\&Element;P}{\mathrm{\&Sigma;}}{S}_i(k+m)e^{-j2\mathrm{\&pi;\&epsiv;}{N}_i/N}{S}_{i+1}^{*}(k+m)e^{j2\mathrm{\&pi;\&epsiv;}{N}_{(i+1)}/N}\left|\right),$ In formula, the subcarrier sequence number collection that P is continuous pilot place in the OFDM symbol, (C, C) is the scope of displacement search;

Obtain the pilot frequency information R in the OFDM symbol _Sp

Estimate remaining decimal frequency bias;

K in the same OFDM symbol that employing receives ₁, k ₂Locational pilot tone y _K1, y _K2Cross-correlation, asked the phase angle computing and be averaging cross correlation value, obtains timing error and estimate , $\widehat{\mathrm{\&sigma;}}=\underset{l=1}{\overset{1-\mathrm{NP}/2}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_l=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{\mathrm{N\&Delta;}{\mathrm{\&phi;}}_l}{2\mathrm{\&pi;\&Delta;k}}=\underset{l=1}{\overset{l=\mathrm{NP}/2}{\mathrm{\&Sigma;}}}\frac{N(\&angle;[y_{k1}{y}_{k2}^{*}\left]\right)}{2\mathrm{\&pi;\&Delta;k}},$ Δ k=k wherein ₂-k ₁, NP is continuous pilot number of sub carrier wave in a symbol, the carrier number that N is each OFDM symbol, σ _lFor the estimated value of utilizing l to use this algorithm to obtain to pilot tone;

Utilize the continuous pilot subcarrier to carry out estimation and the compensation of excess phase;

Utilize pilot signal sampling frequency offset is estimated and proofreaied and correct at frequency domain, at first estimate the caused phase rotating of sampling frequency offset, and then accordingly each sampled value is compensated;

The channel estimation and equalization algorithm of employing based on scattered pilot realized channel estimating and channel equalization; And

Carry out the complex data symbol after chnnel coding obtains normalization according to input code flow.

8. the reception of wireless signals processing method of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 7, is characterized in that, in the frame detecting step, Frame Detection Algorithm is: detection statistic M (n) can be expressed as

$M\left(n\right)=\frac{{\left|c\left(n\right)\right|}^2}{{\left|p\left(n\right)\right|}^2}=\frac{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(k+n)r^{*}(n+k+N)\right|}^2}{{\left|\underset{k=0}{\overset{L-1}{\mathrm{\&Sigma;}}}r(n+k)\right|}^2},$ If it is output signal y that M (n)>TH detects frame start position _kOtherwise, do not detect, continue to detect; Wherein, TH is the judgement threshold, and c (n) is for receiving signal and the cross correlation value that receives signal lag, and p (n) means the energy of corresponding reception signal, and N means the carrier number of OFDM symbol.

9. the reception of wireless signals processing method of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 7, is characterized in that, in frequency domain decimal frequency bias estimating step, for the continuous pilot in the OFDM symbol, when k ∈ P, meets S _I+1(k)=S _i(k); If only there is decimal frequency bias in the reception signal, draw thus the estimated value of decimal frequency bias

, wherein, what N was FFT counts, N _GFor the carrier number at protection interval, ∠ is the tangent of negating.

10. the reception of wireless signals processing method of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 9, it is characterized in that, in order to increase the resistivity to noise, improve the accuracy of calculating, estimated value to a plurality of pilot tones of OFDM symbol inside averages, and obtains corresponding decimal frequency bias: .

11. the reception of wireless signals processing method of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 7, is characterized in that, in the inherent spurious frequency deviation tracking step, and the excess phase estimation

Meet following formula:

${\widehat{\mathrm{\&phi;}}}_m=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{\hat{R}}_{m,n}{\left(P_{m,n}\right)}^{*}\right]$

$=\&angle;\left[\underset{n=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{P}_{m,n}{e}^{j2\mathrm{\&pi;m\&Delta;f}}{\left(P_{m,n}\right)}^{*}\right]$

Wherein,

N pilot tone of m the OFDM symbol that expression receives, P _m,nFor the pilot tone of transmitting terminal, i.e. local known pilot, P _m,nFor+4/3 or-4/3;

Calculate compensating factor

Meet following formula:

$e^{-j{\mathrm{\&phi;}}_m}=e^{-j2\mathrm{\&pi;m\&Delta;f}}$

$=\frac{1}{64}\&times;e^{-j2\mathrm{\&pi;m\&Delta;f}}\&times;\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})$

$=\frac{1}{64}\&times;{\left[\underset{n=1}{\overset{45}{\mathrm{\&Sigma;}}}({\hat{R}}_{m,n}\&times;P_{m,n})\right]}^{*};$

Finally, according to following formula to receiving data

Carry out phase compensation:

.

12. the reception of wireless signals processing method of Coded Orthogonal Frequency Division Multiplexing (COFDM) as claimed in claim 7, is characterized in that, in the sampling frequency offset aligning step, and sampling frequency offset estimated value s _lBe shown below:

, wherein, P _kPilot sub-carrier, wherein k _jFor the subcarrier label, j=0,1,2 ..., M;

The sampling frequency offset compensation

, ${\hat{R}}_k={\hat{Y}}_k{e}^{-j2\mathrm{\&pi;}{s}_{l}k/N}.$

Images