CN103490863B - Space-time code pattern blind-identification method based on partial sequence parameter detecting - Google Patents

Abstract

The present invention discloses a kind of space-time code pattern blind-identification method based on partial sequence parameter detecting.Mainly solve prior art high, the problem of identification of code type narrow range.Implementation step is: (1) extracts pattern set, obtains characteristic quantity set；(2) calculate incidence matrix, and utilize this matrix calculus characteristic quantity functional value vector；(3) utilize characteristic quantity functional value vector pre-estimation characteristic quantity, obtain new space-time code set；(4) the symbolic number vector of pattern in new space-time code set is write out；(5) obtain parameter estimation vector (6) and utilize step (4) and (5), obtain distance decision value vector；(7) pattern corresponding to element that in distance decision value vector, value is minimum is taken as judgement pattern.Instant invention overcomes prior art and cause greatly, due to time lag norm operand of being correlated with, the shortcoming that system complexity is high, increase the identification range of existing space-time code blind-identification method, can be used for the space-time code pattern identification in communication countermeasure.

Description

Space-time code pattern blind-identification method based on partial sequence parameter detecting

Technical field

The invention belongs to communication technical field, the Space-Time Block Coding further relating to signal detection technique field during sky is compiled Pattern blind-identification method, can be used for, in multiple-input and multiple-output mimo system, Space-Time Block Coding being carried out blind recognition.

Background technology

Mimo system is the key technology of next generation wireless communication, and space-time code is the important composition portion of mimo system Point.The blind recognition of space-time code is the field in the urgent need to research, the communication countermeasure field, and it can be mimo system pair Anti-technology provides basis and technical support, has important theory significance and using value, has caused the concern of academia.

The blind recognition of space-time code is an emerging problem, and existing algorithm is divided into maximum likelihood algorithm and time lag phase Close algorithm.Maximum likelihood algorithm construction of function is simple, and computation complexity is low, but it is for block length and packet The pattern None-identified that symbolic number is identical；The discernible pattern of time lag related algorithm is more, but its computation complexity along with Sampling number exponentially increases, and is dfficult to apply in practice detect in real time.

Document [1V.Choqueuse, M.Marazin et al., Blind recognition of linear space time block codes:A likelihoodbased approach.IEEE Trans.Signal Processing,58(3),2010, 1290-1299] in propose code parameter detecting algorithm belong to maximum likelihood algorithm.According to maximum-likelihood criterion to institute There is Candidate Set pattern, construct only relevant with coding parameter likelihood function, by comparing the likelihood letter of different coding pattern Number, makes judgement, and then judges pattern coding parameter.The method decision function simple structure, computation complexity Low, it is widely used in MIMO blind recognition system.But the method carries out MIMO in engineering practice During detection, the deficiency existed is: to multiple have same packets length and often packet internal symbol number coding mode cannot Identify.

Document [2V.Choqueuse, K.Yao et al., Blind recognition of linear space time block codes. IEEE Int.Conf.Acoust.Speech Signal Process, 2008,2833-2836] the middle Decision Classfication inspection proposed Method of determining and calculating belongs to time lag related algorithm.It according to the correlation matrix of different Space-Time Block Codings under different delay The diversity of Frobenius norm, uses and contrasts step by step, it is achieved the blind recognition to Space-Time Block Coding.Due to the method Discernible pattern is relatively wide, and the most superior to the detection performance of orthogonal space time packet, therefore examines at MIMO Survey have also been obtained certain application.But the method there is also a lot of deficiencies in MIMO blind recognition system: main Showing and cannot be distinguished by the multiple pattern with identical F Norm Solution, computation complexity becomes along with sampling time length Geometry multiple increases.

Summary of the invention

Present invention aims to the deficiency of above-mentioned prior art, propose a kind of based on partial sequence parameter detecting Space-time code pattern blind-identification method, to improve space-time code identification of code type scope, reduces the complexity calculated.

