CN105187351A - OFDM timing synchronization detection method under multipath channel - Google Patents

Abstract

The invention provides an OFDM timing synchronization detection method under a multipath channel. The OFDM timing synchronization detection method comprises the following steps: firstly, completing the initialization of a timing synchronization detection system, including the initialization of a sampling position counter and a first-in first-out memory; secondly, comparing a timing measure calculated by use of a timing estimation method based on weighed difference correlation with a set detection threshold; and finally, determining an OFDM timing offset estimator. The OFDM timing synchronization detection method under the multipath channel only requires a simple preamble with two sections of repeated structures and thus avoids the degradation of the power spectrum characteristics of an OFDM signal due to processing such as scrambling; the method is implemented by use of two simplified methods based on polar coordinates, and thus has relatively low implementation complexity; the timing estimation method based on weighted difference correlation is adopted to avoid the influence of a carrier frequency offset in the timing estimation process; besides, an analysis and simulation combined performance assessment method is provided so that the detection threshold meeting the requirements of a system detection probability and a false alarm probability can be set under a multipath fading channel.

Description

OFDM Timing Synchronization detection method under a kind of multipath channel

Technical field

The present invention relates to the OFDM Timing Synchronization detection method under a kind of OFDM Timing Synchronization detection method, particularly a kind of multipath channel, belong to digital wireless communication transmission technique field.

Background technology

OFDM (OrthogonalFrequencyDivisionMultiplexing, OFDM) technology has the advantages such as ability of anti-multipath is strong, the availability of frequency spectrum is high, support Large Copacity information transmission.At moving communicating field, OFDM is the core technology of forth generation land mobile communication.In satellite communication field, OFDM is applicable to the high speed data transfer under wideband satellite communication channel, and such as European satellite standard DVB-SH devises and adopts the mixing ground of OFDM or TDM technology and the communication system of satellite.

In ofdm communication system, the data flow inputted by high speed serialization obtains the sub data flow of low-speed parallel through serial to parallel conversion, and obtains N by subcarrier mapping _sroad parallel sub-stream pairs { X _k.Wherein, the information rate of every sub-data stream is reduced to the 1/N of input traffic _s, symbol period expands to the N of input traffic _sdoubly.Then, by inverse Fourier transform (InverseFastFourierTransform, IFFT) by N _ssub-data stream { X _kbe modulated to N _son individual parallel and mutually orthogonal subcarrier, its result obtains OFDM symbol after parallel serial conversion.In order to reduce the intersymbol interference (Inter-SymbolInterference, ISI) that multipath channel is introduced, before each OFDM symbol, add its end N _gindividual sampling is as Cyclic Prefix.Therefore, OFDM baseband signal x _nbe expressed as

$x_n=\frac{1}{\sqrt{N_s}}\underset{k=0}{\overset{N_s-1}{\Σ}}{X}_k{e}^{j2\mathrm{\π}kn/N_s},n=-N_g,...,0,...,N_s-1$

Wherein, N _sfor the size (generally getting the integral number power of 2) of IFFT/FFT, X _k(0≤k≤N _s-1) for a kth subcarrier raises the data message of system, N _gfor the Cyclic Prefix number of OFDM symbol.

After OFDM transmission signal experience multidiameter fading channel, usually there is the symbol time offset introduced by channel and carrier frequency offset, therefore OFDM baseband receiving signals r (n) can be expressed as

$r\left(n\right)=\underset{m=0}{\overset{L-1}{\Σ}}h\left(m\right)x_{n-\mathrm{\ϵ}-m}{e}^{j(2\mathrm{\π}vn/N)}+w\left(n\right)$

Wherein, ε is unknown symbol time offset, and v is with the normalized carrier frequency offset of subcarrier spacing, and w (n) is independent identically distributed multiple Gaussian process, and h (m) is channel impulse response, and L is the multipath number of channel.

Before receiving demodulation, need the original position by timing synchronization determination Received signal strength.Conventional symbol timing synchronization method adopts the leading symbol with repetitive structure to carry out timing estimation, as S & C method, Minn method etc. usually.S & C method adopts time domain to have the leading symbol of two sections of repetitive structures, carries out conjugation correlation computations timing metric, be expressed as by two sections, front and back to received signal:

$M_{sc}\left(d\right)=\frac{|\underset{n=0}{\overset{N_s/2-1}{\mathrm{\Σ}}}{r}^{*}(d+n)\·r(d+n+N_s/2){|}^2}{{\left(\underset{n=0}{\overset{N_s/2-1}{\mathrm{\Σ}}}\right|r(d+n+N_s/2){|}^2)}^2}$

Wherein, d is length is N _sdata segment in the position of the 1st sampled data.Owing to comprising the repetitive structure of leading symbol in Cyclic Prefix, there is platform in the timing metric of the method, and timing estimation poor-performing.In order to improve the timing metric of S & C method and improve timing estimation accuracy, Minn proposes to adopt the leading symbol C with four sections of repetitive structures, and meet C=[BB-B-B] form, wherein B represents length L _m=N _sthe lead data section of/4.The timing metric of Minn method is expressed as

$M_{mi}\left(d\right)=\frac{|\underset{k=0}{\overset{1}{\Σ}}\underset{n=0}{\overset{L_m-1}{\Σ}}{r}^{*}(d+2L_{m}k+n)\·r(d+2L_{m}k+n+L_m){|}^2}{{\left(\underset{k=0}{\overset{1}{\Σ}}\underset{n=0}{\overset{L_m-1}{\Σ}}\right|r(d+2L_{m}k+n+L_m){|}^2)}^2}$

Minn method improves the timing estimation accuracy of S & C method, but its timing metric still exists multiple peak value, and correct detection probability is lower under a multipath fading channel.

