CN103188186B - Resampling processing unit and method and digital modulation signals generation device - Google Patents
Resampling processing unit and method and digital modulation signals generation device Download PDFInfo
- Publication number
- CN103188186B CN103188186B CN201210470733.2A CN201210470733A CN103188186B CN 103188186 B CN103188186 B CN 103188186B CN 201210470733 A CN201210470733 A CN 201210470733A CN 103188186 B CN103188186 B CN 103188186B
- Authority
- CN
- China
- Prior art keywords
- data
- filter coefficient
- resampling
- filter
- input
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention provides a kind of resampling processing unit and utilize the digital modulation signals generating means of this resampling processing unit and speed ratio to have the resampling processing method of the larger degree of freedom.Resampling processing unit of the present invention is in order to carry out resampling process that speed ratio is m/n and have to the data rows Din inputted with sample rate Fin: FIR type digital filter (21), it carries out the multiply-add operation that up-to-date P the data of data rows Din and P are the filter coefficient of a group, and is exported successively by its operation result; Filter coefficient means of storage (25), (m × P) the individual filter coefficient needed for the data interpolating carrying out each timing of m decile to the data input interval relative to digital filter (21) gathers by each each timing and prestores by it; And filter coefficient switching member (26), it selects P be the filter coefficient of a group and give digital filter (21) with the resampling speed of Finm/n successively from filter coefficient means of storage (25), to be input into digital filter (21) period n data, every n, temporally sequence output m is individual about interpolated data regularly.
Description
Technical field
The present invention relates to a kind of technology of the degree of freedom for expanding the rate conversion in resampling process, this resampling process data rows that synthesis sample rate is different in digital modulation signals generation device etc. also produces when carrying out the signal of digital modulation with the data rows of this synthesis, the data rows of at least one party is converted to data rows time other sample rate (resampling).
Background technology
Produce the data rows of device basically by the base band of generation I, Q2 phase of digital modulation signals, and convert them to analog signal respectively by D/A converter and be input in quadrature modulator, thus generate at desired frequency band the digital modulation signals that phase place/amplitude changes according to I, Q2 value.
The device of this generation digital modulation signals is for utilizing the evaluation of the various communication equipments of digital modulation mode, but in reception is evaluated, be not only expectation ripple, characteristic when also adding disturbing wave also becomes important gradually.
Especially nearest, the demand for the expectation ripple of combination distinct symbols rate and the evaluation of disturbing wave progressively increases.
As generating the method comprising the desired wave of different ratio and the signal of disturbing wave; preparation 2 digital modulation signal generation devices can be expected; the disturbing wave produced in the expectation ripple generated in one side's device and the opposing party's device is carried out closing ripple and the mode being input into evaluation object; but cause the large-scale of system configuration, and cost increases.
Therefore, as simpler mode, the data rows of synthesizing different sample rate in synthesis base band can be expected, and generation carries out the signal of digital modulation to be input into evaluation object with the data rows of this synthesis.
In order to 2 signal trains that synthesis rate like this is different, the ratio carrying out both is needed to coordinate process, as the simplest method of this process, have and both rate conversion are become the common ratio of its least common multiple and the method (referenced patent document 1) of carrying out add operation process.
Patent documentation 1: Japanese Patent Publication 2006-29862 publication
But, when the signal generation device of the data rows for processing plurality of sampling rates, kind for the sample rate of the data rows of synthesis is more, on the other hand, there is the combination entering frequency band limited on hardware and the ratio of least common multiple can be converted to limited, and the problem that the versatility that can be used in the mensuration of the communication system equipment of wide range is short of to some extent.
Especially, when the sample rate for the data rows of carrying out add operation synthesis is close, its least common multiple becomes very large compared with former ratio.Such as, when the ratio of 2 sample rate F1, F2 is 101:99, have to each rate conversion to become 99 times and 101 times, and require the rate conversion process respectively it being carried out to hypervelocity, if former sample rate be at a high speed, then its rate conversion becomes extremely difficult.
Summary of the invention
The present invention completes to solve this problem, its object is to the resampling treatment technology providing a kind of speed ratio to have the larger degree of freedom, and provide a kind of and can carry out the higher digital modulation signals generation device of the versatility of the composite test of the data rows of plurality of sampling rates by utilizing this technology.
In order to realize described object, the resampling processing unit of claim 1 of the present invention, it carries out to the data rows inputted with predetermined sample rate Fin the resampling process that speed ratio is m/n, wherein, m, n are integer, it is characterized in that, this resampling processing unit possesses: FIR(FiniteImpulseResponse, finite impulse response (FIR)) type digital filter (21,21A, 21B), its carry out described in up-to-date P the data of data rows that are transfused to and P be the multiply-add operation of the filter coefficient of a group, and its operation result to be exported successively; Filter coefficient means of storage (25), it prestores m × P filter coefficient; And filter coefficient switching member (26), it selects described P be the filter coefficient of a group and give described digital filter with the resampling speed of Finm/n from described filter coefficient means of storage successively, so that during a described n data are input into described digital filter, every n, temporally sequence output m is individual about interpolated data regularly.
