CN106878225A - The method and device that a kind of device-fingerprint with channel separate - Google Patents

Abstract

The invention discloses a kind of method that device-fingerprint with channel separate, comprise the following steps：S1：By the antenna array receiver signal of receiver；S2：The docking collection of letters number carries out channel estimation, makes k=1；S3：Autocorrelation matrix R (k) is calculated to k-th channel estimation results of carrier wave；S4：K-th autocorrelation matrix R (k) of carrier wave is carried out increasing order treatment, k-th increasing order autocorrelation matrix of carrier wave is obtainedS5：It is rightCalculate multipath signal amplitude；S6：Judge whether k is equal to N, N is the sum of carrier wave：If equal to, then carry out step S7；If it is not, then making k=k+1, step S3 is then return to；S7：Result of calculation to step S6 is ranked up, and selects path, obtains device-fingerprint.The invention also discloses the device that a kind of device-fingerprint with channel separate.The present invention can be widely applied in TDD, FDD communication system, it is also possible to apply in high-speed mobile scene.

Description

The method and device that a kind of device-fingerprint with channel separate

Technical field

The present invention relates to the communications field, the method and device that more particularly to a kind of device-fingerprint with channel separate.

Background technology

With the development of the communication technology, radio communication is widely used.With the hair of the new generation of wireless communication technology Exhibition, the extensive antenna technology that will particularly come into operation greatly increases available antenna amount.The increasing of antenna amount Plus, the handling capacity of wireless communication system is largely increased, while also for complicated signal processing technology is provided firmly Part platform.The conventional Array Signal Processing algorithm based on MUSIC and ESPRIT, antenna amount is the bottleneck of algorithm application.Greatly The application of scale antenna technology provides advantage for Array Signal Processing.

Device-fingerprint is primarily referred to as the comprehensive physical feature of the artificial circuit part (including antenna) of equipment.Its major embodiment Distortion is being caused to signal, especially in broadband system, the unevenness of gain in frequency band is being caused.Device-fingerprint is in communication equipment Between often it is different, so, device-fingerprint can influence the degree of accuracy of channel estimation, so as to reduce systematic function.It is being based on In the safety of physical layer of key generation, the influence of both sides device-fingerprint difference is reduced, the concordance rate to improving key has important shadow Ring.On the other hand, device-fingerprint is the feature on a kind of hardware, it is difficult to replicates or imitates, can be as the hand of equipment identification Section.So, the estimation to device-fingerprint is improving the channel reciprocity of TDD system, improves throughput of system and system safety The aspects such as property, there is important value.

For Wireless Telecom Equipment in itself, in the final tache of equipment production process, hardware calibration can be carried out to equipment, is made Equipment gain size, gain flatness, it is optimal with indexs such as outer energy leakages.But either built using during separating Hardware circuit even integrated circuit, because equipment preparation process has uncertain factor in itself, equipment Alignment cannot reach To optimal, so difference that also cannot be between abatement apparatus.And, aging with equipment, the temperature of use environment, humidity Change, this device-fingerprint can also change.So, in order to realize more preferable communication reliability and validity, it is necessary to pair set Standby fingerprint is extracted.

Calibration program after equipment investment use, is broadly divided into two kinds, and a kind of is the calibration method based on hardware circuit, by In the extra calibration circuit of needs, it is difficult to applied in terminal device, mobile device；One kind is using up-link and descending chain The method that the channel reciprocity on road is calibrated, this method passes through the approach such as signal transacting, signal precoding or feedback of the information, Interactive channel information is calculated device-fingerprint.