Summary of the invention

Technology of the present invention is dealt with problems and is: for the problem that the detection perform of conventional symbols timing synchronization algorithm in OFDM receiver under multidiameter fading channel is poor, the invention provides a kind of OFDM Timing Synchronization and detects and detection perform appraisal procedure.First timing metric is calculated by the timing estimation method relevant based on weighted difference, then according to the detection threshold meeting system requirements, Timing Synchronization detection is carried out to two of this timing metric peak values, wherein, detection threshold is arranged according to the Timing Synchronization detection perform appraisal procedure provided.The leading symbol with repetitive structure that the method only needs employing one general, there is good detection perform under a multipath fading channel, timing estimation process is not by the impact of carrier wave frequency deviation, and there is lower implementation complexity, under a multipath fading channel, the detection perform arranging can assessing OFDM Timing Synchronization detection method by the method meets the detection threshold λ that systems axiol-ogy probability and false alarm probability require.

Technical solution of the present invention is: the OFDM Timing Synchronization detection method under a kind of multipath channel, and step is as follows:

(1) sampling location counter d=0 is established; Initialization length is N _sthe push-up storage of/2, i.e. FIFO are R for storage format _fifo(d)={ b _syn, the data of d}, wherein b _synfor judging whether timing metric M (d) of position d exceedes the flag bit of setting detection threshold λ;

(2) make d=d+1, calculate timing metric M (d) according to timing estimation algorithms, and compare M (d) and detection threshold λ, if M (d)>=λ, then b _syn=1, otherwise b _syn=0;

(3) by data R _fifo={ b _syn, d} is stored in FIFO, if the R in FIFO _fifo(d) information sum N _mmeet N _m=N _s/ 2, then from FIFO sense data R _fifo(d-N _s/ 2) step (4), is entered; Otherwise return step (2);

(4) if R _fifo(d-N _s/ 2) and R _fifod () meets: b _syn(d-N _s/ 2)=b _synd ()=1, then OFDM Timing Synchronization detects successfully, timing slip estimator otherwise return step (2).

Adopt the timing estimation algorithms based on weighted difference is relevant to calculate timing metric M (d) in described step (2), concrete steps are:

(2-1) under polar coordinates, the known targeting signal with two sections of repetitive structures is made to be c (n)=A _{c (n)}exp{j θ _{c (n)}, Received signal strength is r (n+d)=A _{r (n+d)}exp{j θ _{r (n+d)}, wherein, A _{c (n)}for the amplitude of c (n), θ _{c (n)}for the phase place of c (n), A _{r (n+d)}for the amplitude of r (n+d), θ _{r (n+d)}for the phase place of r (n+d);

(2-2) be N by length in Received signal strength _sdata segment and known pilot signal c (n) conjugate multiplication of/2 obtain r ₀(n, d), specifically by formula:

$\begin{array}{c}{r}_0(n,d)=A_{r_0(n+d)}{A}_{c\left(n\right)}\·\exp \left\{{\mathrm{j\θ}}_{r_0(n+d)}\right\}\\ =A_{r(n+d)}{A}_{c\left(n\right)}\·\exp \left\{j\right({\mathrm{\θ}}_{r(n+d)}-{\mathrm{\θ}}_{c\left(n\right)}\}\end{array},d=0,1,...,M_s\×N$

Provide, wherein M _sbe the OFDM symbol number that 1 frame comprises, N=N _s+ N _gbe total number of data and Cyclic Prefix in 1 OFDM symbol, N _sfor IFFT/FFT size, N _gfor the Cyclic Prefix number of OFDM symbol;

(2-3) by sequence r ₀(n, d) with interval m, m=1 ..., M ₀calculate difference to be correlated with p (m, d), obtain M ₀individual difference correlation, specifically by formula:

$p(m,d)=\underset{k=m}{\overset{N_s/2-1}{\Σ}}{r}_0(k,d)r_0^{*}(k-m,d)=\underset{k=m}{\overset{N_s/2-1}{\Σ}}{A}_{r_0(k+d)}{A}_{r_0(k-m+d)}\exp \left\{j({\mathrm{\θ}}_{r_0(k+d)}-{\mathrm{\θ}}_{r_0(k-m+d)})\right\}$

Provide, wherein, N _s/ 2-m is the number of sum term, M ₀for adjustable parameter, and be positive integer, work as M ₀when=1, and difference correlated results p (1, d) be directly used in calculating timing metric, i.e. M (d)=p (1, d); Work as M ₀during >1, for calculating timing metric after being weighted summation to p (m, d).

(2-4) coefficient is adopted to be 1/M ₀average weighted, obtain correlation function P (d), specifically by formula:

$P\left(d\right)=\underset{m=1}{\overset{M_0}{\Σ}}\frac{1}{M_0}\·|p(m,d){|}^2$

Provide;

(2-5) with the energy of data segment to P (d) normalization, obtain based on relevant normalization timing metric M (d) of weighted difference, specifically by formula:

$M\left(d\right)=\frac{P\left(d\right)}{{\left(R\right(d\left)\right)}^2}=\frac{\frac{1}{M_0}\underset{m=1}{\overset{M_0}{\Σ}}|p(m,d){|}^2}{{\left(\underset{n=0}{\overset{N_s/2-1}{\Σ}}\right|r(n+d){|}^2)}^2}$

Provide.

M in described step (2-4) ₀value be: if N _s=64, then M ₀≤ 2; If N _s=128, then M ₀≤ 3; If N _s=256, then M ₀≤ 4; If N _s=512, then M ₀≤ 6; If N _s=1024, then M ₀≤ 8.

