CN104865560A - UVM-based phased array radar digital beam former module verification method and verification platform thereof - Google Patents

Abstract

The invention relates to a UVM-based phased array radar digital beam former module verification method and a verification platform thereof. The verification method constructs a C function model in an excitation generator module by use of a DPI, realizes complex number operation of channel data, weight coefficients, correction coefficients and wave beam directional coefficients and completes conversion from floating numbers to fixed-point numbers for exciting into a DBF module. At the same time, the verification platform automatically acquires wave beam information after operation of the DBF module and performs automatic result comparison with a reference module, and correctly compared results are converted from the fixed-point numbers into floating numbers and then are written into files for subsequent processing. The advantages are as follows: reuse is realized, the maintenance is easy, the efficiency is high, the coverage is high, the code amount of the verification program of the DBF module and the workload of later board debugging can be greatly reduced, and the development efficiency and the quality of FPGA logic codes of the DBF module are improved.

Description

A kind of phased-array radar Digital Beamformer module verification method based on UVM and verification platform thereof

Technical field:

The present invention relates to digital beam froming (DBF, Digital beam forming) technology and FPGA (Field Programmable Gate Array) logic checking field, specifically a kind of reusable, easy care, efficiently, high covering, greatly can reduce the workload that the size of code of DBF module verification program and later stage plate are debugged, and improve the development efficiency of the FPGA logical code of DBF module and the phased-array radar Digital Beamformer module verification method based on UVM (Universal Verification Methodology) of quality and verification platform thereof.

Background technology:

DBF is the gordian technique of phased-array radar always, it belongs to Array Signal Processing category, take full advantage of the spatial information detected by array antenna, the performance of super-resolution and Sidelobe can be obtained easily, realize beam scanning, self calibration and Adaptive beamformer etc., there is the feature of dirigibility and real-time.DBF adopts digital processing method, for a direction incoming signal, compensate and that cause propagation difference cause phase differential different in locus due to sensor, realize in-phase stacking, thus the ceiling capacity realizing this direction receives, and completes the party's upwards Wave beam forming.DBF system is primarily of aerial array, receiver module, A/D module, Digital Beamformer, beam-controller and postsignal processor composition, and as shown in Figure 1, the digital beam froming for a direction will complete lower column operations:

$B_k\left(t\right)=f\left(t\right)\underset{n=0}{\overset{N-1}{\Σ}}\[\exp (j2\mathrm{\π}n\·\frac{dsin\mathrm{\α}}{\mathrm{\λ}})\·\exp (-j\frac{2\mathrm{\π}n}{N}\·{\mathrm{\Ω}}_k)\·C\left(n\right)\·W\]---\left(1\right)$

k＝0,1,…,K-1

In formula (1), f (t) is intended recipient information; K is wave beam number; D is array element distance, as shown in Figure 2; λ is for launching carrier wavelength; α is the incident angle of echo signal relative antenna front normal; N is the n-th passage; N is the array number of Vertical dimension arrangement on antenna array; W is weighting coefficient; C is correction parameter; Ω reflects beam position.

${\mathrm{\Ω}}_k=\frac{N\·d\·\sin \mathrm{\α}}{\mathrm{\λ}}---\left(2\right)$

Obviously, when beam position meets formula (2), the signal that this wave beam receives is maximum, and the beam signal amplitude in other directions declines.

Formula (1) can be reduced to

$B_k=\underset{n=0}{\overset{N-1}{\Σ}}X\left(n\right)\·S_k\left(n\right)\·C\left(n\right)\·W---\left(3\right)$

k＝0,1,…,K-1

In formula (3), X (n) is the input complex signal that the n-th passage comes; S _kfor beam position parameter.

As can be seen from above formula, Beam-former essence has been the function that input signal and corresponding weight coefficient complex multiplication add up, and in engineering, the general FPGA of employing realizes.Plate level debugging after traditional Digital Beamformer design cycle is divided into demand analysis, the design input of hardware description language, simulating, verifying, net table to download.Wherein, in the simulating, verifying stage owing to needing the test and excitation signal of generation comparatively complicated, and its excited data source mostly is the scale-of-two or sexadecimal fixed point vector quantization data that are formed through conversion by floating-point complex, the hardware description languages such as traditional Verilog or VHDL are utilized to be difficult to be realized.Therefore, slip-stick artist only carries out some simple simulating, verifyings in this stage.Simulating, verifying generally as the means of main guarantee designing quality, also can not can not use any verification methodology.But with the raising of DBF system scale and design complexities, only debugging can waste the plenty of time on phase plate rearward, and is difficult to positioning logic mistake, can not ensure that the module designed is through comprehensively checking, can not exist mistake under the condition on border.Therefore, how simply, generate efficiently the test and excitation meeting designing requirement, build the verification platform of reusable a, easy care, to improve the efficiency of simulating, verifying and comprehensive, in the design process of Digital Beamformer, seem particularly important.