And, the resampling processing unit of claim 2 of the present invention, it is characterized in that, be stored in the filter coefficient of described filter coefficient means of storage, is the filter coefficient needed for data interpolating carried out the input interval of the data being input into described digital filter in each timing of m decile.
Further, the digital modulation signals generation device of claim 3 of the present invention, it has: the 1st data rows output link (31), and it exports the 1st data rows for modulating with the 1st sample rate, 2nd data rows output link (32), it exports the 2nd data rows for modulating with the 2nd sample rate different from the 1st sample rate, and resampling handling part (40, 40A, 40B), it is in order to coordinate described 1st data rows with the sample rate of the 2nd data rows, the resampling process that speed ratio is m/n is carried out to the data rows of at least one party, wherein, m, n is integer, this digital modulation signals generation device is to be joined by described resampling process and the data rows of sample rate carries out add operation, and produce the signal being carried out digital modulation by the data rows of this add operation, it is characterized in that, described resampling handling part possesses: FIR(FiniteImpulseResponse, finite impulse response (FIR)) type digital filter (21, 21A, 21B), its carry out with up-to-date P the data of data rows of sample rate Fin input and P individual be the multiply-add operation of the filter coefficient of a group, and its operation result is exported successively, filter coefficient means of storage (25), it prestores m × P filter coefficient, and filter coefficient switching member (26), it selects described P be the filter coefficient of a group and give described digital filter with the resampling speed of Finm/n from described filter coefficient means of storage successively, so that during n data are input into described digital filter, every n, temporally sequence output m is individual about interpolated data regularly.
And, the digital modulation signals generation device of claim 4 of the present invention, it is characterized in that, being stored in the filter coefficient of described filter coefficient means of storage, is the filter coefficient needed for data interpolating carried out the input interval of the data being input into described digital filter in each timing of m decile.
And, the resampling processing method of claim 5 of the present invention, it carries out to the data rows inputted with predetermined sample rate Fin the resampling process that speed ratio is m/n, wherein, m, n are integer, it is characterized in that, this resampling processing method comprises: up-to-date P data of the data rows be transfused to described in carrying out and P is the step of the multiply-add operation of the filter coefficient of a group; Use and export the FIR(FiniteImpulseResponse of its operation result successively, finite impulse response (FIR)) step of type digital filter; Prestore the filter coefficient storing step of m × P filter coefficient; And select the described P of described storage be the filter coefficient of a group and give described digital filter with the resampling speed of Finm/n successively, so that during n data are input into described digital filter, every n, temporally sequence output m is individual about interpolated data regularly, carries out the step that speed ratio is the resampling process of m/n thus.
And, the resampling processing method of claim 6 of the present invention, its filter coefficient stored in described filter coefficient storing step is the filter coefficient needed for data interpolating carried out the input interval of the data being input into described digital filter in each timing of m decile.
So, in the present invention, (m × P) individual filter coefficient needed for the data interpolating input interval of the data relative to digital filter being carried out to each timing of m decile is gathered by each each timing as interpolation candidate and prestores, and from filter coefficient means of storage, select P be the filter coefficient of a group and give digital filter with the resampling speed of Finm/n successively, so that during n data are input into digital filter, every n, temporally sequence output m is individual about interpolated data regularly.
Therefore, resampling process can be carried out with the speed of the wider range that can be represented by the ratio (m/n) of two integers than to the data rows of predetermined sample rate, and can realize can carrying out the higher digital modulation signals generation device of the versatility of add operation to the combination of the data rows of plurality of sampling rates by utilizing this resampling process.
Accompanying drawing explanation
Fig. 1 is the structure chart of the resampling processing unit of embodiments of the present invention.
The action specification figure of Fig. 2 to be execution mode be resampling process of m/n=4.
Fig. 3 represents the storage row of filter coefficient and the figure of reading order.
Fig. 4 is the timing diagram of the resampling process of the m/n=4 of execution mode.
Fig. 5 is the action specification figure of the resampling process of the m/n=4/3 of execution mode.
Fig. 6 is the action specification figure of the resampling process of the m/n=4/3 of execution mode.
Fig. 7 is the action specification figure of the resampling process of the m/n=4/3 of execution mode.
Fig. 8 is the timing diagram of the resampling process of the m/n=4/3 of execution mode.
Fig. 9 is the structure chart of I, Q2 facies pattern resampling processing unit.