In the signal that receiver is received, device-fingerprint is combined together with channel response, can be according to wireless channel Reciprocity obtains device-fingerprint.For TDD system, this method is limited to coherence time, the information feedback overhead of communication.Example Such as, in patent《CN 101015179 B》In refer to using " implicit channel feedback " and to reduce feedback bandwidth.In patent 《CN105052176A》In refer to " using uplink/downlink channel reciprocity at many days with multi-user transmission For the system and method for RF calibration in linear system system ".In patent《CN102158272B》In, one kind is disclosed " according to described Downlink channel condition information and the uplink channel status information calculate the corresponding RF calibration factor of the first network side gusset " Realize the method that device-fingerprint is extracted.The method that above-mentioned patent is referred to using channel reciprocity, it is necessary to realize system calibration.Not Meet the scene of channel reciprocity, such as system translational speed is fast, and in the case that coherence time is short, the implementation result of scheme can be received To influence.Extra feedback of the information, can increase the expense of communication system.It is above-mentioned because communications band is different for FDD system Directly application is difficult to using the method for channel reciprocity.

The content of the invention

Goal of the invention：It is an object of the invention to provide a kind of device-fingerprint that can solve the problem that defect present in prior art The method and device for channel separate.

Technical scheme：To reach this purpose, the present invention uses following technical scheme：

The method that device-fingerprint of the present invention with channel separate, comprises the following steps：

S1：By the antenna array receiver signal of receiver；

S2：The docking collection of letters number carries out channel estimation, makes k=1；

S3：Autocorrelation matrix R (k) is calculated to k-th channel estimation results of carrier wave；

S4：K-th autocorrelation matrix R (k) of carrier wave is carried out increasing order treatment, k-th increasing order auto-correlation of carrier wave is obtained Matrix

S5：It is rightCalculate multipath signal amplitude；

S6：Judge whether k is equal to N, N is the sum of carrier wave：If equal to, then carry out step S7；If it is not, then making K=k+1, is then return to step S3；

S7：Result of calculation to step S6 is ranked up, and selects path, obtains device-fingerprint.

Further, in the step S1 antenna array receiver signal for pilot signal, insertion pilot tone signal or its His fixed signal.

Further, the channel estimation results of the carrier wave in the step S3 are channel estimation results, the part of single carrier wave Appointing in the channel estimation results, the channel estimation results of single symbol, the channel estimation results of multiple symbols after carrier wave merging Meaning is a kind of.

Further, in the step S4, adopt carries out increasing order treatment with the following method to autocorrelation matrix R (k)：

S4.1：The aerial array of receiver is divided into P submatrix, M is the antenna amount of receiving array, and L is number of paths；

S4.2：Calculate the autocorrelation matrix of each submatrix；

S4.3：The P average of the autocorrelation matrix of submatrix is calculated, k-th increasing order autocorrelation matrix of carrier wave is obtained As shown in formula (1)：

In formula (1), R_k,pIt is k-th autocorrelation matrix of p-th submatrix of carrier wave.

Further, the submatrix number P in the step S4.1 is more than multipath number.

Further, in the step S4.1, P=L+1, L are number of paths.The value of so selection P can release multipath letter Number correlation, it is also possible to the order of cross-correlation matrix is increased into P.

Further, in the step S4.1, P<L+1, L are number of paths.So can apply to antenna amount less Situation, the quantity of P can be selected according to path energy.

Further, the computational methods of multipath signal amplitude are comprised the following steps in the step S5：

S5.1：To k-th increasing order autocorrelation matrix of carrier waveFeature decomposition is carried out, is obtainedFeature decomposition Shown in form such as formula (2)：

In formula (2), λ_iIt is characterized value, U_S=[e₁ e₂ … e_L] it is λ₁ArriveIn the larger corresponding spy of L characteristic value Levy the signal subspace of vector, Σ_SIt is λ₁ArriveIn larger L eigenvalue cluster into diagonal matrix,It is λ₁ArriveIn less M₀Noise of the corresponding characteristic vector of the individual characteristic values of-L Space, σ²It is the power of white Gaussian noise, M₀It is the antenna amount in submatrix, L is number of paths；

S5.2：According to ESPRIT algorithms, a unique and nonsingular matrix T is solved, matrix T meets relational expression U_S= AT, A are array response matrix；

S5.3：It is calculated autocorrelation matrixAs shown in formula (3)：

In formula (3), I is unit matrix；

S5.4：It is rightDiagonal entry make even root, obtain k-th multipath amplitude of subcarrier.

