Detailed Description
In order to overcome the defects of the existing method for estimating the modulation precision in the radio frequency consistency test of the terminal, the embodiment of the invention provides the method for estimating the modulation precision, so that the modulation precision can be accurately, simply and effectively estimated.
Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
As shown in fig. 1, a detailed method flow for performing modulation accuracy estimation provided in the embodiment of the present invention is as follows:
step 101: and determining the synchronous position of the test signal and the reference signal, and sampling the test signal according to the synchronous position to determine at least one path of test data.
Preferably, the test data of 2N +1 paths in the test signal is determined by taking the determined synchronous position of the test signal and the reference signal as a center, wherein N is greater than or equal to 0.
In one specific implementation, the specific process of determining the synchronous position of the test signal and the reference signal is as follows: and carrying out correlation operation on the local pilot frequency or the training sequence in the reference signal and the test signal, and determining the synchronous position of the test signal according to the correlation peak value.
According to different specific applications, the position determined by the correlation peak value in the test signal is the synchronous position, or the position in the test signal, which is different from the position determined by the correlation peak value by a set value, is the synchronous position.
In practical application, the test equipment can directly provide a bit data sequence d obtained by detection and demodulation according to test parameter configuration, and perform data recovery on the bit data sequence d to obtain L1Multiplied by the oversampled reference signal X. The specific process of data recovery is as follows: firstly, modulating a bit data sequence d to obtain a modulated symbol data sequence s, and carrying out L on the symbol data sequence1And performing multiple upsampling and pulse shaping filtering to obtain a reference signal X.
Preferably, the test signal and the reference signal are further L2And (4) performing double upsampling. Wherein the test signal itself is subjected to L1The up-sampled signal is multiplied.
In practical application, to ensure that the result of the modulation accuracy estimation is accurate, the local pilot frequency or training sequence in the reference signal and the L of the received test signal may be adjusted1Multiplying the oversampled data, then L2Double oversampling and low-pass filtering to obtain L ═ L1L2The oversampled data is then subjected to correlation and correlation peak search.
In the embodiment of the invention, in order to avoid the influence of inaccurate synchronization on modulation precision estimation, one path of data of the synchronization position in the test signal is taken, and N paths of data on the left and right of the synchronization position in the test signal are taken to perform subsequent modulation precision estimation respectively, and then the final modulation precision estimation is determinedThe modulation accuracy of (1). Wherein, if the test signal for performing the correlation operation is the second upsampling (L)2Multiple upsampling) of the test signal, the determined multipath data belongs to L1The up-sampled test signal.
Step 102: aiming at any path of test data, calculating a frequency offset value and a phase offset value of the test data by using a reference signal, performing frequency offset and phase offset compensation on the test data by using the frequency offset value and the phase offset value, calculating origin offset of the test signal by using the test data after the frequency offset and the phase offset compensation and the reference signal, further calculating an amplitude correction factor, performing amplitude correction and origin offset elimination on the test data after the frequency offset and the phase offset compensation, and calculating mean square error vector amplitude (RMS EVM) by using the test data after the reference signal and the amplitude correction and the origin offset elimination.
One path of test data Y0For example, a specific process for calculating the frequency offset value and the phase offset value provided in the embodiment of the present invention is described as follows: calculating data in reference signal X and test data Y0And removing phase jump caused by the periodicity of 2 pi from the phase difference, performing unary linear regression fitting on the phase difference without the phase jump to obtain an intercept and a slope, and taking the intercept as an initial phase offset value, wherein the slope is a frequency offset value. The phase jump is caused by the periodicity (2 pi) of the phase, and the operation of removing the phase jump is the operation of adding or subtracting the integral multiple of the periodic phase from the phase difference, so that the absolute value of the difference value of two adjacent phase differences is smaller than a set threshold, and preferably, the set threshold belongs to the range of pi being larger than or equal to pi and smaller than 2 pi.