Be N by length in Received signal strength in described step (2-2) _sdata segment and known pilot signal c (n) conjugate multiplication of/2 obtain r ₀(n, d), concrete grammar is: make the amplitude of known pilot signal c (n) be A _{c (n)}=1, then r ₀the phase place of (n, d) is amplitude is conjugate multiplication realizes by means of only adder in FPGA.

Be N by length in Received signal strength in described step (2-2) _sdata segment and known pilot signal c (n) conjugate multiplication of/2 obtain r ₀by sequence r in (n, d) and step (2-3) ₀(n, d) with interval m, m=1 ..., M ₀calculate difference to be correlated with p (m, d), obtain M ₀individual difference correlation, concrete grammar is: the amplitude making Received signal strength r (n) and known pilot signal c (n) is A _{r (n+d)}=A _{c (n)}=1, then r ₀the phase place of (n, d) is ${\mathrm{\θ}}_{r_0(n,d)}={\mathrm{\θ}}_{r(n+d)}-{\mathrm{\θ}}_{c\left(n\right)},$ Amplitude is $A_{r_0(n,d)}=1,$ The phase place of p (m, d) is ${\mathrm{\θ}}_{p(m,d)}={\mathrm{\θ}}_{r_0(k+d)}-{\mathrm{\θ}}_{r_0(k-m+d)},$ Amplitude is A _{p (m, d)}=1, in step (2-2), conjugate multiplication is relevant with difference in step (2-3) is realized by adder and shift register in FPGA.

The targeting signal in described step (2-1) with two sections of repetitive structures is specially: c=[AA], and wherein A is length is N _sthe multiple random sequence of/2.At the frequency domain of ofdm system, be N by length _sto be mapped to length be N to the multiple random sequence of the MPSK/MQAM modulation of/2 _soFDM frequency domain sequence odd subcarriers on, even subcarriers is 0, and to map after frequency domain sequence carry out N _sobtain the leading symbol that time domain has two sections of repetitive structures after the IFFT of point, transmitting terminal time domain has leading symbol c (n) of two sections of repetitive structures, specifically by formula:

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

Provide, wherein C (k) is the data on a frequency domain preamble symbols kth subcarrier, N _sfor the size of IFFT/FFT.

Detection threshold λ in described step (2) is according to the detection probability P of system requirements _dwith false alarm probability P _farrange.

Detection probability P _drepresent the probability that correct detection case occurs, as sampled point ε and ε+N _swhen/2 place's timing metric values all exceed detection threshold, be judged as detecting successfully, detection probability P _dby formula:

P _D＝P _D1·P _D2＝P{M(d)≥λ|d＝ε}·P{M(d)≥λ|d＝ε+N _s/2}

Provide, wherein P _d1and P _d2be respectively the detection probability of two peak points of timing metric M (d), and P _d1and P _d2separate; Detection probability P _d1and P _d2represent sampled point ε and ε+N respectively _s/ 2 places are the peak value of timing metric and exceed the probability of detection threshold.

Described false alarm probability P _fadded up by following four kinds of situations and obtain:

(i) sampled point ε and ε-N _s/ 2 place's timing metric values all exceed detection threshold, then false alarm probability is by formula:

P _F1＝P{M(d)≥λ|d＝ε-N _s/2,d＝ε}

＝P{M(d)≥λ|d＝ε-N _s/2}·P{M(d)≥λ|d＝ε}

Provide;

(ii) sampled point ε+N _s/ 2 with ε+N _splace's timing metric value all exceedes detection threshold, then false alarm probability is by formula:

P _F2＝P{M(d)≥λ|d＝ε+N _s/2,d＝ε+N _s}

＝P{M(d)≥λ|d＝ε+N _s/2}·P{M(d)≥λ|d＝ε+N _s}

Provide;

(iii) sampled point ε+l and ε+l+N _s/ 2 place's timing metric values all exceed detection threshold, and l ∈ [1, L] represents l article of multipath channel, then false alarm probability is by formula:

P _F3(l)＝P{M(d)≥λ|d＝ε+l,d＝ε+l+N _s/2}

＝P{M(d)≥λ|d＝ε+l}·P{M(d)≥λ|d＝ε+l+N _s/2}

Provide;

(iv) sampled point d and d+N _sthe timing metric value at/2 places all exceedes detection threshold, then false alarm probability is by formula:

$\begin{array}{c}{P}_{F4}=P\{M\left(d\right)\≥\mathrm{\λ},M(d+N_s/2)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\\ =P\{M\left(d\right)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\·P\{M(d+N_s/2)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\end{array}$

Provide, wherein, S={ ε-N _s/ 2, ε, ε+N _s/ 2, ε+l} represents the sampling location under situation (i), situation (ii), situation (iii) and correct detection case, D={1 ..., M _s× N} represents the sampling location of 1 frame data, wherein, and M _sbe the OFDM symbol number that 1 frame comprises, N=N _s+ N _gbe total number of data and Cyclic Prefix in 1 OFDM symbol, N _sfor IFFT/FFT size, N _gfor the Cyclic Prefix number of OFDM symbol;

False alarm probability P _fby formula:

$P_F=\frac{P_{F1}+P_{F2}}{N_F}+\frac{1}{N_F}\underset{l=1}{\overset{L}{\Σ}}{P}_{F3}\left(l\right)+\frac{N_F-L-2}{N_F}{P}_{F4}$

Provide, wherein, N _f=M _s× N-1 is the total degree that false alarm condition occurs.