Further, the path of the Path selection ceiling capacity in the step S7.

The device that device-fingerprint of the present invention with channel separate, including：

Array signal receiver module：Received for signal；

Channel estimation module：Channel estimation is carried out for docking the collection of letters number, device-fingerprint is obtained and is superimposed with channel Signal；

Autocorrelation matrix processing module：For calculating autocorrelation matrix, autocorrelation matrix is carried out to increase order treatment；

Multipath signal computing module：For calculating multipath signal amplitude；

Device-fingerprint extraction module：For being recombinated to the signal for obtaining, device-fingerprint is obtained.

Beneficial effect：Compared with prior art, the present invention has following beneficial effect：

1) present invention does not rely on channel reciprocity, in can be widely applied to TDD, FDD communication system.Prior art according to Rely channel reciprocity, in the case where channel reciprocity weakens or ungratified scene is looked into, performance will be greatly reduced or can not use, Rely on the technology of channel reciprocity, it is difficult to apply in FDD scenes.

2) processed present invention only requires to a traffic symbols, obtain device-fingerprint, do not rely on multiple symbols Treatment, can apply in high-speed mobile scene.Under high-speed mobile scene, coherence time is short, channel variation, in adjacent-symbol Between, channel has occurred and that and significantly change, and is difficult to apply in high-speed mobile scene using the method for multiple Symbol processings.

Brief description of the drawings

Fig. 1 is the structured flowchart of the device of the specific embodiment of the invention；

Fig. 2 is the structural representation of the communication system of the specific embodiment of the invention.

Specific embodiment

Technical scheme is further introduced with reference to the accompanying drawings and detailed description.

Assuming that a communication system is made up of a base station (BS) and a user equipment (UE), as shown in Fig. 2 the receipts of BS The transceiver that sender has M root antennas, UE has 1 antenna, and N is passed through between BS and UE_CIndividual subcarrier is communicated.UE By pilot sequence modulates to N_COn individual subcarrier, launched by hardware circuit.

This specific embodiment discloses a kind of method that device-fingerprint with channel separate, and comprises the following steps：

S1：By the antenna array receiver signal of receiver；

S2：The docking collection of letters number carries out channel estimation, makes k=1；

S3：Autocorrelation matrix R (k) is calculated to k-th channel estimation results of carrier wave；

S4：K-th autocorrelation matrix R (k) of carrier wave is carried out to increase order treatment, using it is preceding to space smoothing algorithm, it is front and rear To space smoothing algorithm or other methods that can reach the purpose for increasing autocorrelation matrix order, k-th increasing of carrier wave is obtained Order autocorrelation matrix

S5：It is rightMultipath is calculated using TLS-ESPRIT algorithms or other methods that can be calculated carrier amplitude Signal amplitude；

S6：Judge whether k is equal to N, N is the sum of carrier wave：If equal to, then carry out step S7；If it is not, then making K=k+1, is then return to step S3；

S7：Result of calculation to step S6 is ranked up, selection amplitude maximum path, obtains device-fingerprint.

In step S4, adopt carries out increasing order treatment with the following method to autocorrelation matrix R (k)：

S4.1：The aerial array of receiver is divided into P submatrix, M is the antenna amount of receiving array, and L is number of paths；

S4.2：Calculate the autocorrelation matrix of each submatrix；

S4.3：The P average of the autocorrelation matrix of submatrix is calculated, k-th increasing order autocorrelation matrix of carrier wave is obtainedAs shown in formula (1)：

In formula (1), R_k,pIt is k-th autocorrelation matrix of p-th submatrix of carrier wave.