Figure 10 is the figure of the execution mode representing digital modulation signals generation device of the present invention.
Figure 11 is the figure of another execution mode representing digital modulation signals generation device of the present invention.
Figure 12 be m/n=4/5(reduce sample rate time) action specification figure.
In figure: 20,20 '-resampling processing unit, 21,21A, 21B-digital filter, 22-register, 23-multiplier, 24-adder, 25-filter coefficient means of storage, 26-filter coefficient switching member, 27-filter coefficient write component, 30,30 '-digital modulation signals generation device, 31-the 1st data rows output link, 32-the 2nd data rows output link, 40,40A, 40B-resampling handling part, 41-configuration part, 42-Gain tuning portion, 43-adder, 44-D/A transducer, 45-quadrature modulator.
Embodiment
Below, based on accompanying drawing, embodiments of the present invention are described.
Fig. 1 represents the structure applying resampling processing unit 20 of the present invention.
This resampling processing unit 20 be speed that the data rows Din inputted with predetermined sample rate Fin is specified than m/n(wherein, m, n are integer, be set to m > n at this) the device of resampling process, it has digital filter 21, filter coefficient means of storage 25, filter coefficient switching member 26 and filter coefficient write component 27.
Digital filter 21 is by the register 22 be connected with multiple P level files of level displacement backward successively while of storing input data
1~ 22
p, obtain each register 22
1~ 22
poutput R
1~ R
pwith the filter coefficient h(i that P is a group, 1) ~ h(i, P) the multiplier 23 of long-pending multiple P
1~ 23
p, and obtain multiplier 23
1~ 23
pthe tap number that forms of the adder 24 of output summation be the FIR type digital filter of P.
This digital filter 21 carries out the data of interpolation for being obtained by computing between input data, and utilizes by impulse response F(X) character of the data value of the intermediate point (datum mark) of the summation approximate function waveform of=coefficient that (sinX)/X obtains and data.In addition, can use utilize window function W(X) F(X)=[(sinX)/X] W(X) formula to be to replace the formula of above-mentioned impulse response.
In order to carry out this data interpolating, in filter coefficient means of storage 25, be previously stored with (m × P) the individual filter coefficient needed for the data interpolating that each timing of m decile carried out to the input interval T s=1/Fin of the data relative to digital filter 21 gathered by each each timing.
And, filter coefficient switching member 26 selects P be the filter coefficient of a group and give digital filter 21 with the speed of Finm/n (resampling speed) successively from filter coefficient means of storage 25, so that during n data are input into digital filter 21, every n, temporally sequence output m is individual about interpolated data regularly.
Filter coefficient write component 27 calculates the filter coefficient needed for resampling process based on the sample rate Fin of input data and the speed of specifying than m/n and is stored in filter coefficient means of storage 25.
Then, the action of this resampling processing unit 20 is described.
For the ease of understanding the action in each portion, first illustrate that P=8, m=4, n=1(speed ratio is 4) situation.
By in digital filter 21 input have 8 input data Din(1) ~ Din(8) and timing be set to benchmark timing t=0.Now, as Fig. 2 (a), each register 22
1~ 22
8output R
1~ R
8become R
1=Din(1), R
2=Din(2) ..., R
8=Din(8).
For these 8 continuous print input data, if the register be set to by the initial timing of resampling process near the centre position of tap number (is now set to register 22
4, but also can be register 22
5) output valve R
4=Din(4), then the value of wishing to get in resampling is output valve R
4(=Din(4)) itself, this is known.
Therefore, at this moment, as shown in Fig. 2 (b), as long as to only having and output valve R
4carry out the filter coefficient h(1 of multiplying, 4) be 1 and other filter coefficients h(1,1) ~ h(1,3), h(1,5) ~ h(1,8) be 0 one group of filter coefficient carry out multiplying, the initial output value Dout(1 based on resampling process) become
Dout (1)=
j=1∑
8h (1, j) Rj=1 × Rj(mark
j=1Σ
8represent the summation till j=1 ~ 8).
In addition, emphasize accuracy at this and use itself the group of register coefficient of input-output data, but also can use the coefficient corresponding with the impulse response calculated for interpolated data described later.
In Fig. 2 (a), the speed ratio due to resampling is m/n=4, therefore needs to input data Din(4), Din(5) between Ts carry out 4(=m) obtain the interpolated data of Q1 timing in each timing Q1 ~ Q3 of splitting.