The embodiment of the present invention is not limited to the frequency offset and phase offset estimation method, and other frequency offset and phase offset estimation methods may also be used in the embodiment of the present invention, and the present invention is not limited thereto.
In the embodiment of the invention, a specific process of calculating the origin offset of the test signal by using the test data after the frequency offset and the phase offset compensation and the reference signal is as followsThe following: test data Y after compensating frequency deviation and phase deviation0Respectively carrying out unary linear regression fitting on a real part and unary linear regression fitting on an imaginary part on data in the reference signal X, taking the ratio of the slope and the intercept obtained by the unary linear regression fitting of the real part as the real part of the origin offset, and taking the ratio of the slope and the intercept obtained by the unary linear regression fitting of the imaginary part as the imaginary part of the origin offset, thereby obtaining the origin offset Co。
In the embodiment of the invention, the test data T after the frequency deviation and the phase deviation compensation
0' the specific procedure for performing amplitude correction and origin offset removal is as follows: is calculated so that | Y
0′-A(X+C
0)
2II minimum amplitude correction factor A to obtain
Wherein Y' 0 represents frequency offset and phase offset compensation
Compensated test data, X represents a reference signal, C
oRepresenting an offset of origin, "|
2"represents the square of the modulus; using a formula
And carrying out amplitude correction and origin offset elimination on the test data after the frequency deviation and the phase deviation compensation, wherein Y' represents the test data after the amplitude correction and the origin offset elimination.
When calculating the error vector of the test data Y ' and the reference signal X after amplitude correction and origin offset elimination, if the test data Y ' and the reference signal X are both over-sampled data, respectively filtering and down-sampling the test data Y ' and the reference signal X to obtain the test data Y without over-sampling0'and reference signal X' without oversampling, and then according to the formula e ═ Y0'X' calculates the error vector, then calculates the error vector according to the formulaAnd calculating to obtain the RMSEVM value corresponding to the road test data.
Step 103: and determining the minimum value in the RMS EVM corresponding to each path of test data, and further determining other indexes of the modulation precision according to each intermediate calculation result obtained in the process of calculating the RMS EVM by using the path of test data corresponding to the minimum value.
The intermediate calculation results obtained by the path of test data corresponding to the minimum value in each RMS EVM in the RMS EVM calculation process include, but are not limited to: a frequency offset value, a phase offset value, an origin offset and an error vector; other indicators of modulation accuracy include, but are not limited to: symbol EVM, peak EVM, 95% EVM, and origin offset suppression.
Specifically, if the test data is 2N +1 paths, traversing 2N +1 paths of test data, repeating the process in step 102 to obtain 2N +1 RMS EVM values, selecting the minimum RMS EVM value as the final RMS EVM value, outputting a frequency offset value, a phase offset value and an origin offset value obtained in the calculation process of the path of test data corresponding to the minimum RMS EVM value as the final frequency offset value, the phase offset value and the origin offset value as intermediate results, and calculating other indexes of the modulation accuracy according to the intermediate results, wherein the other indexes include but are not limited to symbol EVM, peak EVM, 95% EVM, OOS and the like.
Preferably, the above steps 101 to 103 may be repeated, and the average value of each index of the modulation precision is calculated after multiple estimations, so as to further improve the accuracy of the estimation.