Realizing the object of the invention ground technical thought is: by using characteristic quantity pre-estimation technology, enter multipath reception signal Row grouping feature amount pre-estimation, utilizes the grouping feature amount estimated, reduces space-time code set；Followed by partial order Row code parameter detecting technology, carries out partial sequence code parameter detecting, utilizes the partial sequence code parameter detected reducing Space-time code set in find judgement pattern.Concrete scheme comprises the steps:

1) receiving terminal receives the signal sequence of a length of N that transmitting terminal sends by r root reception antenna, obtains Reception signal matrix R' of r × N, wherein N >=64, r >=2；

2) utilize the space-time code that be there is a need to identify, form pattern set omega, take the class symbol of every kind of pattern in Ω (s, k) constitutive characteristic duration set (U, V), note i-th is combined as (s in the combination of number s and block length k_i,k_i), I=1,2... Ζ, Ζ are characterized number of combinations in duration set (U, V)；

3) real part and the imaginary part parallel connection of signal matrix R' will be received, it is thus achieved that incidence matrixI.e.

$\tilde{R}=\left[\begin{array}{c}\mathrm{Re}\left(R^{\′}\right)\\ \mathrm{Im}\left(R^{\′}\right)\end{array}\right]$

Wherein Re () represents that the computing for the treatment of excess syndrome portion, Im () expression take imaginary-part operation；

4) characteristic quantity functional value is calculated:

4a) (s is combined for the ith feature amount in characteristic quantity set (U, V)_i,k_i), structure packet correlation matrix R_i:

$R_i=\left[\begin{array}{cccc}\tilde{R}\left(1\right)& \tilde{R}(k_i+1)& ...& \tilde{R}\left(\right(\frac{N}{k_i}-1)k_i+1)\\ \tilde{R}\left(2\right)& \tilde{R}(k_i+2)& ...& \tilde{R}\left(\right(\frac{N}{k_i}-1)k_i+2)\\ \·& \·& & \·\\ \·& \·& & \·\\ \·& \·& & \·\\ \tilde{R}\left(k_i\right)& \tilde{R}\left(2k_i\right)& ...& \tilde{R}\left(\right(\frac{N}{k_i})k_i)\end{array}\right]$

WhereinRepresent incidence matrixτ row, τ=1,2...N；

4b) calculate packet correlation matrix R_iPacket covariance matrix: C_i=E [R_iR_i ^T], wherein the phase is asked in E [] expression Hope computing, ()^TRepresent transposition computing；

4c) to packet covariance matrix C_iDo Eigenvalues Decomposition, the eigenvalue obtained is arranged in descending order, constitutive characteristic Value vectorWherein, ρ_ηFor packet covariance matrix C_iEigenvalue, η=1,2 ... 2rk_i, Take feature value vectorFront 2s_iIndividual eigenvalue constitutes validity feature value vectorNoise is constituted with remaining eigenvalue Feature value vectorI.e.

${\stackrel{\→}{\mathrm{\λ}}}_{i1}=\[{\mathrm{\ρ}}_1,{\mathrm{\ρ}}_2,...,{\mathrm{\ρ}}_{2s_i}\],{\stackrel{\→}{\mathrm{\λ}}}_{i2}=\[{\mathrm{\ρ}}_{2s_i+1},{\mathrm{\ρ}}_{2s_i+2},...,{\mathrm{\ρ}}_{2{\mathrm{rk}}_i}\];$

4d) according to step 4c), obtain characteristic quantity combination (s_i,k_i) characteristic of correspondence flow function value M (s_i,k_i):

Wherein ∏ () represents that vector element connects multiplication；

5) pre-estimation characteristic quantity, obtain new space-time code set Ω ':

5a) every kind of combination in characteristic quantity set (U, V) is repeated step 4, obtain every kind of combination characteristic of correspondence amount Functional value, composition characteristic flow function value vector: Φ=[M (s₁,k₁),M(s₂,k₂)…M(s_m,k_m) ...], wherein, M=1,2 ... Ζ；

5b) find out the element characteristic of correspondence amount combination that in characteristic quantity functional value vector Φ, numerical value is minimumObtain New space-time code set Ω '；

6) write out new space-time code set Ω ' in the symbolic number of the every string of encoder matrix of jth kind pattern, form symbol Number vector P_j, wherein j=1,2...T, T be new space-time code set Ω ' pattern number；