For this reason, as Fig. 2 (c), if use waveform center from the impulse response F(X-Δ T of initial timing only displacement T=Ts/m), and using the value of each sampling timing (Ts interval) being stored in 8 data of each register in the function F (X-Δ T) of this displacement as second group of filter coefficient h(2,1) ~ h(2,8) be set in digital filter 21, then can obtain the output valve Dout(2 of the timing Q1 based on resampling)
Dout（2）＝
j=1Σ
8h（2，j）·Rj。
Similarly, as Fig. 2 (d), if the impulse response F(X-2 Δ T using further displacement T) in the value of each sampling timing (Ts interval) being stored in 8 data of each register as the 3rd group of filter coefficient h(3,1) ~ h(3,8) digital filter 21 is set in, then can obtain the output valve Dout(3 of the timing Q2 based on resampling)
Dout（3）＝
j=1Σ
8h（3，j）·Rj。
In addition, as Fig. 2 (e), if the impulse response F(X-3 Δ T using further displacement T) in the value of each sampling timing (Ts interval) being stored in 8 data of each register as the 4th group of filter coefficient h(4,1) ~ h(4,8) digital filter 21 is set in, then can obtain the output valve Dout(4 of the timing Q3 based on resampling)
Dout（4）＝
j=1Σ
8h（4，j）·Rj。
Thus, can obtain comprising input data Din(4) until next data Din(5) between carry out 4 resampling data Dout(1 of interpolation) ~ Dout(4).And, before input data are updated, period carries out above-mentioned process with the interval of Ts/m, data Din(9 new) be transfused to and the stage upgraded to its up-to-date data Din(2) ~ Din(9) carry out above-mentioned process equally, by continuing to carry out this process, the resampling process of speed than m/n=4 can be carried out to data Din.
4 groups of filter coefficient h(1 in this example needed for resampling process, 1) ~ h(1,8), h(2,1) ~ h(2,8), h(3,1) ~ h(3,8), h(4,1) ~ h(4,8), as long as Fig. 3 is stored in the filter coefficient means of storage 25 be made up of memories such as RAM by sequence of addresses, and they often through Ts/4 time sequencing and are cyclically read be set in digital filter 21.
Fig. 4 is the figure gathering above-mentioned series of actions, as Fig. 4 (a) inputs data Din reproducting periods Ts at 1 (=n), as Fig. 4 (b) for the filter coefficient carrying out 4 groups (=m) needed for interpolation between 2 data with the interval of Ts/4=Δ T temporally sequence order and cyclically changing, thus synchronously export the resampling data Dout of the 4 sampling rates with input as Fig. 4 (c) and its.
In addition, as previously mentioned, as the 1st group of filter coefficient h(1,1) ~ h(1,8), also can at impulse response F(X) in the value of each sampling timing (Ts interval) being stored in 8 data of each register is obtained in groups, and to be set in digital filter 21.
Above-mentioned example is the example of m=4, n=1, about each point of m will be divided between input data, often input 1 input data, each point every 1 all temporally sequence obtain m interpolated data, to be a simple speed ratio be 4 rate conversion process, then the action of m=4, n=3 is described.
As shown in Fig. 5 (a), at 8 data Din(1) ~ Din(8) be stored in digital filter 21 the state of each register under, as Fig. 5 (b) selects the 1st group of filter coefficient h(1,1) ~ h(1,8), initial resampling data Dout(1 is exported)=R1=Din(4).
At this moment, because m=4 as described above, therefore for the treatment of initial timing the 1st group of filter coefficient and about identical with the situation of above-mentioned example to 3 groups of filter coefficients needed for the data interpolating carrying out the timing Q1 ~ Q3 of m decile between 2 input data, but next resampling is spaced apart every 3(=n) the individual timing namely postponing 3 Δ T, therefore as Fig. 5 (c), using from the timing Q3 of initial timing delay 3 Δ T as the impulse response F(X-3 Δ T at waveform center) in, the value being stored in each sampling timing (Ts interval) of 8 data of each register is selected the 4th group of filter coefficient h(4 in the described example of group, 1) ~ h(4, 8), export 2nd the resampling data Dout(2 corresponding with this timing Q3 thus).
Next resampling data are the timings postponing 3 Δ T further, namely in Fig. 5 (a) to the timing of the Q2 ' carried out between register data R5, R6 in each timing Q1 ' ~ Q3 ' of m segmentation, but before carrying out this resampling, through inputting the sampling period Ts of data from the initial timing of resampling process, thus as Fig. 6 (a), each register 22
1~ 22
8output valve R
1~ R
8be updated to data Din(2 respectively) ~ Din(9).
Therefore, from the timing of this input Data Update to the timing of t=Ts+2 Δ T arriving next resampling object timing Q2, as Fig. 6 (b) selects the 3rd group of filter coefficient h(3,1) ~ h(3,8), export thus this register 22
4, 22
5data Din(5), Din(6) between carry out the 3rd resampling data Dout(3 corresponding with timing Q2 in 3 of m decile timing Q1 ~ Q3).