The computational methods of multipath signal amplitude are comprised the following steps in step S5：

S5.1：To k-th increasing order autocorrelation matrix of carrier waveFeature decomposition is carried out, is obtainedFeature decomposition Shown in form such as formula (2)：

In formula (2), λ_iIt is characterized value, U_S=[e₁ e₂ … e_L] it is λ₁ArriveIn the larger corresponding spy of L characteristic value Levy the signal subspace of vector, Σ_SIt is λ₁ArriveIn larger L eigenvalue cluster into diagonal matrix,It is λ₁ArriveIn less M₀Noise of the corresponding characteristic vector of the individual characteristic values of-L Space, σ²It is the power of white Gaussian noise, M₀It is the antenna amount in submatrix, L is number of paths；Signal subspace U_SWith noise Subspace U_NIt is orthogonal, span { U_SRepresent by signal subspace U_SThe space opened, span { A } represents that array response matrix is opened into Space, the presence relational expression span { U of two sub-spaces_S}=span { A }；

S5.2：According to ESPRIT algorithms, a unique and nonsingular matrix T is solved, matrix T meets relational expression U_S= AT, A are array response matrix；

S5.3：It is calculated autocorrelation matrixAs shown in formula (3)：

In formula (3), I is unit matrix；

S5.4：It is rightDiagonal entry make even root, obtain k-th multipath amplitude of subcarrier.

Step S2 can also be omitted.

Step S3 extends also to multiple symbols or multiple carrier waves, or multiple symbols and multiple carrier waves.

Amplitude maximum path can not also be selected in step S7, but according to multipath amplitude size, selects first three larger Amplitude calculate device-fingerprint, or other be based on result of calculation combinations and conversion.

In step S4.1, P can also take P=L+3, or be distributed according to multipath energy, and selection P is equal to other numerical value.

This method can expand to the mimo system of M transmitting antenna, M reception antenna, can be extended to other symbols In the system application of conjunction Multipath Transmission condition.

This specific embodiment also discloses the device that a kind of device-fingerprint with channel separate, as shown in figure 1, including：

Array signal receiver module 501：Received for signal；

Channel estimation module 502：Channel estimation is carried out for docking the collection of letters number, device-fingerprint is obtained and is superimposed upon one with channel The signal for rising；

Autocorrelation matrix processing module 503：For calculating autocorrelation matrix, autocorrelation matrix is carried out to increase order treatment；

Multipath signal computing module 504：For calculating multipath signal amplitude；

Device-fingerprint extraction module 505：For being recombinated to the signal for obtaining, device-fingerprint is obtained.

Claims (10)

1. a kind of method that device-fingerprint with channel separate, it is characterised in that：Comprise the following steps：

S1：By the antenna array receiver signal of receiver；

S2：The docking collection of letters number carries out channel estimation, makes k=1；

S3：Autocorrelation matrix R (k) is calculated to k-th channel estimation results of carrier wave；

S4：K-th autocorrelation matrix R (k) of carrier wave is carried out increasing order treatment, k-th increasing order autocorrelation matrix of carrier wave is obtained

S5：It is rightCalculate multipath signal amplitude；

S6：Judge whether k is equal to N, N is the sum of carrier wave：If equal to, then carry out step S7；If it is not, then making k=k + 1, it is then return to step S3；

S7：Result of calculation to step S6 is ranked up, and selects path, obtains device-fingerprint.

2. the method that device-fingerprint according to claim 1 with channel separate, it is characterised in that：Antenna in the step S1 The signal of array received is pilot signal, the signal or other fixed signals of insertion pilot tone.

3. the method that device-fingerprint according to claim 1 with channel separate, it is characterised in that：Load in the step S3 The channel estimation results of ripple are channel estimation results, the single symbol after channel estimation results, the partial carrier wave merging of single carrier wave Number channel estimation results, the channel estimation results of multiple symbol in any one.