UVM be a new generation based on the high level of authentication methodology of System Verilog language, Utility Engineers it can create easy exploiting, the Verification Components of reusable, robotization and verification platform.UVM, based on System Verilog language, has the characteristic of object based programming, and provides DPI (Direct Programming Interface) interface, can complete the slitless connection of Verification Components in UVM and C model easily.Utilize thought and the framework of UVM, and use C model to complete the generation of complicated test stimulus data, just can provide a kind of outstanding, comprehensive, efficient, reusable verification method for Digital Beamformer module.

Summary of the invention:

The technical problem to be solved in the present invention is, there is provided a kind of reusable, easy care, high-level efficiency, high covering, greatly can reduce the workload that the size of code of DBF module verification program and later stage plate are debugged, and improve the development efficiency of the FPGA logical code of DBF module and the phased-array radar Digital Beamformer module verification method based on UVM of quality and realize the verification platform of this verification method.

Technical solution of the present invention is, provides a kind of phased-array radar Digital Beamformer module verification method based on UVM comprising following sequential steps

1. in excitation generator block, define DPI interface and state the C function that channel data, weighting coefficient, correction coefficient and beam position coefficient generate;

2. C function is according to parameter configuration and algorithm requirement, generates corresponding channel data and coefficient data, and is converted to fixed-point number and passes to excitation generator;

3. the excited data packing that excitation generator block is produced by TLM interface by driver is driven into interface, and then is loaded in DBF module;

4. monitor module gathers the data that DBF module exports, reference model is sent to carry out the automatic comparison of result on the one hand, on the one hand the data of collection are converted in complex plane represent direction and range information by calling DPI interface, and writing in files, for next step data analysis.

Phased-array radar Digital Beamformer module verification method based on UVM of the present invention, wherein, in excitation generator, the production method of passage complex signal is as follows:

1. first, in excitation generator, state DPI interface, and need the passage complex signal called to produce C function:

Import"DPI-C"function void dbf_data_build(

input bit[31:0] ELE_NUM，

input bit[31:0] CH_NUM，

output bit[15:0] DATA_REAL，

output bit[15:0] DATA_IMAG)；

This C function completes the exchange of data with excitation generator by four parameters, be array number (ELE_NUM), port number (CH_NUM), data real part (DATA_REAL) and data imaginary part (DATA_IMAG) respectively.

2. in C function, according to the initial direction (DIR_START) of radar bearing scanning, the step-length (DIR_STEP) terminating direction (DIR_END) and direction change determines the scan angle theta of m range unit

θ＝(DIR_START+DIR_STEP×m)×π/180

And then determine the complex signal of the n-th array element at m range unit

$array\_data(n,m)=e^{j\frac{n\·d}{\mathrm{\λ}}\sin \left(\mathrm{\θ}\right)}$

Wherein, d is the spacing between bay, and λ is wavelength.

3. last, the real part of complex data and empty step are converted to respectively the fixed-point number of sixteen bit, input to excitation generator by DPI interface,

$DATA\_REAL=cos\left(sin\left(\mathrm{\θ}\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

$DATA\_IMAG=\sin \left(sin\left(\mathrm{\θ}\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

The data real part of generation and imaginary data are loaded on interface, to complete the excitation to channel data in beamformer module by excitation generator as required.

Phased-array radar Digital Beamformer module verification method based on UVM of the present invention, wherein, in excitation generator, the production method of weighting coefficient is as follows:

1. first, in excitation generator, state DPI interface, and need the weighting coefficient called to produce C function:

import"DPI-C"function void dbf_coef_build

(input bit[31:0] ELE_NUM，

input bit[31:0] CH_NUM，

input bit[31:0] MAX_BEAM_NUM，

input bit[31:0] BGROUP_NUM，

input bit[31:0] BCOMBIN_NUM，

output bit[15:0] COEF_REAL，

output bit[15:0] COEF_IMAG)；

This C function completes the exchange of data with excitation generator by seven parameters, be array number (ELE_NUM), port number (CH_NUM), maximum numbers of beams (MAX_BEAM_NUM), wave beam group number (BGROUP_NUM, beam collection number (BCOMBIN_NUM), coefficient real part (COEF_REAL) and coefficient imaginary part (COEF_IMAG) respectively.