The present invention's beneficial effect is compared with prior art:

The OFDM Timing Synchronization detection method that the present invention proposes adopts a kind of timing estimation method relevant based on weighted difference, and timing estimation is not by the impact of carrier wave frequency deviation;

Accompanying drawing explanation

Fig. 1 is the bimodal timing testing process of OFDM Timing Synchronization detection method proposed by the invention;

Fig. 2 is the bimodal timing metric of OFDM Timing Synchronization detection method proposed by the invention;

Fig. 3 is the receiver performance characteristics of OFDM Timing Synchronization detection method proposed by the invention;

Fig. 4 is detection probability and the false alarm probability of OFDM Timing Synchronization detection method proposed by the invention.

Embodiment

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is further described in detail.

Under a multipath fading channel, in order to the timing metric solving traditional timing synchronization algorithm exists multiple peak value, the poor problem of timing detection perform, the invention provides a kind of Timing Synchronization to detect and detection perform appraisal procedure, it is characterized in that, for employing, there is the leading ofdm system of two sections of repetitive structures, a kind of timing estimation method of being correlated with based on weighted difference is adopted to obtain the timing metric of bimodal in receiver, and the spacing N of two peak values _s/ 2 samplings, then arrange the detection threshold meeting systems axiol-ogy probability and false alarm probability and require, and can obtain the position of leading symbol in receiving data sequence according to the testing process of synchronization detecting method in the present invention according to detection perform appraisal procedure.

The leading symbol with two sections of repetitive structures has following characteristics respectively at OFDM transmitting terminal and receiving terminal:

1) in OFDM transmitting terminal, the leading symbol that time domain has two sections of repetitive structures is expressed as c=[AA], and wherein A is length is N _slength, at the frequency domain of ofdm system, is N by the multiple random sequence of/2 _sto be mapped to length be N to the multiple random sequence of the MPSK/MQAM modulation of/2 _soFDM frequency domain sequence odd subcarriers on, even subcarriers is 0, and to map after frequency domain sequence carry out N _sobtain the leading symbol that time domain has two sections of repetitive structures after the IFFT of point, transmitting terminal time domain has leading symbol c (n) of two sections of repetitive structures, specifically by formula:

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

Provide, wherein C (k) is the data on a frequency domain preamble symbols kth subcarrier, N _sfor the size of IFFT.

2) in OFDM receiving terminal, the Received signal strength through multidiameter fading channel comprises symbol time offset and carrier frequency offset, is expressed as

$r\left(n\right)=y(n-\mathrm{\ϵ})e^{j(2\mathrm{\π}vn/N_s)}+w\left(n\right)=\underset{m=0}{\overset{L-1}{\Σ}}h\left(m\right)x(n-\mathrm{\ϵ}-m)e^{j(2\mathrm{\π}vn/N_s)}+w\left(n\right)$

Wherein, ε is unknown symbol time offset, and v is normalized carrier frequency offset, and to be variance be w (n) zero-mean complex Gaussian noise, the impulse response that h (m) is multidiameter fading channel, L is channel memory length.

Based on the ofdm signal of above-mentioned transmitting terminal and receiving terminal, the bimodal timing testing process of OFDM Timing Synchronization detection method proposed by the invention as shown in Figure 1.The method meets the detection threshold of systems axiol-ogy probability and false alarm probability by arranging, utilize and complete ofdm system Timing Synchronization detect based on be correlated with bimodal timing metric that timing estimation method obtains of weighted difference, obtain timing estimation amount thus obtain the position of leading symbol in receiving data sequence.The bimodal testing process of OFDM Timing Synchronization detection method proposed by the invention has following steps:

(1) sampling location counter d=0 is established; Initialization length is N _sthe push-up storage of/2, i.e. FIFO are R for storage format _fifo(d)={ b _syn, the data of d}, wherein b _synfor judging whether timing metric M (d) of position d exceedes the flag bit of setting detection threshold λ;

(2) make d=d+1, calculate timing metric M (d) according to timing estimation algorithms, and compare M (d) and detection threshold λ, if M (d)>=λ, then b _syn=1, otherwise b _syn=0;

Described calculating timing metric M (d), concrete steps are:

(2-1) under polar coordinates, the known targeting signal with two sections of repetitive structures is made to be c (n)=A _{c (n)}exp{j θ _{c (n)}, Received signal strength is r (n+d)=A _{r (n+d)}exp{j θ _{r (n+d)}, wherein, A _{c (n)}for the amplitude of c (n), θ _{c (n)}for the phase place of c (n), A _{r (n+d)}for the amplitude of r (n+d), θ _{r (n+d)}for the phase place of r (n+d); This method only needs a general leading symbol with two sections of repetitive structures, avoids the power spectrum characteristic that the process such as scrambling worsen ofdm signal; And this method adopts two kinds to realize based on polar method for simplifying, while raising OFDM receiver Timing Synchronization detection probability, there is lower implementation complexity;

(2-2) be N by length in Received signal strength _sdata segment and known pilot signal c (n) conjugate multiplication of/2 obtain r ₀(n, d), specifically by formula:

$\begin{array}{c}{r}_0(n,d)=A_{r_0(n+d)}{A}_{c\left(n\right)}\·\exp \left\{{\mathrm{j\θ}}_{r_0(n+d)}\right\}\\ =A_{r(n+d)}{A}_{c\left(n\right)}\·\exp \left\{j\right({\mathrm{\θ}}_{r(n+d)}-{\mathrm{\θ}}_{c\left(n\right)}\}\end{array},d=0,1,...,M_s\×N$

Provide, wherein M _sbe the OFDM symbol number that 1 frame comprises, N=N _s+ N _gbe total number of data and Cyclic Prefix in 1 OFDM symbol, N _sfor IFFT/FFT size, N _gfor the Cyclic Prefix number of OFDM symbol;

Above-mentioned computational process can also be: make the amplitude of known pilot signal c (n) be A _{c (n)}=1, then r ₀the phase place of (n, d) is amplitude is conjugate multiplication realizes by means of only adder in FPGA.