7) estimating part symbolic number:

7a) utilize in step (3)With the block length in step (5b)Formations split-phase pass matrix V:

$V=\left[\begin{array}{cccc}\tilde{R}\left(\mathrm{\β}\right)& \tilde{R}(\hat{k}+\mathrm{\β})& ...& \tilde{R}\left(\right(\frac{N}{\hat{k}}-1)\hat{k}+\mathrm{\β})\end{array}\right],$

Wherein, β be new space-time code set Ω ' in the row mark of encoder matrix of pattern,

7b) the part covariance matrix of calculating section correlation matrix V: D=E [VV^T], to part covariance matrix D Do Eigenvalues Decomposition, the eigenvalue obtained is arranged in descending order, constitute partial feature value vector Wherein ε_γFor the eigenvalue of part covariance matrix D,

7c) respectively willBring following formula into and calculate likelihood function value

$\begin{array}{c}L\left(n\right)=-n(4r-2n+1)-\frac{N}{2\hat{k}}\underset{\mathrm{\α}=1}{\overset{2n}{\Σ}}log\left({\mathrm{\ϵ}}_{\mathrm{\α}}\right)\\ -\frac{N}{\hat{k}}(r-n)log\left(\frac{1}{2(r-n)}\underset{\mathrm{\α}=2n+1}{\overset{2r}{\Σ}}{\mathrm{\ϵ}}_{\mathrm{\α}}\right)\end{array}$

And then obtain likelihood function value vector

7d) taking the n that in likelihood function value vector L, numerical value greatest member is corresponding is partial symbols number；

8) computed range decision value vector:

8a) repeat step (7), estimate the partial symbols number of the every string of encoder matrix, obtain parameter estimation vector

Ψ=and [n (1), n (2) ... n (β) ...], wherein n (β) is the partial symbols number of encoder matrix β row；

8b) utilize the symbolic number vector P in step (6)_j, calculate symbolic number vector P_jDistance with parameter estimation vector Ψ Decision value θ_j=(P_j-Ψ)²；

8c) repeat step (8b), obtain new space-time code set Ω ' in the distance of every kind of pattern and parameter estimation vector Ψ sentence Certainly value, constitutes distance decision value vector Π=[θ₁,θ₂...θ_T]；

9) taking the pattern that in distance decision value vector Π, the element of numerical value minimum is corresponding is judgement pattern, completes space-time code mould Formula blind recognition.

The present invention compared with prior art has the advantage that

First, due to the fact that and have employed partial sequence parameter detecting technology, it is estimated that space-time code encoder matrix is every The coding symbol number of string, and identify transmission pattern by the symbolic number of every string, overcome time lag in prior art Related algorithm can not identify that having identical time lag is correlated with the pattern of norm, and maximum likelihood function algorithm can not identify have phase With block length and the deficiency of class symbol number type so that invention increases the identification range of space-time code.

Second, due to the fact that the partial sequence parameter detecting technology that have employed, therefore the partial sequence of estimation can be utilized to join Number directly adjudicates pattern, because of norm of being correlated with without the time lag calculating reception signal, overcomes in time lag related algorithm Norm calculation amount of being correlated with the time lag of space-time code is big, the shortcoming that the system implementation complexity that causes is high, makes the present invention realize Complexity substantially reduce.

Accompanying drawing explanation

Fig. 1 is the system block diagram that the present invention uses；

Fig. 2 is the flowchart of the present invention；

Fig. 3 is with present invention recognition correct rate figure under different parameters；

Fig. 4 is the recognition correct rate comparison diagram of the present invention and existing two kinds of blind-identification methods.

Detailed description of the invention

With reference to Fig. 1, the system that the present invention uses includes: t root launches antenna, and r root reception antenna, modulation system is 4QAM.At transmitting terminal, after serial transmission sequence is space-time encoded, be converted to transmitted in parallel sequence, then by Parallel Sequence Send after modulation.At receiving terminal, reception signal matrix is R':R'=HX+B, wherein, t >=2, r ＞ t, H is the channel matrix obeying multiple Gauss distribution of element independence, and X is the information sequence launched, and B is Gauss white noise Sound matrix, t=3 in this example.