2., in C function, the Initial Azimuth (RCV_BEAM_START) produced according to wave beam, orientation (RCV_BEAM_START) is stopped and step-length (RCV_BEAM_STEP) determines the scan angle theta of m wave beam ₀:

θ ₀＝(RCV_BEAM_START+RCV_BEAM_STEP×m)×π/180

And then determine the coefficient of the n-th array element at m wave beam

$array\_coef(n,m)=e^{-j\frac{n\·d}{\mathrm{\λ}}sin\left({\mathrm{\θ}}_0\right)}$

Wherein, d is the spacing between bay, and λ is wavelength.

3. last, the real part of coefficient and empty step are converted to respectively the fixed-point number of sixteen bit, input to excitation generator by DPI interface

$COEF\_REAL=cos\left(sin\left({\mathrm{\θ}}_0\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

$COEF\_IMAG=sin\left(sin\left({\mathrm{\θ}}_0\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

The weight coefficient real part of generation and imaginary data are loaded on interface, to complete the excitation to weight coefficient in beamformer module by excitation generator as required.

In order to realize the phased-array radar Digital Beamformer module verification method based on UVM, another technical solution of the present invention is, a kind of verification platform based on UVM verification environment is provided, this verification platform comprises test case library, UVM verification environment, data send proxy server, data receiver proxy server and scoring board, data send proxy server and data receiver proxy server is all positioned at UVM verification environment, data send proxy server and comprise the first monitor, data-coefficients generator, first excitation generator and the first driver, data receiver proxy server comprises beam data converter, second monitor, second driver and the second excitation generator, first monitor gathers the port data in DBF module by interface, first excitation generator one end is connected with data-coefficients generator by DPI interface, the first excitation generator other end is connected with first driver one end, the first driver other end is connected with interface one end, the interface other end and DBF model calling, second monitor is connected with beam data converter by DPI interface, second monitor gathers the port data in DBF module by interface, second driver one end and second encourages generator to be connected, the second driver other end is connected with interface, interface and DBF model calling, test case library sends proxy server with data simultaneously and data receiver proxy server is connected, UVM verification environment is connected with scoring board by TLM interface, scoring board comprises reference model and comparer, comparer one end is connected with reference model, the comparer other end is connected with the second monitor in UVM verification environment.

Compared with the verification method of existing phased-array radar Digital Beamformer module, beneficial effect of the present invention is: present invention employs verification methodology UVM, the checking structure of a stratification can be realized, can better simply transplanting the beamformer module of the different configuration of checking, and in excitation generator, the generation of test data is completed by the method calling C model, the test and excitation being driven into DBF module is made not only to have actual physical significance, and the test and excitation generated also possesses reusable, easy care, high covering, the features such as high-level efficiency, greatly can reduce the workload that the size of code of DBF module verification program and later stage plate are debugged, and improve development efficiency and the quality of the FPGA logical code of DBF module.Simultaneously, in the drive equally by the method for DPI interface interchange C model, easily the DBF sampled output signal can being converted to direction and range information in the complex plane be comparatively concerned about, being also convenient to the variation tendency carrying out observation beam signal by drawing directional diagram.

Compared with traditional UVM verification platform, in order to the Complex multiplication accumulating operation of adaptive Digital Beamformer module, the present invention calls DPI interface in excitation generator, C model is utilized to carry out floating point arithmetic, and complete the conversion that floating-point counts to fixed-point number, with realize reusable, easy care, efficiently, test and excitation signal method for building up easily.

Accompanying drawing illustrates:

Fig. 1 is typical DBF radar theory of constitution block diagram;

Fig. 2 is Wave beam forming electrical block diagram;

Fig. 3 is the structural representation of the verification platform of DBF module verification environment.

Specific embodiment:

Below in conjunction with the drawings and specific embodiments, a kind of phased-array radar Digital Beamformer module verification method based on UVM of the present invention and verification platform thereof are described further.