4. the method that device-fingerprint according to claim 1 with channel separate, it is characterised in that：In the step S4, adopt Autocorrelation matrix R (k) is carried out with the following method increase order treatment：

S4.1：The aerial array of receiver is divided into P submatrix；

S4.2：Calculate the autocorrelation matrix of each submatrix；

S4.3：The P average of the autocorrelation matrix of submatrix is calculated, k-th increasing order autocorrelation matrix of carrier wave is obtainedSuch as formula (1) shown in：

$\stackrel{\&OverBar;}{R}\left(k\right)=\frac{1}{P}\underset{p=1}{\overset{P}{\&Sigma;}}\left(R_{k,p}\right)---\left(1\right)$

In formula (1), R_k,pIt is k-th autocorrelation matrix of p-th submatrix of carrier wave.

5. the method that device-fingerprint according to claim 4 with channel separate, it is characterised in that：In the step S4.1 Submatrix number P is more than multipath number.

6. the method that device-fingerprint according to claim 4 with channel separate, it is characterised in that：In the step S4.1, P =L+1, L are number of paths.

7. the method that device-fingerprint according to claim 4 with channel separate, it is characterised in that：In the step S4.1, P <L+1, L are number of paths.

8. the method that device-fingerprint according to claim 1 with channel separate, it is characterised in that：Multipath in the step S5 The computational methods of signal amplitude are comprised the following steps：

S5.1：To k-th increasing order autocorrelation matrix of carrier waveFeature decomposition is carried out, is obtainedFeature decomposition form such as Shown in formula (2)：

$\begin{array}{c}\stackrel{\&OverBar;}{R}\left(k\right)=\underset{i=1}{\overset{M_0}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_i{e}_i{e}_i^H=\underset{i=1}{\overset{L}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_i{e}_i{e}_i^H+\underset{i=1}{\overset{M_0}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}^2{e}_i{e}_i^H=\underset{i=1}{\overset{L}{\mathrm{\&Sigma;}}}\left({\mathrm{\&lambda;}}_i+{\mathrm{\&sigma;}}^2\right)e_i{e}_i^H+\underset{i=L+1}{\overset{M_0}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}^2{e}_i{e}_i^H\\ =U_S{\mathrm{\&Sigma;}}_S{U}_S^H+{\mathrm{\&sigma;}}^2{U}_N{U}_N^H\end{array}---\left(2\right)$

In formula (2), λ_iIt is characterized value, U_S=[e₁ e₂ … e_L] it is λ₁ArriveIn the larger corresponding characteristic vector of L characteristic value The signal subspace opened, Σ_SIt is λ₁ArriveIn larger L eigenvalue cluster into diagonal matrix,It is λ₁ArriveIn less M₀Noise of the corresponding characteristic vector of the individual characteristic values of-L Space, σ²It is the power of white Gaussian noise, M₀It is the antenna amount in submatrix, L is number of paths；

S5.2：According to ESPRIT algorithms, a unique and nonsingular matrix T is solved, matrix T meets relational expression U_S=AT, A It is array response matrix；

S5.3：It is calculated autocorrelation matrixAs shown in formula (3)：

${\stackrel{\&OverBar;}{R}}_S\left(k\right)=T({\mathrm{\&Sigma;}}_S-{\mathrm{\&sigma;}}^{2}I)T^H---\left(3\right)$

In formula (3), I is unit matrix；

S5.4：It is rightDiagonal entry make even root, obtain k-th multipath amplitude of subcarrier.

9. the method that device-fingerprint according to claim 1 with channel separate, it is characterised in that：Road in the step S7 Footpath selects the path of ceiling capacity.

10. the device that a kind of device-fingerprint with channel separate, it is characterised in that：Including：

Array signal receiver module：Received for signal；

Channel estimation module：Channel estimation is carried out for docking the collection of letters number, the letter that device-fingerprint is superimposed with channel is obtained Number；

Autocorrelation matrix processing module：For calculating autocorrelation matrix, autocorrelation matrix is carried out to increase order treatment；

Multipath signal computing module：For calculating multipath signal amplitude；

Device-fingerprint extraction module：For being recombinated to the signal for obtaining, device-fingerprint is obtained.