The present invention is exactly that blind recognition goes out the space-time encoding modes that transmitting terminal uses according to receiving signal matrix R'.

With reference to Fig. 2, the present invention to implement step as follows:

Step 1, receiving terminal receives the signal sequence of a length of N that transmitting terminal sends by r root reception antenna, To reception signal matrix R' of r × N, N >=64, N=1024 or N=512, r=8 or 6 in this example.

Step 2, acquisition characteristic quantity set (U, V):

2a) utilizing the space-time code that be there is a need to identify, form pattern set omega, Ω includes orthogonal space time packet, Quasi-orthogonal space time block code and non-orthogonal space-time block.In this example, pattern set omega is: Ω =BALST (3,1), Tarokh-OSTBC(4,8),Ganesan1-OSTBC(3,4),Ganesan2-OSTBC(3,4),Tarokh-OSTBC( 3,4), Tarokh-OSTBC (4,4) },

Wherein, BLAST is hierarchical space-time code, and OSTBC is orthogonal space time packet, the bracket after space-time code Content represents group separator count s and the combination of block length k of pattern；

(s, k) constitutive characteristic duration set (U, V) gather in this example 2b) to take all of characteristic quantity combination in Ω (U, V) is: (U, V)={ (3,1), (4,8), (3,4), (4,4) }, and in note set, i-th is combined as (s_i,k_i), i=1,2... Ζ, Ζ is characterized number of combinations in duration set (U, V), Z=4 in this example, and each combines (s_i,k_i) corresponding empty time A kind of pattern in code set omega or multiple pattern.

Step 3, by parallel for the real part and imaginary part receiving signal matrix R', it is thus achieved that incidence matrix

$\tilde{R}=\left[\begin{array}{c}\mathrm{Re}\left(R^{\′}\right)\\ \mathrm{Im}\left(R^{\′}\right)\end{array}\right]$

Wherein Re () represents that the computing for the treatment of excess syndrome portion, Im () expression take imaginary-part operation.

Step 4, calculating characteristic quantity functional value:

4a) (s is combined for the ith feature amount in characteristic quantity set (U, V)_i,k_i), structure packet correlation matrix R_i:

$R_i=\left[\begin{array}{cccc}\tilde{R}\left(1\right)& \tilde{R}(k_i+1)& ...& \tilde{R}\left(\right(\frac{N}{k_i}-1)k_i+1)\\ \tilde{R}\left(2\right)& \tilde{R}(k_i+2)& ...& \tilde{R}\left(\right(\frac{N}{k_i}-1)k_i+2)\\ \·& \·& & \·\\ \·& \·& & \·\\ \·& \·& & \·\\ \tilde{R}\left(k_i\right)& \tilde{R}\left(2k_i\right)& ...& \tilde{R}\left(\right(\frac{N}{k_i})k_i)\end{array}\right],$

WhereinRepresent incidence matrixτ row, τ=1,2...N；

4b) calculate packet correlation matrix R_iPacket covariance matrix: C_i=E [R_iR_i ^T], wherein the phase is asked in E [] expression Hope computing, ()^TRepresent transposition computing；

4c) use orthogonal diagonal factorization method to packet covariance matrix C_iDo Eigenvalues Decomposition, i.e. first at packet covariance Matrix C_iBoth sides are multiplied by orthogonal matrix Q and transposed matrix thereof respectively, obtain eigenvalue diagonal matrix Δ=Q^TC_iQ；Again Packet covariance matrix C is extracted from eigenvalue diagonal matrix Δ_iEigenvalue；

4d) eigenvalue obtained is arranged in descending order, constitutive characteristic value vectorWherein, ρ_η For packet covariance matrix C_iEigenvalue, η=1,2 ... 2rk_i, take feature value vectorFront 2s_iIndividual eigenvalue structure Become validity feature value vectorNoise characteristic value vector is constituted with remaining eigenvalueThat is:

${\stackrel{\→}{\mathrm{\λ}}}_{i1}=\[{\mathrm{\ρ}}_1,{\mathrm{\ρ}}_2,...,{\mathrm{\ρ}}_{2s_i}\],$

${\stackrel{\→}{\mathrm{\λ}}}_{i2}=\[{\mathrm{\ρ}}_{2s_i+1},{\mathrm{\ρ}}_{2s_i+2},...,{\mathrm{\ρ}}_{2{\mathrm{rk}}_i}\];$

4e) according to step 4d), obtain characteristic quantity combination (s_i,k_i) characteristic of correspondence flow function value M (s_i,k_i):

Wherein ∏ () represents that vector element connects multiplication.