Next resampling is the timing postponing 3 Δ T from the state of Fig. 6 further again, namely in figure 6 (a) to the timing of the Q1 ' carried out between register data R5, R6 in each timing Q1 ' ~ Q3 ' of m segmentation, but before carrying out this resampling, through inputting the time (2Ts) of 2 times of the sampling period Ts of data from the initial timing of resampling process, thus as Fig. 7 (a), each register 22
1~ 22
8output valve R
1~ R
8be updated to data Din(3 respectively) ~ Din(10).
Therefore, from the timing of this input Data Update to the timing of t=2Ts+ Δ T arriving next resampling object timing Q1, as Fig. 7 (b) selects second group of filter coefficient h(2,1) ~ h(2,8), export thus this register 22
4, 22
5data Din(6), Din(7) between carry out the 4th resampling data Dout(4 corresponding with timing Q1 in 3 of m decile timing Q1 ~ Q3).
Further, next resampling is the timing postponing 3 Δ T from the state of Fig. 7 further, namely in the output timing of the middle register data R5 of Fig. 7 (a), when carrying out this resampling, just in time through 3Ts from the initial timing of resampling process, thus each register 22
1~ 22
8output valve R
1~ R
8be updated to data Din(4 respectively) ~ Din(11).
This state is except register data is updated, identical with the initial condition of Fig. 5, be equivalent to the initial condition of getting back to resampling process, therefore this input Data Update timing is coordinated, the 1st group of filter coefficient h(1 is selected in the same manner as described Fig. 5 (b), 1) ~ h(1,8), export and this register 22 thus
4, 22
5data Din(7) equal the 5th resampling data Dout(5).
Namely, to input the 3/4(=n/m of the sampling period Ts of data in this resampling process) cycle doubly by filter coefficient with h(1, j) → h(4, j) → h(3, j) → h(2, j) sequential loop reads and switches, thus to input data Din, to carry out speed ratio be 4/3(=m/n constantly) resampling process.
Fig. 8 represents that the input data of above-mentioned example, filter coefficient switch and the timing diagram of relation of resampling data, 3(=n is upgraded at the input data Din of the sampling period Ts of Fig. 8 (a)) secondary period, as Fig. 8 (b), 4 groups of filter coefficients are with 3Ts/4(=Tsn/m) interval change according to the sequence, as Fig. 8 (c) exports the resampling data calculated by the filter coefficient of this change.In addition, if the cycle will be converted into speed, then the switch speed of filter coefficient and resampling speed will become Finm/n.
Above, the simple example of P=8, m=4, n=3 is illustrated, as long as but the tap number P of digital filter 21 carries out displacement and the number of the interpolation error of asymmetry that produces for being enough to reduce based on being responded by paired pulses, then and can be arbitrary.
Such as, be 16 bits at data length, when tap number P=28, to the W-CDMA modulating data of symbol rate 3.84MHz and the GSM(registered trade mark of symbol rate 270.833kHz being carried out the 3.25MHz of 12 times of over-samplings) with modulating data carry out add operation synthesize time, to GSM(registered trade mark) carry out speed than the resampling process of m/n=384/325 with modulating data and carry out conjunction ripple, its result is about GSM(registered trade mark) to determine phase error be 0.22deg, determining Modulation Error Rate about W-CDMA is (EVM) 0.51%, can confirm to have almost no change with or without resampling process.
Further, determine that the speed of resampling process is arbitrary than the scope of integer m, n of m/n, if but the upper limit of m, n is set as about 1024, then can tackle the combination of the symbol rate of the many digital modulations used now.
In above-mentioned each example, describe the resampling process that a series of input data are carried out, but for the I of digital modulation, during Q2 phase data row, if as Fig. 9 resampling processing unit 20 ' to as described in two digital filter 21A of data filter 21 same structure, 21B input is with 2 phase data Iin of predetermined sample rate (symbol rate) synchronously input, Qin, and the common filter coefficient for its multiply-add operation is similarly switched as above-mentioned by filter coefficient switching member 25, then can obtain the 2 phase data Iout carrying out resampling with speed than m/n, Qout.
Figure 10 represents the structure of the digital modulation signals generation device 30 possessed as resampling handling part 40 by the resampling processing unit 20 ' of I, Q2 facies pattern shown in Fig. 9.
This digital modulation signals generation device 30 has: the 1st data rows output link 31, and it exports for modulating and the 1st data rows D1 be made up of the group of I, Q2 phase with the 1st sample rate Fin1; And the 2nd data rows output link 32, it exports with the 2nd sample rate Fin2 different from the 1st sample rate Fin1 and to be made up of the group of I, Q2 phase and the 2nd data rows D2 for modulating equally.