In this specific embodiment, a kind of phased-array radar Digital Beamformer module verification method based on UVM of the present invention comprises the following steps

1. in excitation generator block, define DPI interface and state the C function that channel data, weighting coefficient, correction coefficient and beam position coefficient generate;

2. C function is according to parameter configuration and algorithm requirement, generates corresponding channel data and coefficient data, and is converted to fixed-point number and passes to excitation generator;

3. the excited data packing that excitation generator block is produced by TLM interface by driver is driven into interface, and then is loaded in DBF module;

4. monitor module gathers the data that DBF module exports, reference model is sent to carry out the automatic comparison of result on the one hand, on the one hand the data of collection are converted in complex plane represent direction and range information by calling DPI interface, and writing in files, for next step data analysis.

In this specific embodiment, in excitation generator, the production method of passage complex signal is as follows:

1. first, in excitation generator, state DPI interface, and need the passage complex signal called to produce C function:

Import"DPI-C"function void dbf_data_build(

input bit[31:0] ELE_NUM，

input bit[31:0] CH_NUM，

output bit[15:0] DATA_REAL，

output bit[15:0] DATA_IMAG)；

This C function completes the exchange of data with excitation generator by four parameters, be array number (ELE_NUM), port number (CH_NUM), data real part (DATA_REAL) and data imaginary part (DATA_IMAG) respectively.

2. in C function, according to the initial direction (DIR_START) of radar bearing scanning, the step-length (DIR_STEP) terminating direction (DIR_END) and direction change determines the scan angle theta of m range unit

θ＝(DIR_START+DIR_STEP×m)×π/180

And then determine the complex signal of the n-th array element at m range unit

$array\_data(n,m)=e^{j\frac{n\·d}{\mathrm{\λ}}\sin \left(\mathrm{\θ}\right)}$

Wherein, d is the spacing between bay, and λ is wavelength.

3. last, the real part of complex data and empty step are converted to respectively the fixed-point number of sixteen bit, input to excitation generator by DPI interface,

$DATA\_REAL=cos\left(sin\left(\mathrm{\θ}\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

$DATA\_IMAG=sin\left(sin\left(\mathrm{\θ}\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

The data real part of generation and imaginary data are loaded on interface, to complete the excitation to channel data in beamformer module by excitation generator as required.

In this specific embodiment, in excitation generator, the production method of weighting coefficient is as follows:

1. first, in excitation generator, state DPI interface, and need the weighting coefficient called to produce C function:

import"DPI-C"function void dbf_coef_build

(input bit[31:0] ELE_NUM，

input bit[31:0] CH_NUM，

input bit[31:0] MAX_BEAM_NUM，

input bit[31:0] BGROUP_NUM，

input bit[31:0] BCOMBIN_NUM，

output bit[15:0] COEF_REAL，

output bit[15:0] COEF_IMAG)；

This C function completes the exchange of data with excitation generator by seven parameters, be array number (ELE_NUM), port number (CH_NUM), maximum numbers of beams (MAX_BEAM_NUM), wave beam group number (BGROUP_NUM, beam collection number (BCOMBIN_NUM), coefficient real part (COEF_REAL) and coefficient imaginary part (COEF_IMAG) respectively.

2., in C function, the Initial Azimuth (RCV_BEAM_START) produced according to wave beam, orientation (RCV_BEAM_START) is stopped and step-length (RCV_BEAM_STEP) determines the scan angle theta of m wave beam ₀:

θ ₀＝(RCV_BEAM_START+RCV_BEAM_STEP×m)×π/180

And then determine the coefficient of the n-th array element at m wave beam

$array\_coef(n,m)=e^{-j\frac{n\·d}{\mathrm{\λ}}sin\left({\mathrm{\θ}}_0\right)}$

Wherein, d is the spacing between bay, and λ is wavelength.

3. last, the real part of coefficient and empty step are converted to respectively the fixed-point number of sixteen bit, input to excitation generator by DPI interface

$COEF\_REAL=cos\left(sin\left({\mathrm{\θ}}_0\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

$COEF\_IMAG=sin\left(sin\left({\mathrm{\θ}}_0\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

The weight coefficient real part of generation and imaginary data are loaded on interface, to complete the excitation to weight coefficient in beamformer module by excitation generator as required.

Concrete to implement at this, due to other coefficient as the generation method of correction coefficient, beam position coefficient and the generation method of weighting coefficient similar, therefore no longer this repeats.