Step 5, pre-estimation characteristic quantity, obtain new space-time code set Ω '.

5a) every kind of combination in characteristic quantity set (U, V) is repeated step 4, obtain every kind of combination characteristic of correspondence amount Functional value, composition characteristic flow function value vector: Φ=[M (s₁,k₁),M(s₂,k₂)…M(s_m,k_m) ...], wherein, M=1,2 ... Ζ；

5b) find out the element characteristic of correspondence amount combination that in characteristic quantity functional value vector Φ, numerical value is minimum Obtain new space-time code set Ω ', it by group separator count all in space-time code set Ω isAnd block length ForPattern composition set, this set includes one or more patterns, wherein

Step 6, write out new space-time code set Ω ' in the symbolic number of the every string of encoder matrix of jth kind pattern, composition Symbolic number vector P_j, the encoder matrix of this jth kind pattern and new space-time code set Ω ' in the encoder matrix of other patterns Having identical columns, but the coding symbol number that every string contains is not quite similar, wherein j=1, when 2...T, T are new sky Code set omega ' pattern number.

Step 7, estimating part symbolic number.

7a) utilize the incidence matrix in step (3)With the block length in step (5b)Formations divides correlation matrix V:

$V=\left[\begin{array}{cccc}\tilde{R}\left(\mathrm{\β}\right)& \tilde{R}(\hat{k}+\mathrm{\β})& ...& \tilde{R}\left(\right(\frac{N}{\hat{k}}-1)\hat{k}+\mathrm{\β})\end{array}\right],$

Wherein, β be new space-time code set Ω ' in the row mark of encoder matrix of pattern,

7b) the part covariance matrix of calculating section correlation matrix V: D=E [VV^T], use orthogonal diagonal factorization Part covariance matrix D is done Eigenvalues Decomposition by method, is arranged in descending order by the eigenvalue obtained, and constitutes partial feature value VectorWherein ε_γFor the eigenvalue of part covariance matrix D,

7c) respectively willBring following formula into and calculate likelihood function value

$\begin{array}{c}L\left(n\right)=-n(4r-2n+1)-\frac{N}{2\hat{k}}\underset{\mathrm{\α}=1}{\overset{2n}{\Σ}}log\left({\mathrm{\ϵ}}_{\mathrm{\α}}\right)\\ -\frac{N}{\hat{k}}(r-n)log\left(\frac{1}{2(r-n)}\underset{\mathrm{\α}=2n+1}{\overset{2r}{\Σ}}{\mathrm{\ϵ}}_{\mathrm{\α}}\right)\end{array},$

And then obtain likelihood function value vector

7d) take the n that in likelihood function value vector L, numerical value greatest member is corresponding, as partial symbols number.

Step 8, computed range decision value vector.

8a) repeat step (7), estimate the partial symbols number of the every string of encoder matrix, obtain parameter estimation vectorial:

Ψ=and [n (1), n (2) ... n (β) ...], wherein n (β) is the partial symbols number of encoder matrix β row；

8b) utilize the symbolic number vector P in step (6)_j, calculate symbolic number vector P_jDistance with parameter estimation vector Ψ Decision value θ_j=(P_j-Ψ)²；

8c) repeat step (8b), obtain new space-time code set Ω ' in the distance of every kind of pattern and parameter estimation vector Ψ sentence Certainly value, constitutes distance decision value vector Π=[θ₁,θ₂...θ_T]。

Step 9, takes pattern corresponding to element that in distance decision value vector Π, numerical value is minimum as judgement pattern, completes Space-time code pattern blind recognition.

The effect of the present invention can be further described by following emulation.