At this, be set to the 2nd sample rate Fin2 lower than the 1st sample rate Fin1(Fin1 > Fin2), the 2nd lower for sample rate data rows D2 is input into the resampling handling part 40 with the resampling processing unit 20 ' same structure of described I, Q2 facies pattern, and carrys out command speed by the configuration part 41 of the various parameters for setting device and compare m/n=Fin1/Fin2.
Thus, export the original waveform of I, Q of data rows D2 respectively with Fin2(m/n from resampling handling part 40) speed, the data rows D2 ' be made up of the group of I, Q2 phase that namely equal with the sample rate Fin1 of the 1st data rows D1 ratio carries out resampling and obtains.
The data rows D2 ' obtained by this resampling process is input into Gain tuning portion 42 with same ratio (Fin1) together with the 1st data rows D1, and be limited to be treated to and prevent the position based on add operation from overflowing and making both amplitude ratio α/βs become the amplitude ratio of being specified by configuration part 41, and be input into adder 43 to carry out add operation with same ratio (Fin1).
The data rows Dsum be made up of the group of I, Q2 phase as this add operation result is input into dual-channel type D/A converter 44 with the 1st sample rate Fin1, and converts the analog signal Asum that is made up of the group of I, Q2 phase to and be input into quadrature modulator 45.
The I composition of quadrature modulator 45 couples of analog signal Asum and the sum Q composition of predetermined local signal and phase shift local signal local signal being carried out 90 degree of phase shifts are long-pendingly carries out conjunctions ripple, and the high-frequency digital modulation signal RFsum that modulates is carried out in output with the amplitude/phase determined by mutually orthogonal I, Q composition.The frequency of digital modulation signals RFsum is set by configuration part 41.
This digital modulation signals RFsum is applied in the modulation of the add operation result corresponding to the 1st data rows D1 and the 2nd data rows D2, by carrying out various change to the gain relative to two data rows, two ripple receiving action of the equipment receiving this digital modulation signals RFsum can be verified.
Above-mentioned digital modulation signals generation device 30 side to the data rows of 2 for carrying out add operation carry out resampling process and make it consistent with the sample rate of the opposing party, but digital modulation signals generation device 30 ' as shown in figure 11, if two groups of resampling handling parts 40A, 40B arranging same structure can carry out resampling process to the data rows of 2 for carrying out add operation respectively, then can make it have the larger degree of freedom relative to the restriction of the frequency field on the hardware of the reprocessing of D/A converter 44.
Such as, the frequency field of allowing in the process after D/A converter 44 is Fa ± Δ F, when the 1st above-mentioned sample rate Fin1 does not enter this frequency field, need to carry out rate conversion to two data rows D1, D2.
Now, ', n, n ' following condition of demand fulfillment about integer m, m.
Fa-ΔF≤Fin1·（m/n）=Fin2·（m’/n’）≤Fa+ΔF
If find to meet integer m, m of above-mentioned condition ', n, n ', then at resampling handling part 40A, the resampling process that speed ratio is m/n is carried out to data rows D1, at resampling handling part 40B, the resampling process that speed ratio is m '/n ' is carried out to data rows D2, the sample rate of two data can be made thus consistent in the frequency field of allowing, even if there is the restriction on hardware, also can carry out conjunction ripple to the data rows of various sample rate, and high versatility can be given.
In addition, in above-mentioned explanation, the situation improving sample rate (m > n) is illustrated, but equally also can carry out m < n(reduce sample rate) resampling process, now, as long as obtain the filter coefficient for obtaining the interpolated data to each timing carrying out m decile between former data, and when often inputting n data rows switching filter coefficient so that every n temporally sequence output m the interpolated data of each timing.
Such as, initial treatment in the resampling processing unit 20 of Fig. 1 structure when m=4, n=5 as described in Fig. 5 (a) and (b) shown in, obtain about to the interpolated data of each timing Q1 ~ Q3 carrying out 4 deciles between 2 data, and utilize the 1st filter coefficient h(1,1) ~ h(1,8) export initial resampling value Dout(1)=Din(4).
And, due to the timing Q1 ' that next resampling value is delay 5 Δ T (=n Δ T), therefore the Δ T after such as Figure 12 (a) next one input Data Update is through timing Q1(t=Ts+ Δ T), if Figure 12 (b) is by switching to the 2nd group of filter coefficient h(2 by filter coefficient, 1) ~ h(2,8) export the 2nd resampling value Dout(2).Similarly in the following, in input 5 periods of Data Update, switching filter coefficient so as every 5 temporally sequence export 4 interpolated datas, carry out the resampling process that speed ratio is 4/5 thus, and increase the degree of freedom of data rows combination further by using it for above-mentioned digital modulation signals generation device 30,30 '.