The flow of the preferred embodiment of the present invention is shown in fig. 2, and the specific process of the embodiment is as follows: after the operation is started, receiving a test signal, performing data recovery according to demodulation bits provided by test equipment to obtain a reference signal, performing correlation operation and correlation peak value search on the received test signal and the reference signal after up-sampling, determining a synchronous position in the up-sampled test signal according to the correlation peak value, then performing down-sampling on the up-sampled test signal, and determining 2N +1 paths of test data in the up-sampled test signal; calculating a frequency offset value and a phase offset value by using a reference signal aiming at any path of test data and compensating the test data; calculating an origin offset value of the test signal by using the compensated test data and the reference signal, further calculating an amplitude correction factor, and performing amplitude correction and origin offset compensation on the compensated test signal; because the test data is the data after up-sampling, the test data and the reference signal after amplitude correction and origin offset compensation need to be down-sampled to obtain the test data and the reference signal without oversampling after the amplitude correction and origin offset compensation without oversampling, and then the RMS EVM value corresponding to the test data is further calculated; traversing the determined 2N +1 paths of test data to obtain 2N +1 RMSEVM values, selecting the minimum RMS EVM value as a final RMS EVM value, and outputting an intermediate result obtained in the calculation process of one path of test data corresponding to the minimum RMS EVM value to calculate other indexes of modulation precision; and repeating the process of calculating the RMS EVM value to carry out multiple measurements, and respectively calculating and outputting the average value of modulation precision indexes such as the RMS EVM and the like obtained by the multiple measurements so as to further improve the estimation accuracy, and ending the estimation process.
The following takes an 8PSK modulation method under the EGPRS system as an example, and details a specific implementation process of the modulation accuracy estimation will be described with reference to fig. 3.
Step 1: replacing TSC (Training Sequence Code) bits in the middle of a bit data Sequence d obtained by demodulation and provided by test equipment with known local TSC bits (78 bits), then carrying out modulation, OSR (optical signal to noise ratio) up-sampling and pulse shaping filtering to obtain a local reference signal rcIntercepting sampling points corresponding to the positions of the 26 TSC symbols as a local TSC code reference signal rc_TSC。
Step 2: reference signal r using local TSC codec_TSCPerforming a correlation operation with the received test signal r to determine an accurate synchronization position I in the test signal based on the correlation peakmaxWhere the test signal r itself is an OSR times up-sampled signal.
The specific process is as follows:
2 a: for test signal r and local TSC code reference signal rc_TSCCarry out NOSRMultiple upsampling, local TSC code reference signal rc_TSCThe sampling rate after upsampling is OSR' ═ OSR × NOSRRespectively to give r'OSR' and rc″_TSC_OSR′Wherein r isc″_TSC_OSR′The sampling points corresponding to 16 symbols in the middle of the TSC sequence are intercepted, and the up-sampling can be carried out by NOSRDouble interpolation 0 and low pass filtering implementation.
2 b: calculating rO' and rc″_TSC_OSR′Correlation value of CTSC=[CTSC,0,CTSC,1,…,CTSC,M-1]Where M ═ OSR' × 2P generally requires a synchronization timing process before modulation quality estimation is performed, it is assumed that the error of coarse synchronization is within P symbols.
2 c: searching for sample point number I corresponding to maximum square value of absolute value of correlation valuemax: I.e. the correlation peak position.
2 d: re-acquisition by correlation peak position for NOSRAnd (3) performing down-sampling on the data of one time slot n in the test signal subjected to the up-sampling, and acquiring 2 × Δ I +1 paths of test data of the test signal subjected to the OSR up-sampling after accurate synchronization.
In order to ensure the accuracy of the estimation in this step, ImaxAnd (3) taking the delta I test data on the left and right sides of the central position respectively to obtain 2 x delta I +1 test data, respectively performing the calculation of the steps (3) to (5), and then selecting the test data with the minimum RMS EVM to perform the calculation of other indexes.
The following description of steps 3 to 5 is given by taking any one of the obtained 2 × Δ I +1 paths of test data r' as an example
And step 3: and calculating the frequency offset value and the phase offset value of the received test data r 'and compensating the test data r'.
The method comprises the following specific steps:
3 a: calculating an error signal r
e=[r
e(0),r
e(1),…,r
e(N-1)]Wherein
And N is the number of sampling points of one time slot after the guard interval is removed.
3 b: calculate the phase of the error signal: phie=[φe(0),φe(1),…,φe(N-1)]Wherein phie(i) Is re(i) Removes the phase after the phase jump due to the 2 pi periodicity.