As shown in Figure 3, in this specific embodiment, the phased-array radar Digital Beamformer module verification platform based on UVM that the present invention creates adopts system-level hardware description language System Verilog to complete, this verification platform comprises test case library (Testcase Library) 1, UVM verification environment (Enviroment) 2, data send proxy server 3, data receiver proxy server 4 and scoring board (scoreboard) 5, data send proxy server 3 and data receiver proxy server 4 is all positioned at UVM verification environment 2, data send proxy server 3 and comprise the first monitor (monitor) 6, data-coefficients generator 7, first excitation generator (sequencer) 8 and the first driver (driver) 9, data receiver proxy server 4 comprises beam data converter 10, second monitor (monitor) 11, second driver (driver) 12 and the second excitation generator (sequencer) 13, first monitor 6 gathers the port data in DBF module 15 by interface 14, first excitation generator 8 one end is connected with data-coefficients generator 7 by DPI interface, first excitation generator 8 other end is connected with first driver 9 one end, first driver 9 other end is connected with interface 14 one end, interface 14 other end is connected with DBF module 15, second monitor 11 is connected with beam data converter 10 by DPI interface, second monitor 11 gathers the port data in DBF module 15 by interface 14, second driver 12 one end encourages generator 13 to be connected with second, second driver 12 other end is connected with interface 14, interface 14 is connected with DBF module 15, test case library 1 sends proxy server 3 with data simultaneously and data receiver proxy server 4 is connected, UVM verification environment 2 is connected with scoring board 5 by TLM interface, scoring board 5 comprises reference model 16 and comparer 17, comparer 17 one end is connected with reference model 16, comparer 17 other end is connected with the second monitor 11 in UVM verification environment 2.

The operation of the verification platform in the present invention comprises the step of following order:

A, the data-coefficients generator 7 realized by C language, to channel data, weighting coefficient, correction coefficient and beam position coefficient carry out complex operation, and complete the conversion that floating-point counts to fixed-point number.

B, the excited data bag of the first excitation generator 8 by producing in DPI interface interchange data-coefficients generator 7, and by it in a certain order and quantity, send to the first driver 9.

The excited data bag that C, the first driver 9 will obtain, is driven in DBF module by interface 14.

D, the second monitor 11 gather the port data of DBF, and by DPI interface, the data sampled are converted to complex field signal by the beam data converter 10 realized by C language equally, and writing in files.

Whether E, scoring plug (Scoreboard) 5 be consistent with the output of reference model 16 by design more to be measured, verifies whether design to be measured normally runs, and provide the result.

F, last, whether traveled through by the statistical study measuring ability of coverage rate, correct to ensure the function designed in practice environment.

Above-described embodiment is only be described the preferred embodiment of the present invention; not scope of the present invention is limited; under not departing from the present invention and designing the prerequisite of spirit; the various distortion that those of ordinary skill in the art make technical scheme of the present invention and improvement, all should fall in protection domain that claims of the present invention determines.

Claims (4)

1., based on a phased-array radar Digital Beamformer module verification method of UVM, its spy is being: this verification method comprises following sequential steps:

1. in excitation generator block, define DPI interface and state the C function that channel data, weighting coefficient, correction coefficient and beam position coefficient generate;

2. C function is according to parameter configuration and algorithm requirement, generates corresponding channel data and coefficient data, and is converted to fixed-point number and passes to excitation generator;

3. the excited data packing that excitation generator block is produced by TLM interface by driver is driven into interface, and then is loaded in DBF module;

4. monitor module gathers the data that DBF module exports, reference model is sent to carry out the automatic comparison of result on the one hand, on the one hand the data of collection are converted in complex plane represent direction and range information by calling DPI interface, and writing in files, for next step data analysis.

2. the phased-array radar Digital Beamformer module verification method based on UVM according to claim 1, is characterized in that: in described excitation generator, the production method of passage complex signal is as follows:

1. first, in excitation generator, state DPI interface, and need the passage complex signal called to produce C function:

This C function completes the exchange of data with excitation generator by four parameters, be array number (ELE_NUM), port number (CH_NUM), data real part (DATA_REAL) and data imaginary part (DATA_IMAG) respectively.