Emulation 1: under three groups of different parameters, diplomatic copy to the identification of space-time code in space-time code set Ω is just being invented Really rate, simulation result such as Fig. 3.Wherein: it is 3 that the solid line in Fig. 3 represents transmitting antenna, and reception antenna is 8, sends out Sending the system identification accuracy of sequence length N=1024, it is 3 that the zone circle solid line in Fig. 3 represents transmitting antenna, receives Antenna is 6, sends the system identification accuracy of sequence length N=1024, and the dotted line in Fig. 3 represents transmitting antenna and is 3, reception antenna is 8, sends the system identification accuracy of sequence length N=512,

As can be seen from Figure 3: solid line, on zone circle solid line, illustrates can be carried by the quantity improving reception antenna The high present invention recognition correct rate to space-time code set Ω；Solid line is on dotted line, and illustrating to increase transmission sequence length can To improve the present invention recognition correct rate to space-time code set Ω.

Emulation 2: with existing two kinds of blind-identification methods, space-time code set Ω is identified by the present invention.

Existing two kinds of blind-identification methods are:

Document [1V.Choqueuse, M.Marazin et al., Blind recognition of linear space time block codes:A likelihoodbased approach.IEEE Trans.Signal Processing,58(3),2010, 1290-1299] the middle code parameter detecting algorithm proposed, it, constructs all Candidate Set patterns according to maximum-likelihood criterion Only relevant with coding parameter likelihood function, by comparing the likelihood function of different coding pattern, makes coding parameter Judge, and then draw judgement pattern, this emulation is abbreviated as a yard parameter detection method.

Document [2V.Choqueuse, K.Yao et al., Blind recognition of linear space time block Codes.IEEE Int.Conf.Acoust.Speech Signal Process, 2008,2833-2836] the middle decision-making proposed Classification and Detection algorithm, it is according to Frobenius norm under different delay of the correlation matrix of different Space-Time Block Codings Diversity, uses and contrasts step by step, it is achieved the blind recognition to Space-Time Block Coding, is abbreviated as Decision Classfication inspection in this emulation Survey method.

If emulation SNR ranges is-5dB～10dB, emulates 1000 Monte Carlo Experiments every 1dB, cover every time Space-time code in space-time code set Ω is sent identification by special Carlow experiment successively, records the correct knowledge under each signal to noise ratio Other number of times, and then obtain the recognition correct rate under each signal to noise ratio, i.e. the ratio of recognizable code type duty time-code set omega, Simulation result such as Fig. 4.Wherein solid line represents the recognition correct rate of the present invention, and zone circle solid line represents code parameter detection method Recognition correct rate, dotted line represents the recognition correct rate of Decision Classfication detection method.

As can be seen from Figure 4: solid line, far away on zone circle solid line and dotted line, illustrates under same signal to noise ratio, this The bright recognition correct rate to space-time code set Ω is far above existing two kinds of blind-identification methods.

It can also be seen that from Fig. 4: zone circle solid line finally levels off to 4/6, dotted line finally levels off to 5/6, and solid line is Level off to 1 eventually, the 4/6 of the pattern duty time-code set omega that description code parameter detection method can identify, Decision Classfication The 5/6 of the pattern duty time-code set omega that detection method can identify, and the present invention can be by space-time code set Ω Pattern all identifies, i.e. the present invention can be than code parameter detection method and the more pattern of Decision Classfication detection method identification.

Claims (5)

1. a space-time code pattern blind-identification method based on partial sequence parameter detecting, comprises the steps:

1) receiving terminal receives the signal sequence of a length of N that transmitting terminal sends by r root reception antenna, obtains r × N Reception signal matrix R', wherein N >=64, r >=2；

2) utilize the space-time code that be there is a need to identify, form pattern set omega, take the class symbol of every kind of pattern in Ω (s, k) constitutive characteristic duration set (U, V), note i-th is combined as (s in the combination of number s and block length k_i,k_i) I=1,2...Z, Z are characterized number of combinations in duration set (U, V)；

3) real part and the imaginary part parallel connection of signal matrix R' will be received, it is thus achieved that incidence matrixI.e.