The specific process of removing the phase jump can be identified by a pseudo code as follows:
for i=1:N-1
ifφe(i)-φe(i-1)>Threshold_unwarp
φe(i:end)=φe(i:end)-2π
elseifφe(i)-φe(i-1)<-Threshold_unwarp
φe(i:end)=φe(i:end)+2π
end
end
wherein Threshold _ unwarp is a Threshold value set for eliminating phase jump, and is generally a number in a range of pi being equal to or greater than pi and smaller than 2 pi, and can be 300 pi/180, for example. Wherein phi ise(i) At the first occurrence (i.e. initial value) is re(i) The second occurrence is the value updated with 2 pi for the previous time.
3 c: using the formula y ═ a0+x*a1Performing linear regression fitting on the phase of the error signal to obtain an estimated frequency offset value f from the slopeeObtaining an estimated initial phase offset value phi from the intercept0Then, the test data r' is subjected to phase offset and frequency offset compensation to obtain r ″. Wherein, the concrete formula of the unary linear regression fitting is represented as: wherein i=1,2,…,N
And 4, step 4: test data r' and reference signal r compensated using frequency offset and phase offsetcAnd calculating the origin offset of the test signal and further calculating an amplitude correction factor, performing IQ imbalance calculation, and performing origin offset compensation and amplitude correction on r'.
The method comprises the following specific steps:
4 a: r' and rcRespectively carrying out unary linear regression fitting operation on the real part and the imaginary part of each sequence data Fitting the real part to obtain an intercept b01And slope b1Fitting the imaginary part to obtain the intercept b02And slope b2。
Wherein, n, see 3c above, and r "may be taken to reduce complexitycA portion of the data in each sequence data was fit calculated.
4 b: calculating origin offset Co:
4 c: calculating an amplitude correction factor C1First, the origin offset is added to the reference signal, and the formula is expressed as: r isco=rc+CoThen, an amplitude correction factor is calculated, which is expressed as:
4 d: carrying out origin offset compensation and amplitude correction on the test data r 'after the frequency offset and phase offset compensation to obtain r':
4 e: the amplitude gains of the I path and the Q path can be output: gI=bl,GQ=b2。
And 5: RMS EVM calculation.
The method comprises the following specific steps:
5 a: will reference the signal r locallycAnd the test data r '″ after phase offset frequency offset and origin offset compensation is subjected to a measurement filter and OSR multiple down sampling to obtain r'cAnd rm′。
5 b: calculating the error vector e-rm′-rc' (the trailing bits of each 3 symbols before and after removal).
5 c: the RMS EVM value is calculated by the following specific formula:
step 6: traversing 2 × Δ I +1 paths of test data in step 2.4, respectively repeating the steps 3-5 to obtain 2 × Δ I +1 RMS EVM values, taking the smallest one of the 2 × Δ I +1 RMS EVM values as the final RMS EVM value of the time slot n, taking an intermediate result obtained in the RMS EVM estimation process of one path of data corresponding to the smallest RMS EVM value as the final output, and calculating the rest modulation accuracy indexes by using the path of data and the intermediate result.
The calculation of the remaining modulation accuracy indexes specifically includes, but is not limited to, the following:
6 a: the symbol EVM (n, i) of slot n is obtained, and the formula is as follows:
wherein L issymThe number of symbols after the guard interval is removed for one slot.
6 b: obtaining the peak value EVM of the time slot n:
6 c: obtaining an OOS linear value of the time slot n:
and 7: repeating the steps 1-6 to repeat the test Nburst_EVM_OOSAnd (4) carrying out time slot, respectively averaging all test indexes (frequency offset, phase offset, EVM, OOS and the like) to be used as test results to be output, and obtaining 95% EVM by counting the cumulative distribution of the EVM.
Based on the same inventive concept, as shown in fig. 4, the embodiment of the present invention further provides a modulation accuracy estimation apparatus, and the specific implementation of the apparatus may refer to the specific implementation of the above method, and repeated details are not repeated, and the apparatus mainly includes the following units:
a determining unit 401, configured to determine a synchronization position of a test signal and a reference signal, and sample the test signal according to the synchronization position to determine at least one path of test data;
a processing unit 402, configured to calculate, for any path of test data, a frequency offset value and a phase offset value of the test data by using a reference signal, perform frequency offset and phase offset compensation on the test data by using the frequency offset value and the phase offset value, calculate an origin offset of the test signal by using the test data after the frequency offset and the phase offset compensation and by using the reference signal, further calculate an amplitude correction factor, perform amplitude correction and origin offset cancellation on the test data after the frequency offset and the phase offset compensation, and calculate a mean square error vector amplitude by using the test data after the reference signal and the amplitude correction and the origin offset cancellation;
a selecting unit 403, configured to determine a minimum value in the RMS EVM corresponding to each path of test data, and further determine other indexes of the modulation accuracy according to intermediate calculation results obtained in the process of calculating the RMS EVM by using one path of test data corresponding to the minimum value, where the intermediate calculation results at least include: frequency offset value, phase offset value, origin offset, and error vector.
In one particular implementation, the processing unit is specifically configured to: calculating the phase difference between the data in the reference signal and the test data, removing the phase jump caused by the periodicity of 2 pi to the phase difference, and performing unary linear regression fitting on the phase difference without the phase jump to obtain an intercept and a slope, wherein the intercept is an initial phase deviation value, and the slope is a frequency deviation value.
In one particular implementation, the processing unit is specifically configured to: and respectively carrying out unary linear regression fitting of a real part and unary linear regression fitting of an imaginary part on the test data and the reference signal after the frequency offset and the phase offset compensation, wherein the ratio of the slope to the intercept obtained by the unary linear regression fitting of the real part is used as the real part of the origin offset, and the ratio of the slope to the intercept obtained by the unary linear regression fitting of the imaginary part is used as the imaginary part of the origin offset.
Wherein the processing unit is specifically configured to: is calculated so that | Y0′-A(X+C0)‖2Minimum amplitude correction factor A to obtain Wherein, Y0' denotes test data after frequency offset and phase offset compensation, X denotes a reference signal, CoRepresenting origin offset, "|2"denotes the square of the modulus.
Wherein the processing unit is further to: using a formula
And carrying out amplitude correction and origin offset elimination on the test data after the frequency deviation and the phase deviation compensation, wherein Y' represents the test data after the amplitude correction and the origin offset elimination.
Wherein the determining unit is specifically configured to: and carrying out correlation operation on the local pilot frequency or the training sequence in the reference signal and the test signal, and determining the synchronous position of the test signal according to the correlation peak value.
Wherein the determining unit is specifically configured to: and determining the test data of 2N +1 paths in the test signal by taking the determined synchronous position as the center, wherein N is greater than or equal to 0.
Based on the technical scheme, in the embodiment of the invention, after the synchronous position of the test signal and the reference signal is determined, at least one path of test data is determined according to the synchronous position, for any path of test data, the frequency offset value and the phase offset value of the test data are calculated by using the reference signal, after the test data is subjected to frequency offset and phase offset compensation, the origin offset of the test signal is calculated and an amplitude correction factor is further calculated, after the test data subjected to frequency offset and phase offset compensation is subjected to amplitude correction and origin offset elimination, the mean square error vector amplitude RMS EVM is calculated by using the test data subjected to reference signal and amplitude correction and origin offset elimination. The method does not need a complex iterative process, does not need complex quadratic linear regression fitting, is simple to implement, determines more than one path of test data after determining the synchronous position of the test signal and the reference signal, respectively calculates the RMS EVM value of each path of test data, and further determines other indexes of the modulation precision according to the intermediate calculation results obtained by the path of test data corresponding to the minimum RMS EVM value in the RMS EVM calculation process, thereby ensuring the estimation accuracy.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.