Claims (6)
1. a resampling processing unit, it carries out to the data rows inputted with predetermined sample rate Fin the resampling process that speed ratio is m/n, and wherein, m, n are integer, it is characterized in that, this resampling processing unit possesses:
FIR (FiniteImpulseResponse, finite impulse response (FIR)) type digital filter (21,21A, 21B), it carries out the multiply-add operation that up-to-date P data of the data rows of described input and P are the filter coefficient of a group, and is exported successively by its operation result;
Filter coefficient means of storage (25), it prestores m × P filter coefficient; And
Filter coefficient switching member (26), it selects described P be the filter coefficient of a group and give described digital filter with the resampling speed of Finm/n from described filter coefficient means of storage successively, so that during n data are input into described digital filter, every n, temporally sequence output m is individual about interpolated data regularly.
2. resampling processing unit as claimed in claim 1, is characterized in that,
Being stored in the filter coefficient of described filter coefficient means of storage, is the filter coefficient needed for data interpolating carried out the input interval of the data being input into described digital filter in each timing of m decile.
3. a digital modulation signals generation device, it has:
1st data rows output link (31), it exports the 1st data rows for modulating with the 1st sample rate;
2nd data rows output link (32), it exports the 2nd data rows for modulating with the 2nd sample rate different from the 1st sample rate; And
Resampling handling part (40,40A, 40B), it is in order to coordinate described 1st data rows with the sample rate of the 2nd data rows, and carry out to the data rows of at least one party the resampling process that speed ratio is m/n, wherein, m, n are integer,
This digital modulation signals generation device carries out add operation to the data rows coordinated by described resampling process and sample rate, and produces the signal being carried out digital modulation by the data rows of this add operation, and it is characterized in that, described resampling handling part possesses:
FIR (FiniteImpulseResponse, finite impulse response (FIR)) type digital filter (21,21A, 21B), its carry out with up-to-date P the data of data rows of sample rate Fin input and P individual be the multiply-add operation of the filter coefficient of a group, and its operation result to be exported successively;
Filter coefficient means of storage (25), it prestores m × P filter coefficient; And
Filter coefficient switching member (26), it selects described P be the filter coefficient of a group and give described digital filter with the resampling speed of Finm/n from described filter coefficient means of storage successively, so that during n data are input into described digital filter, every n, temporally sequence output m is individual about interpolated data regularly.
4. digital modulation signals generation device as claimed in claim 3, is characterized in that,
Being stored in the filter coefficient of described filter coefficient means of storage, is the filter coefficient needed for data interpolating carried out the input interval of the data being input into described digital filter in each timing of m decile.
5. a resampling processing method, it carries out to the data rows inputted with predetermined sample rate Fin the resampling process that speed ratio is m/n, and wherein, m, n are integer, it is characterized in that, this resampling processing method comprises:
Carry out the step that up-to-date P data of the data rows of described input and P are the multiply-add operation of the filter coefficient of a group;
Use the step exporting FIR (FiniteImpulseResponse, finite impulse response (FIR)) the type digital filter of its operation result successively;
Prestore the filter coefficient storing step of m × P filter coefficient; And
The described P of described storage is selected to be the filter coefficient of a group and to give described digital filter with the resampling speed of Finm/n successively, so that during n data are input into described digital filter, every n, temporally sequence output m is individual about interpolated data regularly, carries out the step that speed ratio is the resampling process of m/n thus.