2. in C function, according to the initial direction (DIR_START) of radar bearing scanning, the step-length (DIR_STEP) terminating direction (DIR_END) and direction change determines the scan angle theta of m range unit

θ＝(DIR_START+DIR_STEP×m)×π/180

And then determine the complex signal of the n-th array element at m range unit

$\mathrm{array}\_\mathrm{data}(n,m)=e^{j\frac{n\·d}{\mathrm{\λ}}\sin \left(\mathrm{\θ}\right)}$

Wherein, d is the spacing between bay, and λ is wavelength.

3. last, the real part of complex data and empty step are converted to respectively the fixed-point number of sixteen bit, input to excitation generator by DPI interface,

$\mathrm{DATA}\_\mathrm{REAL}=\cos \left(\sin \left(\mathrm{\θ}\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

$\mathrm{DATA}\_\mathrm{IMAG}=\sin \left(\sin \left(\mathrm{\θ}\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

The data real part of generation and imaginary data are loaded on interface, to complete the excitation to channel data in beamformer module by excitation generator as required.

3. the phased-array radar Digital Beamformer module verification method based on UVM according to claim 1, is characterized in that: in described excitation generator, the production method of weighting coefficient is as follows:

1. first, in excitation generator, state DPI interface, and need the weighting coefficient called to produce C function:

This C function completes the exchange of data with excitation generator by seven parameters, be array number (ELE_NUM), port number (CH_NUM), maximum numbers of beams (MAX_BEAM_NUM), wave beam group number (BGROUP_NUM, beam collection number (BCOMBIN_NUM), coefficient real part (COEF_REAL) and coefficient imaginary part (COEF_IMAG) respectively.

2., in C function, the Initial Azimuth (RCV_BEAM_START) produced according to wave beam, orientation (RCV_BEAM_START) is stopped and step-length (RCV_BEAM_STEP) determines the scan angle theta of m wave beam ₀:

θ ₀＝(RCV_BEAM_START+RCV_BEAM_STEP×m)×π/180

And then determine the coefficient of the n-th array element at m wave beam

$\mathrm{array}\_\mathrm{coef}(n,m)=e^{-j\frac{n\·d}{\mathrm{\λ}}\sin \left({\mathrm{\θ}}_0\right)}$

Wherein, d is the spacing between bay, and λ is wavelength.

3. last, the real part of coefficient and empty step are converted to respectively the fixed-point number of sixteen bit, input to excitation generator by DPI interface

$\mathrm{COEF}\_\mathrm{REAL}=\cos \left(\sin \left({\mathrm{\θ}}_0\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

$\mathrm{COEF}\_\mathrm{IMAG}=\sin \left(\sin \left({\mathrm{\θ}}_0\right)\frac{n\·d}{\mathrm{\λ}}\right)\·2^{15}-1$

The weight coefficient real part of generation and imaginary data are loaded on interface, to complete the excitation to weight coefficient in beamformer module by excitation generator as required.

4. one kind for realizing as claimed in claim 1 based on the verification platform of the phased-array radar Digital Beamformer module verification method of UVM, it is characterized in that: this verification platform comprises test case library (1), UVM verification environment (2), data send proxy server (3), data receiver proxy server (4) and scoring board (5), described data send proxy server (3) and data receiver proxy server (4) is all positioned at UVM verification environment (2), described data send proxy server (3) and comprise the first monitor (6), data-coefficients generator (7), first excitation generator (8) and the first driver (9), described data receiver proxy server (4) comprises beam data converter (10), second monitor (11), second driver (12) and the second excitation generator (13), described first monitor (6) gathers the port data on DBF module (15) by interface (14), described first excitation generator (8) one end is connected with data-coefficients generator (7) by DPI interface, first excitation generator (8) other end is connected with the first driver (9) one end, described first driver (9) other end is connected with interface (14) one end, interface (14) other end is connected with DBF module (15), second monitor (11) is connected with beam data converter (10) by DPI interface, second monitor (11) gathers the port data on DBF module (15) by interface (14), described second driver (12) one end encourages generator (13) to be connected with second, second driver (12) other end is connected with interface (14), described interface (14) is connected with DBF module (15), described test case library (1) sends proxy server (3) with data simultaneously and data receiver proxy server (4) is connected, described UVM verification environment (2) is connected with scoring board (5) by TLM interface, described scoring board (5) comprises reference model (16) and comparer (17), described comparer (17) one end is connected with reference model (16), comparer (17) other end is connected with the second monitor (11) in UVM verification environment (2).