(2-3) by sequence r ₀(n, d) with interval m, m=1 ..., M ₀calculate difference to be correlated with p (m, d), obtain M ₀individual difference correlation, specifically by formula:

$p(m,d)=\underset{k=m}{\overset{N_s/2-1}{\Σ}}{r}_0(k,d)r_0^{*}(k-m,d)=\underset{k=m}{\overset{N_s/2-1}{\Σ}}{A}_{r_0(k+d)}{A}_{r_0(k-m+d)}\exp \left\{j({\mathrm{\θ}}_{r_0(k+d)}-{\mathrm{\θ}}_{r_0(k-m+d)})\right\}$

Provide, wherein, N _s/ 2-m is the number of sum term, M ₀for adjustable parameter, and be positive integer, work as M ₀when=1, and difference correlated results p (1, d) be directly used in calculating timing metric, i.e. M (d)=p (1, d); Work as M ₀during >1, for calculating timing metric after being weighted summation to p (m, d).Be specially:

(2-3-1) M is worked as ₀time less, employing coefficient is 1/M ₀average weighted, obtain correlation function P (d), specifically by formula:

$P\left(d\right)=\underset{m=1}{\overset{M_0}{\Σ}}\frac{1}{M_0}\·|p(m,d){|}^2$

Provide; M ₀value be: if N _s=64, then M ₀≤ 2; If N _s=128, then M ₀≤ 3; If N _s=256, then M ₀≤ 4; If N _s=512, then M ₀≤ 6; If N _s=1024, then M ₀≤ 8;

(2-3-2) with the energy of data segment to P (d) normalization, obtain based on relevant normalization timing metric M (d) of weighted difference, specifically by formula:

$M\left(d\right)=\frac{P\left(d\right)}{{\left(R\right(d\left)\right)}^2}=\frac{\frac{1}{M_0}\underset{m=1}{\overset{M_0}{\Σ}}|p(m,d){|}^2}{{\left(\underset{n=0}{\overset{N_s/2-1}{\Σ}}\right|r(n+d){|}^2)}^2}$

Provide.

Computational process in step (2-2) and step (2-3) can also be: the amplitude making Received signal strength r (n) and known pilot signal c (n) is A _{r (n+d)}=A _{c (n)}=1, then r ₀the phase place of (n, d) is ${\mathrm{\θ}}_{r_0(n,d)}={\mathrm{\θ}}_{r(n+d)}-{\mathrm{\θ}}_{c\left(n\right)},$ Amplitude is $A_{r_0(n,d)}=1,$ The phase place of p (m, d) is ${\mathrm{\θ}}_{p(m,d)}={\mathrm{\θ}}_{r_0(k+d)}-{\mathrm{\θ}}_{r_0(k-m+d)},$ Amplitude is A _{p (m, d)}=1, conjugate multiplication is relevant with difference to be realized respectively by adder and shift register in FPGA.(when conjugate multiplication is under rectangular coordinate, need four multipliers, can be multiplied by amplitude under polar coordinates, phase place is added and subtracted, i.e. expression formula in step (2-2) (2-3), puts 1 here eliminated amplitude and be multiplied by amplitude, only remaining phase place plus-minus, therefore only need to be realized by adder, as shift LD, during this realizes for concrete FPGA, data sampling continues the process entered, and can realize data and local data respective operations by shift LD.)

Described detection threshold λ is according to detection probability P given in advance _dwith false alarm probability P _farrange, the detection perform of OFDM Timing Synchronization detection method can be assessed by the method, and the detection threshold λ meeting systems axiol-ogy probability and false alarm probability and require is set.Be specially: need first to emulate acquisition

Detection probability P{M (the d) >=λ at ε place | d=ε },

ε+N _sdetection probability P{M (the d)>=λ at/2 places | d=ε+N _s/ 2},

ε-N _sfalse alarm probability P{M (the d)>=λ at/2 places | d=ε-N _s/ 2},

ε+N _sfalse alarm probability P{M (the d)>=λ at place | d=ε+N _s,

False alarm probability P{M (d) >=λ that ε+l locates | d=ε+l},

ε+l+N _sfalse alarm probability P{M (the d)>=λ at/2 places | d=ε+l+N _s/ 2},

The false alarm probability at d place $P\{M\left(d\right)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\},$

D+N _sthe false alarm probability at/2 places $P\{M(d+N_s/2)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}.$

After obtaining above detection probability and false alarm probability, can according to detection probability P _d, false alarm probability P _fand P _f1, P _f2, P _f3, P _f4calculating formula obtain detection probability under detection threshold λ and false alarm probability performance.Thus detection threshold λ is set according to the detection probability required by system and false alarm probability performance.

Described detection probability P _drepresent the probability that correct detection case occurs, as sampled point ε and ε+N _swhen/2 place's timing metric values all exceed detection threshold, be judged as inspection result, detection probability P _dby formula:

P _D＝P _D1·P _D2＝P{M(d)≥λ|d＝ε}·P{M(d)≥λ|d＝ε+N _s/2}

Provide, wherein P _d1and P _d2be respectively the detection probability of two peak points of timing metric M (d), and P _d1and P _d2separate; Detection probability P _d1and P _d2represent sampled point ε and ε+N respectively _s/ 2 places are the peak value of timing metric and exceed the probability of detection threshold.

Described false alarm probability P _fadded up by following four kinds of situations and obtain:

(i) sampled point ε and ε-N _s/ 2 place's timing metric values all exceed detection threshold, then false alarm probability is by formula:

P _F1＝P{M(d)≥λ|d＝ε-N _s/2,d＝ε}

＝P{M(d)≥λ|d＝ε-N _s/2}·P{M(d)≥λ|d＝ε}

Provide;

(ii) sampled point ε+N _s/ 2 with ε+N _splace's timing metric value all exceedes detection threshold, then false alarm probability is by formula:

P _F2＝P{M(d)≥λ|d＝ε+N _s/2,d＝ε+N _s}

＝P{M(d)≥λ|d＝ε+N _s/2}·P{M(d)≥λ|d＝ε+N _s}

Provide;

(iii) sampled point ε+l and ε+l+N _s/ 2 place's timing metric values all exceed detection threshold, and l ∈ [1, L] represents l article of multipath channel, then false alarm probability is by formula:

P _F3(l)＝P{M(d)≥λ|d＝ε+l,d＝ε+l+N _s/2}

＝P{M(d)≥λ|d＝ε+l}·P{M(d)≥λ|d＝ε+l+N _s/2}

Provide;

(iv) sampled point d and d+N _sthe timing metric value at/2 places all exceedes detection threshold, then false alarm probability is by formula:

$\begin{array}{c}{P}_{F4}=P\{M\left(d\right)\≥\mathrm{\λ},M(d+N_s/2)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\\ =P\{M\left(d\right)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\·P\{M(d+N_s/2)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\end{array}$

Provide, wherein, S={ ε-N _s/ 2, ε, ε+N _s/ 2, ε+l} represents the sampling location under situation (i), situation (ii), situation (iii) and correct detection case, D={1 ..., M _s× N} represents the sampling location of 1 frame data, wherein, and M _sbe the OFDM symbol number that 1 frame comprises, N=N _s+ N _gbe total number of data and Cyclic Prefix in 1 OFDM symbol, N _sfor IFFT/FFT size, N _gfor the Cyclic Prefix number of OFDM symbol;

False alarm probability P _fby formula:

$P_F=\frac{P_{F1}+P_{F2}}{N_F}+\frac{1}{N_F}\underset{l=1}{\overset{L}{\Σ}}{P}_{F3}\left(l\right)+\frac{N_F-L-2}{N_F}{P}_{F4}$

Provide, wherein, N _f=M _s× N-1 is the false-alarm sum obtained of sampling.

(3) by data R _fifo={ b _syn, d} is stored in FIFO, if the R in FIFO _fifo(d) information sum N _mmeet N _m=N _s/ 2, then from FIFO sense data R _fifo(d-N _s/ 2) step (4), is entered; Otherwise return step (2);

(4) if R _fifo(d-N _s/ 2) and R _fifod () meets: b _syn(d-N _s/ 2)=b _synd ()=1, then OFDM Timing Synchronization detects successfully, timing slip estimator otherwise return step (2).Main thought of the present invention is: provide a kind of timing detection method being applicable to OFDM wireless communication system, and the performance estimating method of a kind of Conjoint Analysis of the method for deriving and emulation.The leading symbol that the timing detection method proposed only needs employing one to have two sections of repetitive structures obtains bimodal timing metric by the timing estimation method relevant based on weighted difference, synchronously detecting by arranging the detection threshold completion timing meeting systems axiol-ogy probability and false alarm probability, obtaining timing estimation amount under a multipath fading channel, select suitable detection threshold by the performance estimating method of Conjoint Analysis and emulation, the OFDM Timing Synchronization detection method proposed can obtain good detection perform.

Specific embodiment

In ofdm communication system, if sub-carrier number N _s=256, the effective sub-carrier number N of user _u=180, Cyclic Prefix number N _g=24, signal bandwidth is B _w=3MHz, subcarrier spacing is Δ f=15kHz, carrier wave frequency deviation v=3.4 Δ f.System adopts the block leading symbol with two sections of repetitive structures.Emulation adopts multipath number to be the rayleigh fading channel of L=8, the time delay τ in every footpath _ibe 0,2,4 ..., 14 samplings, channel has index power delay profile, namely for path gain A _ihave: wherein i represents i-th multipath, 0≤i≤L-1.

The selection of detection threshold needs between detection probability and false alarm probability compromise.Fig. 3 is the receiver performance characteristics (ReceiverOperatingCharacteristic, ROC) of OFDM timing estimation method 2 under different signal to noise ratio under provided timing detection method, each false alarm probability P _fa corresponding alarm dismissal probability P _md=1-P _d.As shown in Figure 3, when signal to noise ratio is SNR=5dB, adopt the false alarm probability P that the timing of timing estimation method 2 detects _f=6.2 × 10 ^-5corresponding alarm dismissal probability P _md=8.7 × 10 ^-4.When signal to noise ratio is SNR=10dB, adopt the false alarm probability P that the timing of timing estimation method 2 detects _f=5.3 × 10 ^-6, corresponding alarm dismissal probability P _md=3.5 × 10 ^-4.

Fig. 4 is the detection probability P that under different detection threshold, under λ, timing detects _dwith false alarm probability P _f.According to false alarm probability P _fwith alarm dismissal probability P _md, the detection threshold of performance requirement can be met from Fig. 4.False alarm probability P during SNR=5dB _f=6.2 × 10 ^-5, alarm dismissal probability P _md=8.7 × 10 ^-4corresponding detection threshold λ=0.048, false alarm probability P during SNR=10dB _f=5.3 × 10 ^-6, alarm dismissal probability P _md=3.5 × 10 ^-4corresponding detection threshold λ=0.081.

Can there is carrier wave frequency deviation in dark place from simulated conditions, the result of its timing estimation is analogous diagram 2, even if clearly there is larger carrier wave frequency deviation, timing metric is also cleaner bimodal, and the method therefore in the present invention is very little by the impact of carrier wave frequency deviation;

The content be not described in detail in specification of the present invention belongs to the known technology of professional and technical personnel in the field.

Claims (9)

1. the OFDM Timing Synchronization detection method under multipath channel, is characterized in that step is as follows:

(1) sampling location counter d=0 is established; Initialization length is N _sthe push-up storage of/2, i.e. FIFO are R for storage format _fifo(d)={ b _syn, the data of d}, wherein b _synfor judging whether timing metric M (d) of position d exceedes the flag bit of setting detection threshold λ;

(2) make d=d+1, calculate timing metric M (d) according to timing estimation algorithms, and compare M (d) and detection threshold λ, if M (d)>=λ, then b _syn=1, otherwise b _syn=0;

(3) by data R _fifo={ b _syn, d} is stored in FIFO, if the R in FIFO _fifo(d) information sum N _mmeet N _m=N _s/ 2, then from FIFO sense data R _fifo(d-N _s/ 2) step (4), is entered; Otherwise return step (2);

(4) if R _fifo(d-N _s/ 2) and R _fifod () meets: b _syn(d-N _s/ 2)=b _synd ()=1, then OFDM Timing Synchronization detects successfully, timing slip estimator otherwise return step (2).

2. the OFDM Timing Synchronization detection method under a kind of multipath channel according to claim 1, it is characterized in that: adopt the timing estimation algorithms based on weighted difference is relevant to calculate timing metric M (d) in described step (2), concrete steps are:

(2-1) under polar coordinates, the known targeting signal with two sections of repetitive structures is made to be c (n)=A _{c (n)}exp{j θ _{c (n)}, Received signal strength is r (n+d)=A _{r (n+d)}exp{j θ _{r (n+d)}, wherein, A _{c (n)}for the amplitude of c (n), θ _{c (n)}for the phase place of c (n), A _{r (n+d)}for the amplitude of r (n+d), θ _{r (n+d)}for the phase place of r (n+d);

(2-2) be N by length in Received signal strength _sdata segment and known pilot signal c (n) conjugate multiplication of/2 obtain r ₀(n, d), specifically by formula:

$\begin{array}{c}{r}_0(n,d)=A_{r_0(n+d)}{A}_{c\left(n\right)}\·\exp \left\{{\mathrm{j\θ}}_{r_0(n+d)}\right\}\\ =A_{r(n+d)}{A}_{c\left(n\right)}\·\exp \left\{j\right({\mathrm{\θ}}_{r(n+d)}-{\mathrm{\θ}}_{c\left(n\right)}\}\end{array},d=0,1,...,M_s\×N$

Provide, wherein M _sbe the OFDM symbol number that 1 frame comprises, N=N _s+ N _gbe total number of data and Cyclic Prefix in 1 OFDM symbol, N _sfor IFFT/FFT size, N _gfor the Cyclic Prefix number of OFDM symbol;

(2-3) by sequence r ₀(n, d) with interval m, m=1 ..., M ₀calculate difference to be correlated with p (m, d), obtain M ₀individual difference correlation, specifically by formula:

$p(m,d)=\underset{k=m}{\overset{N_s/2-1}{\Σ}}{r}_0(k,d)r_0^{*}(k-m,d)=\underset{k=m}{\overset{N_s/2-1}{\Σ}}{A}_{r_0(k+d)}{A}_{r_0(k-m+d)}\exp \left\{j({\mathrm{\θ}}_{r_0(k+d)}-{\mathrm{\θ}}_{r_0(k-m+d)})\right\}$

Provide, wherein, N _s/ 2-m is the number of sum term, M ₀for adjustable parameter, and be positive integer, work as M ₀when=1, and difference correlated results p (1, d) be directly used in calculating timing metric, i.e. M (d)=p (1, d); Work as M ₀during >1, for calculating timing metric after being weighted summation to p (m, d).

(2-4) coefficient is adopted to be 1/M ₀average weighted, obtain correlation function P (d), specifically by formula:

$P\left(d\right)=\underset{m=1}{\overset{M_0}{\Σ}}\frac{1}{M_0}\·|p(m,d){|}^2$

Provide;

(2-5) with the energy of data segment to P (d) normalization, obtain based on relevant normalization timing metric M (d) of weighted difference, specifically by formula:

$M\left(d\right)=\frac{P\left(d\right)}{{\left(R\right(d\left)\right)}^2}=\frac{\frac{1}{M_0}\underset{m=1}{\overset{M_0}{\Σ}}|p(m,d){|}^2}{{\left(\underset{n=0}{\overset{N_s/2-1}{\Σ}}\right|r(n+d){|}^2)}^2}$

Provide.

3. the OFDM Timing Synchronization detection method under a kind of multipath channel according to claim 2, is characterized in that: M in described step (2-4) ₀value be: if N _s=64, then M ₀≤ 2; If N _s=128, then M ₀≤ 3; If N _s=256, then M ₀≤ 4; If N _s=512, then M ₀≤ 6; If N _s=1024, then M ₀≤ 8.

4. the OFDM Timing Synchronization detection method under a kind of multipath channel according to claim 2, is characterized in that: be N by length in Received signal strength in described step (2-2) _sdata segment and known pilot signal c (n) conjugate multiplication of/2 obtain r ₀(n, d), concrete grammar is: make the amplitude of known pilot signal c (n) be A _{c (n)}=1, then r ₀the phase place of (n, d) is amplitude is conjugate multiplication realizes by means of only adder in FPGA.

5. the OFDM Timing Synchronization detection method under a kind of multipath channel according to claim 2, is characterized in that: be N by length in Received signal strength in described step (2-2) _sdata segment and known pilot signal c (n) conjugate multiplication of/2 obtain r ₀by sequence r in (n, d) and step (2-3) ₀(n, d) with interval m, m=1 ..., M ₀calculate difference to be correlated with p (m, d), obtain M ₀individual difference correlation, concrete grammar is: the amplitude making Received signal strength r (n) and known pilot signal c (n) is A _{r (n+d)}=A _{c (n)}=1, then r ₀the phase place of (n, d) is ${\mathrm{\θ}}_{r_0(n,d)}={\mathrm{\θ}}_{r(n+d)}-{\mathrm{\θ}}_{c(n)},$ Amplitude is $A_{r_0(n,d)}=1,$ The phase place of p (m, d) is ${\mathrm{\θ}}_{p(m,d)}={\mathrm{\θ}}_{r_0(k+d)}-{\mathrm{\θ}}_{r_0(k-m+d)},$ Amplitude is A _{p (m, d)}=1, in step (2-2), conjugate multiplication is relevant with difference in step (2-3) is realized by adder and shift register in FPGA.

6. the OFDM Timing Synchronization detection method under a kind of multipath channel according to claim 2, is characterized in that: the targeting signal in described step (2-1) with two sections of repetitive structures is specially: c=[AA], and wherein A is length is N _sthe multiple random sequence of/2.At the frequency domain of ofdm system, be N by length _sto be mapped to length be N to the multiple random sequence of the MPSK/MQAM modulation of/2 _soFDM frequency domain sequence odd subcarriers on, even subcarriers is 0, and to map after frequency domain sequence carry out N _sobtain the leading symbol that time domain has two sections of repetitive structures after the IFFT of point, transmitting terminal time domain has leading symbol c (n) of two sections of repetitive structures, specifically by formula:

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

Provide, wherein C (k) is the data on a frequency domain preamble symbols kth subcarrier, N _sfor the size of IFFT/FFT.

7. the OFDM Timing Synchronization detection method under a kind of multipath channel according to claim 1, is characterized in that: the detection threshold λ in described step (2) is according to the detection probability P of system requirements _dwith false alarm probability P _farrange.

8. the OFDM Timing Synchronization detection method under a kind of multipath channel according to claim 7, is characterized in that: detection probability P _drepresent the probability that correct detection case occurs, as sampled point ε and ε+N _swhen/2 place's timing metric values all exceed detection threshold, be judged as detecting successfully, detection probability P _dby formula:

P _D＝P _D1·P _D2＝P{M(d)≥λ|d＝ε}·P{M(d)≥λ|d＝ε+N _s/2}

Provide, wherein P _d1and P _d2be respectively the detection probability of two peak points of timing metric M (d), and P _d1and P _d2separate; Detection probability P _d1and P _d2represent sampled point ε and ε+N respectively _s/ 2 places are the peak value of timing metric and exceed the probability of detection threshold.

9. the OFDM Timing Synchronization detection method under a kind of multipath channel according to claim 7, is characterized in that: described false alarm probability P _fadded up by following four kinds of situations and obtain:

(i) sampled point ε and ε-N _s/ 2 place's timing metric values all exceed detection threshold, then false alarm probability is by formula:

P _F1＝P{M(d)≥λ|d＝ε-N _s/2,d＝ε}

＝P{M(d)≥λ|d＝ε-N _s/2}·P{M(d)≥λ|d＝ε}

Provide;

(ii) sampled point ε+N _s/ 2 with ε+N _splace's timing metric value all exceedes detection threshold, then false alarm probability is by formula:

P _F2＝P{M(d)≥λ|d＝ε+N _s/2,d＝ε+N _s}

＝P{M(d)≥λ|d＝ε+N _s/2}·P{M(d)≥λ|d＝ε+N _s}

Provide;

(iii) sampled point ε+l and ε+l+N _s/ 2 place's timing metric values all exceed detection threshold, and l ∈ [1, L] represents l article of multipath channel, then false alarm probability is by formula:

P _F3(l)＝P{M(d)≥λ|d＝ε+l,d＝ε+l+N _s/2}

＝P{M(d)≥λ|d＝ε+l}·P{M(d)≥λ|d＝ε+l+N _s·2}

Provide;

(iv) sampled point d and d+N _sthe timing metric value at/2 places all exceedes detection threshold, then false alarm probability is by formula:

$\begin{array}{c}{P}_{F4}=P\{M\left(d\right)\≥\mathrm{\λ},M(d+N_s/2)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\\ =P\{M\left(d\right)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\·P\{M(d+N_s/2)\≥\mathrm{\λ}|d\∈D\∩\stackrel{\‾}{S}\}\end{array}$

Provide, wherein, S={ ε-N _s/ 2, ε, ε+N _s/ 2, ε+l} represents the sampling location under situation (i), situation (ii), situation (iii) and correct detection case, D={1 ..., M _s× N} represents the sampling location of 1 frame data, wherein, and M _sbe the OFDM symbol number that 1 frame comprises, N=N _s+ N _gbe total number of data and Cyclic Prefix in 1 OFDM symbol, N _sfor IFFT/FFT size, N _gfor the Cyclic Prefix number of OFDM symbol;

False alarm probability P _fby formula:

$P_F=\frac{P_{F1}+P_{F2}}{N_F}+\frac{1}{N_F}\underset{l=1}{\overset{L}{\Σ}}{P}_{F3}\left(l\right)+\frac{N_F-L-2}{N_F}{P}_{F4}$

Provide, wherein, N _f=M _s× N-1 is the total degree that false alarm condition occurs.