$\tilde{R}=\left[\begin{array}{c}\mathrm{Re}\left(R^{\′}\right)\\ \mathrm{Im}\left(R^{\′}\right)\end{array}\right]$

Wherein Re () represents that the computing for the treatment of excess syndrome portion, Im () expression take imaginary-part operation；

4) characteristic quantity functional value is calculated:

4a) (s is combined for the ith feature amount in characteristic quantity set (U, V)_i,k_i), structure packet correlation matrix R_i:

$R_i=\left[\begin{array}{cccc}\tilde{R}\left(1\right)& \tilde{R}(k_i+1)& ...& \tilde{R}\left(\right(\frac{N}{k_i}-1)k_i+1)\\ \tilde{R}\left(2\right)& \tilde{R}(k_i+2)& ...& \tilde{R}\left(\right(\frac{N}{k_i}-1)k_i+2)\\ \·& \·& & \·\\ \·& \·& ...& \·\\ \·& \·& & \·\\ \tilde{R}\left(\mathrm{\τ}\right)& \tilde{R}(k_i+\mathrm{\τ})& ...& \tilde{R}\left(\right(\frac{N}{k_i}-1)k_i+\mathrm{\τ})\\ \·& \·& & \·\\ \·& \·& ...& \·\\ \·& \·& & \·\\ \tilde{R}\left(k_i\right)& \tilde{R}\left(2k_i\right)& ...& \tilde{R}\left(\right(\frac{N}{k_i})k_i)\end{array}\right]$

WhereinRepresent incidence matrixτ row, τ=1,2...k_i, k_iRepresent block length；

4b) calculate packet correlation matrix R_iPacket covariance matrix: C_i=E [R_iR_i ^T], wherein the phase is asked in E [] expression Hope computing, ()^TRepresent transposition computing；

4c) use orthogonal diagonal factorization method to packet covariance matrix C_iDo Eigenvalues Decomposition, i.e. first at packet covariance Matrix C_iBoth sides are multiplied by orthogonal matrix Q and transposed matrix thereof respectively, obtain eigenvalue diagonal matrix Δ=Q^TC_iQ；Again Packet covariance matrix C is extracted from eigenvalue diagonal matrix Δ_iEigenvalue；The eigenvalue obtained is arranged in descending order Row, constitutive characteristic value vectorWherein, ρ_ηFor packet covariance matrix C_iEigenvalue, η=1,2 ... 2rk_i, take feature value vectorFront 2s_iIndividual eigenvalue constitutes validity feature value vectorWith remaining Eigenvalue constitutes noise characteristic value vectorI.e.

${\stackrel{\→}{\mathrm{\λ}}}_{i1}=\[{\mathrm{\ρ}}_1,{\mathrm{\ρ}}_2,...,{\mathrm{\ρ}}_{2s_i}\],{\stackrel{\→}{\mathrm{\λ}}}_{i2}=\[{\mathrm{\ρ}}_{2s_i+1},{\mathrm{\ρ}}_{2s_i+2},...,{\mathrm{\ρ}}_{2{\mathrm{rk}}_i}\];$

4d) according to step 4c), obtain characteristic quantity combination (s_i,k_i) characteristic of correspondence flow function value M (s_i,k_i):

Wherein Π () represents that vector element connects multiplication；

5) pre-estimation characteristic quantity, obtain new space-time code set Ω ':

5a) every kind of combination in characteristic quantity set (U, V) is repeated step 4, obtain every kind of combination characteristic of correspondence amount Functional value, composition characteristic flow function value vector: Φ=[M (s₁,k₁),M(s₂,k₂)···M(s_m,k_m)], wherein, M=1,2, Z；

5b) find out the element characteristic of correspondence amount combination that in characteristic quantity functional value vector Φ, numerical value is minimumObtain New space-time code set Ω '；

6) write out new space-time code set Ω ' in the symbolic number of the every string of encoder matrix of jth kind pattern, form symbol Number vector P_j, wherein j=1,2...T, T be new space-time code set Ω ' pattern number；

7) estimating part symbolic number:

7a) utilize in step (3)With the block length in step (5b)Formations split-phase pass matrix V:

$V=\left[\begin{array}{cccc}\tilde{R}\left(1\right)& \tilde{R}(\hat{k}+1)& ...& \tilde{R}\left(\right(\frac{N}{\hat{k}}-1)\hat{k}+)\\ \tilde{R}\left(2\right)& \tilde{R}(\hat{k}+2)& ...& \tilde{R}\left(\right(\frac{N}{\hat{k}}-1)\hat{k}+)\\ \·& \·& & \·\\ \·& \·& ...& \·\\ \·& \·& & \·\\ \tilde{R}\left(\mathrm{\β}\right)& \tilde{R}(\hat{k}+\mathrm{\β})& ...& \tilde{R}\left(\right(\frac{N}{\hat{k}}-1)\hat{k}+\mathrm{\β})\\ \·& \·& & \·\\ \·& \·& ...& \·\\ \·& \·& & \·\\ \tilde{R}(\hat{k})& \tilde{R}(2\hat{k})& ...& \tilde{R}\left(N\right)\end{array}\right];$

Wherein, β be new space-time code set Ω ' in the row mark of encoder matrix of pattern,

7b) the part covariance matrix of calculating section correlation matrix V: D=E [VV^T], to part covariance matrix D Do Eigenvalues Decomposition, the eigenvalue obtained is arranged in descending order, constitute partial feature value vector Wherein ε_γFor the eigenvalue of part covariance matrix D,

7c) respectively by n=1,Bring following formula into and calculate likelihood function value L (1),

$\begin{array}{c}L\left(n\right)=-n(4r-2n+1)-\frac{N}{2\hat{k}}\underset{\mathrm{\α}=1}{\overset{2n}{\Σ}}log\left({\mathrm{\ϵ}}_{\mathrm{\α}}\right)\\ -\frac{N}{\hat{k}}(r-n)log\left(\frac{1}{2(r-n)}\underset{\mathrm{\α}=2n+1}{\overset{2r}{\Σ}}{\mathrm{\ϵ}}_{\mathrm{\α}}\right)\end{array}$

And then obtain likelihood function value vector

7d) taking the n that in likelihood function value vector L, numerical value greatest member is corresponding is partial symbols number；

8) computed range decision value vector:

8a) repeat step (7), estimate the partial symbols number of the every string of encoder matrix, obtain parameter estimation vector

Ψ=and [n (1), n (2) ... n (β) ...], n (β) is the partial symbols number of encoder matrix β row；

8b) utilize the symbolic number vector P in step (6)_j, calculate symbolic number vector P_jDistance with parameter estimation vector Ψ Decision value θ_j=(P_j-Ψ)²；

8c) repeat step (8b), obtain new space-time code set Ω ' in the distance of every kind of pattern and parameter estimation vector Ψ sentence Certainly value, constitutes distance decision value vector Π=[θ₁,θ₂...θ_T]；

9) taking the pattern that in distance decision value vector Π, the element of numerical value minimum is corresponding is judgement pattern, completes space-time code mould Formula blind recognition.

Space-time code pattern blind-identification method based on partial sequence parameter detecting the most according to claim 1, It is characterized in that, space-time code set Ω in described step (2), including orthogonal space time packet, quasi-orthogonal space Time block code and non-orthogonal space-time block.

Space-time code pattern blind-identification method based on partial sequence parameter detecting the most according to claim 1, It is characterized in that, the combination (s in described step (2)_i,k_i), in its corresponding space-time code set Ω of each combination A kind of pattern or multiple pattern.

Space-time code pattern blind-identification method based on partial sequence parameter detecting the most according to claim 1, It is characterized in that, described step 5b) in new space-time code set Ω ', be by space-time code set Ω all points Group code number isAnd block length isPattern composition set, this set includes one or more patterns, Wherein

Space-time code pattern blind-identification method based on partial sequence parameter detecting the most according to claim 1, It is characterized in that, new space-time code set Ω in described step (6) ' in the encoder matrix of jth kind pattern, its with New space-time code set Ω ' in the encoder matrix of other patterns there is identical columns, but the coding that every string contains Symbolic number is not quite similar.