CN103746954B - Associated synchronization and frequency offset estimation method for OFDM (Orthogonal Frequency Division Multiplexing) system - Google Patents

Abstract

The invention discloses an associated synchronization and frequency offset estimation method for an OFDM (Orthogonal Frequency Division Multiplexing) system. The method comprises the steps of obtaining sampling point data of two continuous OFDM symbols taking an ith sampling point of a received signal as a starting point; carrying out self-correlation operation on a sampling point in a cyclic prefix of the latter OFDM symbol and a sampling point, and solving average values of the obtained G values, wherein the first sampling point is away from the second sampling point for N sampling points at the tail of the same symbol; carrying out self-correlation operation on the sampling point in the cyclic prefix of the latter OFDM symbol and a sampling point in the precious OFDM symbol, and solving average values of the obtained G values, wherein the first sampling point is away from the second sampling point for N sampling points; adding the obtained results together, searching M addition results according to the number of the OFDM symbols, and solving average values to obtain phi (i) and phi (i), wherein i corresponding to the maximum phi (i) in a watch window provides an OFDM symbol synchronous position, and the phase angle of phi (i) corresponding to the synchronous position provides a frequency offset estimation value. The associated synchronization and frequency offset estimation method for the OFDM system is free of multi-path interference and low in computation complexity, and no training sequence is needed.

Description

A kind of combined synchronization for ofdm system and frequency deviation estimating method

Technical field

The present invention relates to the simultaneous techniquess field in mobile communication, more particularly, to one kind is applied to OFDM (OFDM) symbol of system and frequency synchronization method.

Background technology

In mobile communication system, sign synchronization and Nonlinear Transformation in Frequency Offset Estimation be the premise that can correctly demodulate of receiving data it One.At present, OFDM has become the main flow air interface technologies of forth generation (4G) mobile communication system.4G system is in up-downgoing respectively Using DFT-s-OFDM and OFDM technology.In an ofdm system, in order to eliminate intersymbol interference (ISI), each OFDM symbol A Cyclic Prefix above will be inserted.Cyclic Prefix is the duplication of symbol afterbody, it be more than multidiameter fading channel maximum when Prolong extension, and the linear convolution of sending signal and multipath channel is changed into cyclic convolution simultaneously.In an ofdm system, traditional determine When synchronous method mainly by auto-correlation computation is carried out to training sequence, then look for the method for peak value to find rising of symbol Initial point.This time synchronization method not only needs training sequence known to transmitting and receiving terminal, and due to cannot remove between multipath Interfere, so the synchronizing information obtaining all have in multidiameter fading channel inclined.

Content of the invention

In order to overcome the deficiencies in the prior art, the present invention provides a kind of combined synchronization for ofdm system and frequency Bias estimation, does not have between multipath interference, and computation complexity low it is not necessary to any training sequence；Fast and reliable of the present invention, estimate Count high precision, implementation complexity is low, be not only suitable to ofdm system but also be suitable to the same of other mobile communication system with Cyclic Prefix Step method of estimation

For achieving the above object, the present invention adopts the following technical scheme that：

A kind of combined synchronization for ofdm system and frequency deviation estimating method, comprise the steps：

(1) with receipt signal ith sample point in observation window as starting point, two continuous OFDM symbol m-1 are extracted, m's The initial value of sample point data, wherein i is 0, i ∈ Ω, and Ω is the number of sampled point in observation window；The initial value of m is 1, m ∈ Γ, Γ are the number of OFDM symbol in receipt signal；Hypothesis observation window is 100 sampled points, and the OFDM symbol of reception has 1,2, 3,4,5,6,80 sampled points of each OFDM symbol, first assume that the 1st sampled point in observation window starts for OFDM symbol, first With in observation window the 1st sampled point, as starting point, takes out 160 sampled datas, is equivalent to two OFDM symbol, corresponding to OFDM symbol Number 1 and 2, carry out operation below.Then more ensuing OFDM symbol 3 and 4 is operated, be finally OFDM symbol 5 He 6.Assume again that the in observation window the 2nd sampled point starts for OFDM symbol, circulate aforesaid operations.

(2) set the number of m-th OFDM symbol cyclic prefix samples point as G, the sub-carrier number of ofdm system is N, successively Sampled point in m-th OFDM symbol Cyclic Prefix at a distance of N number of sampled point and is located at m-th OFDM symbol afterbody with it Sampled point carries out autocorrelation operation, is then averaging G obtained autocorrelation value；

(3) will adopt at a distance of N number of with it in the sampled point in m-th OFDM symbol Cyclic Prefix and the m-1 OFDM symbol The sampled point of sampling point carries out autocorrelation operation, and G obtained autocorrelation value is averaging；

(4) results added that will obtain in step (2) and step (3)；

(5) m=m+2, repeat step (1) to step (4), and flat is asked to the described addition result of all steps (4) obtaining All, thus obtaining the unbiased estimator of unbiased estimator φ (i) to channel autocorroelation function and carrier wave frequency deviation

(6) i=i+1, to each sampled point of receipt signal in observation window successively execution step (1) to step (5), then observes In window, the maximum i corresponding to φ (i) is the sync bit of OFDM symbol, and this sync bit is correspondingPhase angle be carry The inclined estimated value of wave frequency.

Beneficial effect：(1) present invention passes through to introduce the associative operation between continuous two OFDM symbol, thoroughly gets rid of Interfering between channel multi-path, has obtained synchronizing information and the offset estimation value of unbiased；(2) this method does not need to utilize and appoints What pilot signal or training sequence known to receiving terminal, does not increase any overhead to system；(3) joint proposed by the present invention Synchronous and frequency deviation estimating method can be used for the various mobile communication system with Cyclic Prefix including OFDM.

Brief description

Fig. 1 is the workflow diagram of the present invention.

Fig. 2 is the transposition structural representation realizing the inventive method.

Specific embodiment

Below in conjunction with the accompanying drawings the present invention is further described.

As shown in figure 1, a kind of combined synchronization for ofdm system of present invention offer and frequency deviation estimating method, including such as Lower step：

(1) with receipt signal ith sample point in observation window as starting point, two continuous OFDM symbol m-1 are extracted, m's The initial value of sample point data, wherein i is 0, i ∈ Ω, and Ω is the number of sampled point in observation window；The initial value of m is 1, m ∈ Γ, Γ are the number of OFDM symbol in receipt signal；

(2) set the number of m-th OFDM symbol cyclic prefix samples point as G, the sub-carrier number of ofdm system is N, successively Sampled point in m-th OFDM symbol Cyclic Prefix at a distance of N number of sampled point and is located at m-th OFDM symbol afterbody with it Sampled point carries out autocorrelation operation, is then averaging G obtained autocorrelation value；

(3) will adopt at a distance of N number of with it in the sampled point in m-th OFDM symbol Cyclic Prefix and the m-1 OFDM symbol The sampled point of sampling point carries out autocorrelation operation, and G obtained autocorrelation value is averaging；

(4) results added that will obtain in step (2) and step (3)；

(5) m=m+2, repeat step (1) to step (4), and flat is asked to the described addition result of all steps (4) obtaining All, thus obtaining the unbiased estimator of unbiased estimator φ (i) to channel autocorroelation function and carrier wave frequency deviation

(6) i=i+1, to each sampled point of receipt signal in observation window successively execution step (1) to step (5), then observes In window, the maximum i corresponding to φ (i) is the sync bit of OFDM symbol, and sync bit is correspondingPhase angle be carrier wave The estimated value of frequency deviation.

Assume that an ofdm system comprises N number of subcarrier, a width of B of band of each subcarrier, the length of Cyclic Prefix is G. One OFDM symbol comprises N+G sampled point, and label is from 0 to N+G-1.The sampling period of each sampled point is T, and T=1/NB. With in observation window, ith sample point obtains the sample point data of continuous two OFDM symbol for starting point.First by rear OFDM symbol Sampled point in number Cyclic Prefix carries out autocorrelation operation in same symbol afterbody at a distance of the sampled point of N number of sampled point with it, Again by the sampled point in a rear OFDM symbol Cyclic Prefix and previous OFDM symbol with its at a distance of N number of sampled point sampling Point carries out autocorrelation operation, then this two results added is obtained meansigma methodss.Finally, by maximum value search device in window, The starting point of estimate symbol and the frequency deviation of carrier wave.

Specific algorithm is described as follows：

Assume that m-th OFDM symbol receiving can be expressed as：

$\begin{array}{c}{y}_m\left(n\right)=e^{j2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n\]/N}\underset{l=0}{\overset{L-1}{\Σ}}{h}_l^m\left(n\right)x_m{\left(\right(n-G-l-\mathrm{\θ}\left)\right)}_N+v_m\left(n\right)\\ \mathrm{\θ}+G\≤n\≤\mathrm{\θ}+N+G-1\end{array}---\left(1\right)$

In formula, ε represent by B normalization after carrier wave frequency deviation；θ represents timing offset；L is the multipath number of channel, (())_N Represent the circulative shift operation with N as base.v_mN () is the white Gaussian noise of additive zero, variance is Represent the channel parameter n-th moment for the l paths, be the Gaussian random variable of zero-mean, variance isThe auto-correlation function of l paths is：

$E\{h_l\left(m\right)\·h_l^{*}(m+n)\}={\mathrm{\σ}}_l^2{J}_0\left(2{\mathrm{\πf}}_{d}nT\right)---\left(2\right)$

In formula, J₀() represents zero-order Bessel (Bessel) function, f_dFor maximum doppler frequency.

With the ith sample point of receipt signal as starting point, construct following estimator：

$\begin{array}{c}{\mathrm{\ρ}}_m^i=\frac{1}{G}\underset{n=0}{\overset{G-1}{\Σ}}{\mathrm{\λ}}_m^i\left(n\right)\\ =\frac{1}{G}\underset{n=0}{\overset{G-1}{\Σ}}\[y_m(n+i)y_m^{*}(n+N+i)+y_m(n+i)y_{m-1}^{*}(n+G+i)\]\end{array}---\left(3\right)$

Assume i=θ, as 0≤n ＜ L-1,

$\begin{array}{c}E\{y_m(n+\mathrm{\θ})y_m^{*}(n+N+\mathrm{\θ})+y_m(n+\mathrm{\θ})y_{m-1}^{*}(n+G+\mathrm{\θ})\}\\ =E\{e^{j\frac{2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n\]}{N}}\underset{l=0}{\overset{n}{\Σ}}{x}_m(N-G+n-l)h_l^m\left(n\right)\\ {\[e^{j\frac{2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n+N\]}{N}}{x}_m(N-G+n-l)h_l^m(n+N)\]}^{*}\}\\ +E\{e^{j\frac{2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n\]}{N}}\underset{l=n+1}{\overset{L-1}{\Σ}}{x}_{m-1}(N+n-l)h_l^m\left(n\right)\\ {\[e^{j\frac{2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n+N\]}{N}}{x}_m(N-G+n-l)h_l^m(n+N)\]}^{*}\}\\ +E\{e^{j\frac{2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n\]}{N}}\underset{l=0}{\overset{n}{\Σ}}{x}_m(N-G+n-l)h_l^m\left(n\right)\\ {\[e^{j\frac{2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n-N\]}{N}}{x}_{m-1}(n-l)h_l^{m-1}(n+G)\]}^{*}\}\\ +E\{e^{j\frac{2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n\]}{N}}\underset{l=n+1}{\overset{L-1}{\Σ}}{x}_{m-1}(N+n-l)h_l^m\left(n\right)\\ {\[e^{j\frac{2\mathrm{\π}\mathrm{\ϵ}\[(N+G)m+n-N\]}{N}}{x}_{m-1}(N+n-l)h_l^{m-1}(n+G)\]}^{*}\}\end{array}---\left(4\right)$

Sampling { x due to m-1 symbol_m-1(n) } with the sampling { x of m-th symbol_m(n) } separate, and channel General power is 1, and we obtain：

$\begin{array}{c}E\{y_m(n+\mathrm{\θ})y_m^{*}(n+N+\mathrm{\θ})+y_m(n+\mathrm{\θ})y_{m-1}^{*}(n+G+\mathrm{\θ})\}\\ =e^{-j2\mathrm{\π}\mathrm{\ϵ}}\underset{i=0}{\overset{L-1}{\Σ}}{\mathrm{\σ}}_l^2{J}_0\left(2{\mathrm{\πf}}_{d}NT\right)\\ =e^{-j2\mathrm{\π}\mathrm{\ϵ}}{J}_0\left(2{\mathrm{\πf}}_{d}NT\right)\end{array}---\left(5\right)$

As n >=L-1,

$\begin{array}{c}E\{y_m(n+\mathrm{\θ})y_m^{*}(n+N+\mathrm{\θ})+y_m(n+\mathrm{\θ})y_{m-1}^{*}(n+G+\mathrm{\θ})\}\\ =e^{-j2\mathrm{\π}\mathrm{\ϵ}}{J}_0\left(2{\mathrm{\πf}}_{d}NT\right)\end{array}---\left(6\right)$

Then,

$E\left\{{\mathrm{\ρ}}_m^{\mathrm{\θ}}\right\}=e^{-j2\mathrm{\π}\mathrm{\ϵ}}{J}_0\left(2{\mathrm{\πf}}_{d}NT\right)---\left(7\right)$

From formula (7),The unbiased esti-mator being contemplated to be to channel autocorroelation function and carrier wave frequency deviation.Specific In Project Realization,Expectation can be by averagely asking for M result, wherein the size of M is according in receipt signal The number of OFDM symbol calculates.Hypothesis observation window is 100 sampled points, and the OFDM symbol of reception has 1,2,3,4,5,6, first Assume that the 1st sampled point in observation window starts for OFDM symbol, first the with observation window the 1st sampled point, as starting point, takes out 160 Individual sampled data, is equivalent to two OFDM symbol, corresponding to OFDM symbol 1 and 2, carries out operation below.Then dock down again The OFDM symbol 3 and 4 come is operated, and is finally OFDM symbol 5 and 6, i.e. M=3 analogizes in the same manner, during odd number, does not consider last One OFDM symbol.Thus, we obtain：

$\mathrm{\φ}\left(i\right)=\frac{1}{M}\underset{m=1}{\overset{M}{\Σ}}\left|{\mathrm{\ρ}}_m^i\right|---\left(8\right)$

Finally started with the next sampled point in observation window for OFDM symbol, time aforesaid operations of circulation are pre-defined Window in, the maximum sampled point i corresponding to φ (i) is the starting point of OFDM symbol.Starting point is correspondingPhase Angle is the estimated result of carrier wave frequency deviation.

Fig. 2 is the transposition structural representation realizing the inventive method, and multiplier completes the related fortune comprising in formula (3) Calculate, adder completes the additive operation comprising in formula (3) and formula (8), and in window, maximum value search device completes maximum Search and starting point determination.

The above be only the preferred embodiment of the present invention it should be pointed out that：Ordinary skill people for the art For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should It is considered as protection scope of the present invention.

Claims (1)

1. a kind of combined synchronization for ofdm system and frequency deviation estimating method are it is characterised in that comprise the steps：

(1) with receipt signal ith sample point in observation window as starting point, two continuous OFDM symbol m-1, the sampling of m are extracted The initial value of point data, wherein i is 0, i ∈ Ω, and Ω is the number of sampled point in observation window；The initial value of m is 1, m ∈ Γ, Γ Number for OFDM symbol in receipt signal；

(2) set the number of m-th OFDM symbol cyclic prefix samples point as G, the sub-carrier number of ofdm system is N, successively by m Sampled point in individual OFDM symbol Cyclic Prefix with its at a distance of N number of sampled point and be located at m-th OFDM symbol afterbody sampled point Carry out autocorrelation operation, then G obtained autocorrelation value is averaging；

(3) by the sampled point in m-th OFDM symbol Cyclic Prefix and the m-1 OFDM symbol with it at a distance of N number of sampled point Sampled point carry out autocorrelation operation, and G obtained autocorrelation value is averaging；

(4) results added that will obtain in step (2) and step (3)；

(5) m=m+2, repeat step (1) to step (4), and the described addition result of all steps (4) obtaining is averaging, from And obtain the unbiased estimator of unbiased estimator φ (i) to channel autocorroelation function and carrier wave frequency deviation

(6) i=i+1, to each sampled point of receipt signal in observation window successively execution step (1) to step (5), then watch window The interior maximum i corresponding to φ (i) is the sync bit of OFDM symbol, and this sync bit is correspondingPhase angle be carrier frequency Inclined estimated value.