6. resampling processing method as claimed in claim 5, is characterized in that,
The filter coefficient stored in described filter coefficient storing step is the filter coefficient needed for data interpolating carried out the input interval of the data being input into described digital filter in each timing of m decile.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2011-285939 | 2011-12-27 | ||
| JP2011285939A JP2013135401A (en) | 2011-12-27 | 2011-12-27 | Resampling device, digital modulation signal generation device using the same, and resampling method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103188186A CN103188186A (en) | 2013-07-03 |
| CN103188186B true CN103188186B (en) | 2016-04-13 |
Family
ID=48679156
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210470733.2A Active CN103188186B (en) | 2011-12-27 | 2012-11-20 | Resampling processing unit and method and digital modulation signals generation device |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JP2013135401A (en) |
| CN (1) | CN103188186B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104734668B (en) * | 2015-03-31 | 2017-11-07 | 中国科学院微电子研究所 | A kind of interpolation filter |
| CN106972840B (en) * | 2017-02-28 | 2020-05-05 | 深圳市鼎阳科技股份有限公司 | Sampling rate conversion method and device |
| CN108572266B (en) * | 2017-12-11 | 2020-09-15 | 深圳市鼎阳科技股份有限公司 | Waveform generating device |
| JPWO2019131162A1 (en) * | 2017-12-27 | 2020-12-17 | ソニーセミコンダクタソリューションズ株式会社 | Amplifier and signal processing circuit |
| CN110335615B (en) * | 2019-05-05 | 2021-11-16 | 北京字节跳动网络技术有限公司 | Audio data processing method and device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1259006A2 (en) * | 2001-05-18 | 2002-11-20 | SANYO ELECTRIC Co., Ltd. | Transmission timing control device, digital roll-off filter and mobile radio terminal for digital radio communication |
| JP2008113123A (en) * | 2006-10-30 | 2008-05-15 | Fujitsu Ten Ltd | Sampling conversion device |
| CN101339610A (en) * | 2008-08-13 | 2009-01-07 | 哈尔滨工业大学 | Particle filtering resampling method suitable for non-linear probabilistic system posture |
| JP2009232079A (en) * | 2008-03-21 | 2009-10-08 | Japan Radio Co Ltd | Interpolation filter |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH02149011A (en) * | 1988-11-30 | 1990-06-07 | Toshiba Corp | Sampling frequency converting device |
| JP4722266B2 (en) * | 2000-08-16 | 2011-07-13 | 富士通セミコンダクター株式会社 | Oversampling FIR filter, oversampling FIR filter control method, semiconductor integrated circuit having oversampling FIR filter, and communication system for transmitting data filtered by oversampling FIR filter |
| JP2003051736A (en) * | 2001-06-01 | 2003-02-21 | Sanyo Electric Co Ltd | Digital roll-off filter and mobile radio terminal employing the same |
| JP4413697B2 (en) * | 2004-07-13 | 2010-02-10 | アンリツ株式会社 | Waveform generator |
| JP4376222B2 (en) * | 2005-10-27 | 2009-12-02 | シャープ株式会社 | Wave shaping digital filter circuit |
-
2011
- 2011-12-27 JP JP2011285939A patent/JP2013135401A/en active Pending
-
2012
- 2012-11-20 CN CN201210470733.2A patent/CN103188186B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1259006A2 (en) * | 2001-05-18 | 2002-11-20 | SANYO ELECTRIC Co., Ltd. | Transmission timing control device, digital roll-off filter and mobile radio terminal for digital radio communication |
| JP2008113123A (en) * | 2006-10-30 | 2008-05-15 | Fujitsu Ten Ltd | Sampling conversion device |
| JP2009232079A (en) * | 2008-03-21 | 2009-10-08 | Japan Radio Co Ltd | Interpolation filter |
| CN101339610A (en) * | 2008-08-13 | 2009-01-07 | 哈尔滨工业大学 | Particle filtering resampling method suitable for non-linear probabilistic system posture |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103188186A (en) | 2013-07-03 |
| JP2013135401A (en) | 2013-07-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103188186B (en) | Resampling processing unit and method and digital modulation signals generation device | |
| US4748578A (en) | Process and apparatus for translating the sampling rate of a sampling sequence | |
| US7394415B2 (en) | Time-interleaved analog-to-digital converter and high speed signal processing system using the same | |
| CN101296014B (en) | Broadband digital beam forming method | |
| CN105991137A (en) | Systems and methods of variable fractional rate digital resampling | |
| CN103259602B (en) | Method and system for signal generation | |
| CN105116424A (en) | Beidou satellite navigation signal simulation source based on file generation and playback | |
| CN108111450B (en) | Method for simulating dynamic characteristics of target by adopting incoherent demodulation | |
| JP4542935B2 (en) | A / D converter | |
| US6442580B1 (en) | Resampling method and resampler circuit | |
| CN102468831A (en) | Using multi-level pulse width modulated signal for real time noise cancellation | |
| CN107484197B (en) | Wireless fading channel simulation device and method of any Doppler power spectrum | |
| US7830217B1 (en) | Method and system of vector signal generator with direct RF signal synthesis and parallel signal processing | |
| EP3191912B1 (en) | Generation of high-rate sinusoidal sequences | |
| US20080224750A1 (en) | Digital delay architecture | |
| KR20010101889A (en) | Digital-Analog Converter and Method, And Data Interpolation Device and Method | |
| US8242829B1 (en) | Multichannel interpolator | |
| CN1203620C (en) | Method for system-independent digital evaluation of mobile communication reception signals of various mobile radio standards | |
| JP3074603B2 (en) | Fading simulator | |
| CN103716018B (en) | Digital broadband excitation source implement device and method | |
| JP5450580B2 (en) | Digital modulation signal generator and digital modulation signal generation method | |
| JP4413697B2 (en) | Waveform generator | |
| CN115085693B (en) | Multi-channel multi-phase interpolation processing architecture | |
| KR20150040111A (en) | Apparatus and method for coverting data rate | |
| CN110190829B (en) | Filter and filtering method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |