CN103493039A - Data processing method and related device - Google Patents
Data processing method and related device Download PDFInfo
- Publication number
- CN103493039A CN103493039A CN201280000317.4A CN201280000317A CN103493039A CN 103493039 A CN103493039 A CN 103493039A CN 201280000317 A CN201280000317 A CN 201280000317A CN 103493039 A CN103493039 A CN 103493039A
- Authority
- CN
- China
- Prior art keywords
- base
- grades
- butterfly computations
- butterfly
- roads
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
- G06F17/141—Discrete Fourier transforms
- G06F17/142—Fast Fourier transforms, e.g. using a Cooley-Tukey type algorithm
Landscapes
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Discrete Mathematics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Traffic Control Systems (AREA)
- Complex Calculations (AREA)
Abstract
The invention provides a data processing method and a related device. The data processing method comprises the steps of inputting y channel of data to be subject to z-stage radix-y butterfly computation into a radix-y butterfly arithmetic unit for z-stage radix-y butterfly computation, y and z being positive integers; in the case where the z-stage radix-y butterfly computation is a last stage of radix-y butterfly computation, outputting result data of the z-stage radix-y butterfly computation; in the case where the z-stage radix-y butterfly computation is not the last stage of radix-y butterfly computation, inputting the result data of the z-stage radix-y butterfly computation into the radix-y butterfly arithmetic unit for (z+1)-stage radix-y butterfly computation. The technical solutions provided by the embodiments of the present invention are conducive to reducing resources consumed during the butterfly computation of the data.
Description
Data processing method and relevant apparatus
Technical field
The present invention relates to technical field of data processing, and in particular to data processing method and relevant apparatus.Background technology
At present, channel estimation, signal analysis and signal in the field of communications needs to carry out data processing during comparing etc., the method for needing widely to apply to Fourier analysis, and the data after Fourier analysis are handled accordingly so that the process such as channel estimation, signal analysis obtains cylinder.Fourier analysis can pass through Fast Fourier Transform (FFT)（FFT, Fast Fourier Transform) operation processing unit realizes also there is similar application in other data processing fields.
FFT operation processing units in existing data handling system can pass through many algorithms（Such as butterfly computation of base 2, radix-4 butterfly computing and mixed base scheduling algorithm）To realize.But, current butterfly computation resource utilization is also relatively low, and such as 256 point datas need 4 grades of radix-4 butterfly computings, then needs to expend 4 sets of computing circuits and storage resource, resource cost is relatively more.The content of the invention
The embodiment of the present invention provides a kind of data processing method and relevant apparatus, to reduce to the resource consumption needed for data progress butterfly computation.
On the one hand the embodiment of the present invention provides a kind of data processing method, it may include：
The data input base y butterfly computations device of the pending z grades of bases y butterfly computations in y roads is subjected to z grades of base y butterfly computations, wherein, the y and z are positive integer；
In the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained is exported；
In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained inputs the base y butterfly computations device and carries out z+1 grades of base y butterfly computations.
On the other hand the embodiment of the present invention also provides a kind of data processing equipment, including：
Controller and base y butterfly computation devices；
Wherein, the controller, for by the data input institute of the pending z grades of bases y butterfly computations in y roads
State base y butterfly computations device and carry out z grades of base y butterfly computations, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, export the result data that the z grades of bases y butterfly computations are obtained；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained, input the base y butterfly computations device and carry out z+1 grades of base y butterfly computations, wherein, the y and z are positive integer.
On the other hand the embodiment of the present invention also provides and is disposed in a kind of access device, the access device just like the data processing equipment described in above-described embodiment.
On the other hand the embodiment of the present invention also provides and is disposed in a kind of user equipment, the user equipment just like the data processing equipment described in above-described embodiment.
On the other hand the embodiment of the present invention also provides a kind of computer-readable storage medium,
The computer-readable storage medium has program stored therein, and described program includes the part or all of step of such as above-mentioned data processing method when performing.
Therefore, the embodiment of the present invention is by the data of the pending z grades of bases y butterfly computations in y roads, input base y butterfly computations device and carry out z grades of base y butterfly computations, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained inputs base y butterfly computations device and carries out z+1 grades of base y butterfly computations.Because the mechanism is multiplexed to base y butterfly computation devices, two-stage base y butterfly computations are multiplexed same base y butterfly computation devices, the time-multiplexed base y butterfly computations device of such as availability data input carries out computing, so advantageously reduce the time of butterfly computation and reduce resource consumption.Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, cylinder will be made to the required accompanying drawing used in embodiment and description of the prior art below singly to introduce, apparently, drawings in the following description are only some embodiments of the present invention, for those of ordinary skill in the art, without having to pay creative labor, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1-a are the butterfly computation model schematics of a kind of base 2 provided in an embodiment of the present invention；
Fig. 1-b are a kind of radix-4 butterfly operational model schematic diagrames provided in an embodiment of the present invention；
Fig. 2 is a kind of configuration diagram of 4 grades of radix-4 butterfly computing circuits provided in an embodiment of the present invention；
Fig. 3 is a kind of schematic diagram of the read-write operation of shift register；
Fig. 4 is a kind of schematic diagram of data processing method provided in an embodiment of the present invention；
Fig. 5 is a kind of schematic diagram of data processing equipment provided in an embodiment of the present invention；
Fig. 6 is a kind of schematic diagram of shift register operations provided in an embodiment of the present invention；
Fig. 7-a are a kind of schematic diagrames of butterfly computation sequence arrangement provided in an embodiment of the present invention；
Fig. 7-b are the schematic diagrames of another butterfly computation sequence arrangement provided in an embodiment of the present invention；Fig. 8-a are a kind of schematic diagrames of data processing equipment provided in an embodiment of the present invention；
Fig. 8-b are the schematic diagrames of another data processing equipment provided in an embodiment of the present invention；
Fig. 9 is the schematic diagram of a kind of electronic equipment provided in an embodiment of the present invention.
Embodiment
The embodiment of the present invention provides a kind of data processing method and relevant apparatus, to reduce to the resource consumption needed for data progress butterfly computation.
To enable goal of the invention, feature, the advantage of the present invention more obvious and understandable, below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme in the embodiment of the present invention is clearly and completely described, obviously, the embodiments described below are only a part of embodiment of the invention, and not all embodiment.Based on the embodiment in the present invention, all other embodiment that those of ordinary skill in the art are obtained under the premise of creative work is not made belongs to the scope of protection of the invention.
First, it can be derived from by FFT formula：
W— 1
N ,
2 N
j kn
[x(n) +WN N
n=0
.2π j
, - /—— kn kN r i i, kc is even number
wkn = e N
Wherein, WN 2=(- l=k k are strange severe several, and Q≤k≤N -1 is to treatFFTThe data of computing, are O)FFTOperation result data, N is to treat
The points of the data of FFT computings.Wherein, the butterfly computation mode of base 2 can be obtained from above can be as shown in Fig. 1-a.
It is appreciated that the FFT flowing water of 2^, data can be obtained by the k grades of butterfly computation of base 2 cascades；
The butterfly computation of two-stage base 2 is equivalent to one-level radix-4 butterfly computing, and N point input datas are divided into 4 parts, take the corresponding data of each several part to be calculated every time by radix-4 butterfly computing, and radix-4 butterfly computing mode can be as shown in Fig. 1-b.Wherein, ^ point datas are carried out！^ computings, need k grades of radix-4 butterfly computing cascades, and such as 256 point data FFT computings need 4 grades of radix-4 butterflies.
Referring to Fig. 2, Fig. 2 is a kind of configuration diagram of 4 grades of radix-4 butterfly computing circuits of citing.Typically in data are carried out with radix-4 butterfly calculating process, assuming that there is N point datas to need to carry out radix-4 butterfly computing, and 1 point data of write-in needs 1 clock, then 3N/4 point datas are sequentially stored into three shift registers by the preceding 3N/4 cycle, last N/4 point datas are inputted to start, radix-4 butterfly computing circuit starts to carry out computing to data, the time that such then radix-4 butterfly computing circuit has nearly 3/4 is in idle state, calculation resources utilization rate is relatively low, such as 256 point datas need 4 grades of radix-4 butterfly computings, then need to expend 4 sets of computing circuits and storage resource, resource
Expend relatively many.
For ease of understanding the working mechanism of framework shown in Fig. 2, below exemplified by carrying out one-level radix-4 butterfly calculating process, common N point datas are inputted, it is assumed that each clock inputs a point data, it is necessary to N number of clock cycle；The preceding N/4 clock cycle, input data Slow is deposited to shift register（R0 );N/4 ~ N/2 clock cycle, input data Slow is deposited to shift register（ R1 );N/2 ~ 3N/4 clock cycle, input data Slow is deposited to shift register（R2 );The last N/4 cycle, data are read from three shift registers simultaneously and radix-4 butterfly arithmetic unit are input to N/4 last data syn-chronization and carry out butterfly computation, computing obtains N/4,4 road data simultaneously, need the FFT computings for continuing to do N/4 points, the first via N/4 that computing is obtained is sent to rear class butterfly computation device, its excess-three road is written back to during the corresponding Slow of this grade of butterfly computation device deposits, N/4 data of operation result etc. the first via are all sent to after rear class radix-4 butterfly arithmetic unit, N/4 data are read from shift register -0 be sent to rear class radix-4 butterfly arithmetic unit again, then N/4 data of correspondence are read from shift register -1 again and is sent to rear class radix-4 butterfly arithmetic unit, read corresponding data from shift register -2 again afterwards and be sent to rear class radix-4 butterfly arithmetic unit, by that analogy.
Below so that rear class does the radix-4 butterfly computing of 8 point datas as an example, the operation to shift register can be shown such as Fig. 3.As shown in P1 ~ P6 states in Fig. 3, A1 ~ A8 is 8 point datas, and A_1 ~ A_8 is data of the A1 ~ A8 after the butterfly computation of base 4, and data are sequentially stored into 3 shift registers by the preceding 3N/4 cycle（This is sentenced realized with shift register exemplified by）, the rear N/4 cycle（P7, P8) complete butterfly computation, A7 and A8 do not bother to see me out shift register, but the data directly preserved with before are sent to radix-4 butterfly arithmetic unit.Latter 3 in obtain 4 result datas are sent back to shift register after each computing（Backfill, as shown in P8）, by the 1st（That is A_l, A_2) the lower 1 grade of progress radix-4 butterfly computing of feeding（Such as P9).Then N/4 data feeding next stage radix-4 butterfly internalarithmetic is taken out from 3 shift registers successively, now, next group of N point input datas can be write toward shift register simultaneously（That is B1 ~ B8, as shown in P10 and P11）.
Wherein, in Fig. 3, P1 ~ P6 is data preparation stage, and now radix-4 butterfly arithmetic unit is idle, and P7 ~ P9 is the data operation stage, and P10 ~ P11 is write back data state.The calculating process of the data of other points is then by that analogy.
The scheme that the embodiment of the present invention is proposed advantageously reduces the resource consumption carried out to data needed for butterfly computation.
Below by specific embodiment, it is described in detail respectively.
One embodiment of the data processing method of the present invention, wherein, this method may include：By the data of the pending z grades of bases y butterfly computations in y roads, input base y butterfly computations device carries out z grades of base y butterfly computations, wherein, the y and z are positive integer；In the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained inputs base y butterfly computations device and carries out 1 grade of base y butterfly computation of z+.
Referring to Fig. 4, a kind of data processing method provided in an embodiment of the present invention, it may include herein below：401st, by the data of the pending z grades of bases y butterfly computations in y roads, input base y butterfly computations device carries out z grades of base y butterfly computations；
It is appreciated that y and z is positive integer.Assuming that needing at 256 points（ 44) data progress FFT computings, if carrying out FFT computings by the butterfly computation of base 2, then y is 2, if carrying out FFT computings by radix-4 butterfly computing, then y is 4;If carrying out FFT computings by the butterfly computation of base 3, then y is 3;If carrying out FFT computings by the butterfly computation of base 8, then y is 8;If carrying out FFT computings by the butterfly computation of base 16, shellfish ' Jy be 16, by that analogy.
In some embodiments of the invention, for example there is N point FFT computings, handled according to base y butterfly computation devices, then need to carry out M (M=log altogetheryN) level base y butterfly computations, z is less than or equal to ^1.General, N is 2 positive integer power herein, when N is not 2 positive integer power, N can be supplemented to the positive integer power for 2, such as can by mend 0 method, mended on the basis of original N point datas several 0 so that mend 0 after data always points for 2 positive integer power.
402nd, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained inputs base y butterfly computations device and carries out z+1 grades of base y butterfly computations.
It can be seen that, by the data of the pending z grades of bases y butterfly computations in y roads in the present embodiment, input base y butterfly computations device and carry out z grades of base y butterfly computations, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained inputs base y butterfly computations device and carries out z+1 grades of base y butterfly computations.Because the mechanism is to base y
Butterfly computation device is multiplexed, two-stage base y butterfly computations are multiplexed same base y butterfly computation devices, the time-multiplexed base y butterfly computations device of such as availability data input carries out computing, so advantageously reduce the time of butterfly computation and reduce resource consumption.
In some embodiments of the invention, intermediate result data that may be incorporated into Slow storages to store base y butterfly computations etc..For example, in the case where above-mentioned z grades of base y butterfly computations are not afterbody base y butterfly computations, the obtained y roads result data Zhong Mei roads result data of the z grades of base y butterfly computations can be divided into y part, and this y part is respectively written into y Ge Slowcun areas；The y part that the y Ge Slowcun areas Zhong Gaimei roads result data is divided into will be then respectively written into, as the data of the pending z+1 grades of bases y butterfly computations in y roads, base y butterfly computations device is inputted and carries out 1 grade of base y butterfly computation of z+.
In addition, in the case where z+1 grades of base y butterfly computations are last 1 grade of base y butterfly computations, the result data of exportable above-mentioned z+1 grades of bases y butterfly computations；And in the case where the z+1 grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that can obtain the z+1 grades of base y butterfly computations inputs above-mentioned base y butterfly computation devices again（Or other base y butterfly computations devices) carry out z+2 grades of base y butterfly computations.Wherein, in the case where above-mentioned z+1 grades of base y butterfly computations are not afterbody base y butterfly computations, the obtained y roads result data Zhong Mei roads result data of the z+1 grades of base y butterfly computations can be for example divided into y part, and this y part is respectively written into y Ge Slowcun areas；The y part that the y Ge Slowcun areas Zhong Gaimei roads result data is divided into will be then respectively written into, it is used as the data of the pending z+2 grades of bases y butterfly computations in y roads, input base y butterfly computations device and carry out z+2 grades of base y butterfly computations, if result data subsequently also needs to carry out base y butterfly computations, can accordingly circular treatment to complete all grades of base y butterfly computations.
Wherein, above-mentioned y Ge Slowcun areas can be located in y Slow storage respectively；Or, the y Ge Slowcun areas may be alternatively located at least one Slow storage, wherein, the Slow storages for example can select displacement Slow storages（Such as shift register or other displacement Slow storages）.
It is appreciated that, if also failing to obtain FFT operation results after carrying out this grade of base y butterfly computation, i.e. in the case where this grade of base y butterfly computation is not afterbody base y butterfly computations, above-mentioned mechanism circular treatment can be based on, read out respectively from above-mentioned y Ge Slowcun areas, write the y part that the result data Zhong Mei roads result data of this grade base y butterfly computation of the y Ge Slowcun areas in is divided into, the data of the pending lower 1 grade of base y butterfly computations in y roads are partly used as using this y, input base y butterfly computations device and carry out next stage base y butterflies fortune
Calculate, circular treatment most obtains the FFT operation results of data at last accordingly.
For example, with to 256 points（44Point）Data are carried out exemplified by FFT computings, 256 points（ 44) data need carry out 4 grades of radix-4 butterfly computings.Wherein, the 1st grade by 256 point datas carry out radix-4 butterfly computing, obtain 64 point datas/group 4 groups of data；2nd grade by 64 point datas/group 4 groups of data carry out radix-4 butterfly computing, obtain 16 point datas/group 16 groups of data；And 3rd level by 16 point datas/group 16 groups of data carry out radix-4 butterfly computing, obtain 4 point datas/group 64 groups of data；And the 4th grade by 4 point datas/group 64 groups of data carry out radix-4 butterfly computing, obtain the FFT operation results of 256 point datas.It is appreciated that the FFT computings mode of the data of other points is by that analogy.
Referring to Fig. 5, the data processing operation device in Fig. 5 includes：4 depth are at least 64 and (at least store 64 point datas）Shift register and a radix-4 butterfly arithmetic unit, operation result data are write back in the shift register after completing computing every time, when next stage radix-4 butterfly computing starts, the data for reading pending subordinate's radix-4 butterfly computing from the shift register again carry out subordinate's radix-4 butterfly computing, until obtaining the output of FFT operation results.
In some embodiments of the invention, for example can be after the data of all pending every grade of base y butterfly computations all carry out this grade of base y butterfly computation, then carry out the next stage base y butterfly computations of the data of all pending next stage base y butterfly computations.
For example, after the result data for the z+1 grades of base y butterfly computations in y*y roads that above-mentioned z+1 grades of bases y butterfly computations are obtained is carried out, in the case where z+1 grades of base y butterfly computations are not afterbody base y butterfly computations, can the result data Zhong Mei roads result datas of z+1 grades of Jiang Gai y*y roads base y butterfly computations be divided into y part, using this y part as the data of the pending z+2 grades of bases y butterfly computations in y roads, above-mentioned base y butterfly computation devices are inputted（Or other butterfly computation devices) carry out z+2 grades of base y butterfly computations, for example, the obtained y*y roads result data Zhong Mei roads result data of the z+1 grades of base y butterfly computations can be divided into y part, and this y part is respectively written into y Ge Slowcun areas；The y part that the y Ge Slowcun areas Zhong Gaimei roads result data is divided into will be then respectively written into, it is used as the data of the pending z+2 grades of bases y butterfly computations in y roads, input base y butterfly computations device and carry out z+2 grades of base y butterfly computations, if the result data of z+2 grades of base y butterfly computations also needs to carry out base y butterfly computations, can accordingly circular treatment to complete all grades of base y butterfly computations, the data of all pending every grade of base y butterfly computations are all carried out after this grade of base y butterfly computation, the next stage base y butterfly computations of the data of all pending next stage base y butterfly computations are carried out again.
With to 256 points（44Point）Data carry out FFT computings exemplified by, 256 point datas need carry out 4 grades of radix-4 butterfly computings, wherein, the 1st grade completion 256 point datas radix-4 butterfly computing after obtain 64 point datas/group 4 groups of data；Then carry out the 2nd grade of 64 point data/group 4 groups of data radix-4 butterfly computing（Wherein, 64 point datas/group 4 groups of data in 4 parts being divided into of every group of data, 4 parts that every group of data are all divided into input radix-4 butterfly arithmetic unit and carry out the 2nd grade of butterfly computation of base 4 as the data of the pending 2nd grade of radix-4 butterfly computing in 4 roads）, obtain 16 point datas/group 16 groups of data；64 point datas/group 4 groups of data radix-4 butterfly computing all complete after, then carry out 16 point datas of 3rd level/group 16 groups of data radix-4 butterfly computing, obtain 4 point datas/group 64 groups of data；16 point datas/group 16 groups of data radix-4 butterfly computing all complete after, then carry out the 4th grade 4 point datas/group 64 groups of data radix-4 butterfly computing, obtain the FFT operation results of 256 point datas.The FFT computings mode of the data of other points is by that analogy.
, for example can also be formerly by pending kth level in other embodiments of the present invention（Such as the 2nd or 3rd level or non-last 1 grade of other levels）The data of base y butterfly computations are divided into several parts（Such as 2 parts, 3 parts or more part）, first carry out the kth level base y butterfly computations of certain partial data therein, after the completion of remaining all grades of base y butterfly computations of certain partial data are all calculated, the kth level base y butterfly computations of another part of the data of pending kth level base y butterfly computations are carried out again, until remaining all grades of base y butterfly computations of another part data all calculate completion, if there is another partial data in pending kth level, the kth level base y butterfly computations a part of again of the data of the pending kth level base y butterfly computations are carried out again, until remaining all grades of base y butterfly computations of another partial data all calculate completion, by that analogy, it is finally completed all grades of base y butterfly computations of all data, obtain FFT result of calculations.
For example, first carrying out y circuit-switched datas in the y roads result data that z grades of base y butterfly computations are obtained, the wherein m roads result data in y Ge Slowcun areas is first written to（It is written to the result data that the result data in y Ge Slowcun areas is namely first calculated）Y part being divided into respectively of every circuit-switched data, z+1 grades of base y butterfly computations are carried out to input above-mentioned base y butterfly computations device as the data of the pending z+1 grades of bases y butterfly computations in y roads, the result data of the z+1 grades of base y butterfly computations in y*m roads is can obtain；After the result data of the z+1 grades of base y butterfly computations in y*m roads is obtained, by in the result data of above-mentioned y roads, the y part that the every circuit-switched data being written to afterwards in the remaining n roads result data in the y Ge Slowcun areas is divided into respectively, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input base y butterfly computations device and carry out z+1 grades of base y butterfly computations, the result data of the z+1 grades of base y butterfly computations in y*n roads is can obtain, wherein
Above-mentioned m is less than y, and above-mentioned m is equal to y with n sums;Above-mentioned m is less than or equal to or more than n.
Wherein, in the case where z+1 grades of base y butterfly computations are not afterbody base y butterfly computations, in the y roads result data for obtaining above-mentioned z grades of bases y butterfly computations, the y part that the remaining n roads result data Zhong Mei roads result data in the y Ge Slowcun areas is divided into respectively is written to afterwards, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input before z+1 grades of base y butterfly computations of the base y butterfly computations device progress, the result data Zhong Mei roads result data of z+1 grades of y*m roads base y butterfly computations is divided into y part, the data of the pending z+2 grades of bases y butterfly computations in y roads are partly used as using this y, input base y butterfly computations device and carry out z+2 grades of base y butterfly computations, it can obtain the result data of the z+2 grades of base y butterfly computations in y*y*m roads.
For example, still with to 256 points（44Point）Data carry out FFT computings exemplified by, 256 point datas need carry out 4 grades of radix-4 butterfly computings, wherein, the 1st grade complete 256 point datas radix-4 butterfly computing after obtain 64 point datas/group 4 groups of data；First carry out the 2nd grade of 64 point data/group 4 groups of data in wherein two groups data radix-4 butterfly computing, obtain 16 point datas/group 8 groups of data；64 point datas/group 2 groups of data radix-4 butterfly computing all after the completion of, carry out again 16 point datas/group 8 groups of data 3rd level radix-4 butterfly computing, obtain 4 point datas/group 32 groups of data, then carry out again 4 point datas/group 32 groups of data the 4th grade of radix-4 butterfly computing, obtain the FFT operation results of 128 point datas；Then, carry out the 2nd grade of 64 point data/group 4 groups of data in remaining two groups of data radix-4 butterfly computing, obtain 16 point datas/group 8 groups of data；64 point datas/group remaining 2 groups of data radix-4 butterfly computing all after the completion of, 16 point datas obtained again/group 8 groups of data 3rd level radix-4 butterfly computing, obtain 4 point datas/group 32 groups of data, 4 point datas then obtained again/group 32 groups of data the 4th grade of radix-4 butterfly computing, obtain the FFT operation results of 128 point datas, so far, 256 point datas need the 4 grades of radix-4 butterfly computings carried out to be fully completed.It is appreciated that the FFT computing modes of the data of other points can be by that analogy.
Wherein, scheme of the embodiment of the present invention to the data of pending same level base y butterfly computations during carrying out with one-level base y butterfly computations, the data that first write-in Slow is deposited in area can first be inputted base y butterfly computations device and carry out computing, rear write-in Slow deposited the progress computing of base y butterfly computations device is inputted after the data in area.
It is understood that the step in such scheme can be implemented under the control of the controller, its part or all of function can be also integrated in base y butterfly computations device and Slow storages by the controller independently of base y butterfly computations device and Slow storages.
For ease of being better understood from and implementing such scheme, it is described below exemplified by carrying out the process of radix-4 butterfly computing of 256 point datas.
Referring to Fig. 6, in a kind of feasible embodiment, the shift register storage operation mechanism of the data processing equipment of framework shown in Fig. 5 can be as follows, wherein, example is mainly saved as using different Slow storages progress Slow with every circuit-switched data herein and is described.
First, 4 circuit-switched datas on input 64 point datas/road carry out the radix-4 butterfly computing of 256 point datas, obtain 4 tunnel result of calculations and deposit in respectively in four shift registers（Shift register R0 ~ R3).
4 kinds of shadings in Fig. 6 represent 4 circuit-switched datas that radix-4 butterfly computing is obtained respectively.Wherein, 4 parts of radix-4 butterfly computings to carry out next stage are divided into 1 circuit-switched data, in order to be able to which reading with 1 tunnel simultaneously（Identical shading in figure）4 numbers according to carry out radix-4 butterfly computing, when depositing data, the shift register for storing the circuit-switched data can be switched over after 16 data for often having deposited 1 tunnel, ultimately form in low 16 addresses that preceding 16 data in the 1st circuit-switched data have shift register R0, there is shift register R1 16 ~ address of address 31 in the 17th ~ 32 data in the 1st circuit-switched data, there is shift register R2 32 ~ address of address 47 in the 33rd ~ 48 data in the 1st circuit-switched data, there is shift register R3 48 ~ address of address 63 in the 49th ~ 64 data in the 1st circuit-switched data；Other circuit-switched datas（Other shadings in figure）Storage mode by that analogy.
When needing to carry out subordinate's radix-4 butterfly computing, the part feeding radix-4 butterfly arithmetic unit of same shading is read from 4 shift registers respectively, the data obtained after computing can be stored in the corresponding address read out again（If shift register is sufficiently large, the data obtained after computing can also be stored in other clear positions）, for example, the point data of 1st tunnel 64 carries out the data obtained after radix-4 butterfly computing, still it is storable in shift register R0 0 ~ 15 address location, shift register R1 16 ~ 31 address locations, shift register R 2 32 ~ 47 address locations and shift register R3 48 ~ 63 address locations, the storage mode on other several roads is by that analogy, in 4 tunnels that i.e. radix-4 butterfly computing is obtained, the data on same 1 tunnel, which are divided into 4 parts, to be present in different shift registers, to ensure subsequently read out to carry out subordinate's radix-4 butterfly computing simultaneously.The address storage mode for the data that the third level and fourth stage radix-4 butterfly computing are obtained can by that analogy, and here is omitted.
To obtain as far as possible high operation efficiency, also radix-4 butterfly priority of operations order can be carried out rationally cleverly to arrange.The mode of several arrangement radix-4 butterfly priority of operations orders provided in an embodiment of the present invention is described below.
Obtained it is appreciated that the premise of rear class butterfly computation is prime butterfly computation result, it is contemplated that meter
Situations such as delay and control delay is calculated, the sequencing that each circuit-switched data of proper arrangement carries out every grade of butterfly computation can be carried out as needed.
Below with to 256 point datas（44) carry out FFT computings exemplified by, the sequence arrangement that a kind of butterfly is calculated is for example shown in Fig. 7-a.
1. 256 points of first order radix-4 butterfly computing is performed, 4 circuit-switched datas on 64 point datas/road are obtained, 64 clock cycle are needed（Cyde );
2. wherein 1 tunnel is performed（Such as the 1st tunnel）64 point datas second level radix-4 butterfly computing, obtain 4 circuit-switched datas on 16 point datas/road, need 16 cycle;
3. the third level radix-4 butterfly computing of 4 circuit-switched datas on above-mentioned 16 point datas/road is performed, 16 circuit-switched datas on 4 point datas/road are obtained, 16 cycle are needed;
4. the radix-4 butterfly computing of the above-mentioned point data of 16 tunnel 4 is performed, it is necessary to 16 cycle;
So far, 4 circuit-switched datas that 256 points of the 1st grade of radix-4 butterfly computing is obtained are performed（64 point datas/road）Wherein 1 circuit-switched data calculate finish.2. 3. 4. by that analogy, 3 times steps are repeated, 4 circuit-switched datas that just can obtain the 1st grade of butterfly computation of base 4 obtain its excess-three road computing and finished.
Assuming that 64 radix-4 butterfly computings, equivalent to step in Fig. 5 2. in calculating process, 4 circuit-switched datas of input need 16cycles, and 64 data labels for example can be as follows：
0、 1、 2、 3、 4、 5、 6、 7、 8、 9、 10、 11、 12、 13、 14、 15
16th, 17,18,19,20,21,22,23,24,25,26,27,28,29,30,31 32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47 48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63 progress radix-4 butterfly computings obtain the 1st circuit-switched data of 4 circuit-switched datas for for example labeled as：
A, B, C, D, E, F, G, H, I, J, K, L, M, N, 0, P, 16 data need to carry out lower 1 grade of radix-4 butterfly computing；
Carrying out 4 tunnels of lower 1 grade of radix-4 butterfly computing, totally 16 input datas are as follows：
A、 B、 C、 D
E、 F、 G、 H
1、 J、 K、 L
M、 N、 0、 P
In view of when carrying out butterfly computation plus-minus, twiddle factor multiplication need several clock cycle, in the radix-4 butterfly computing for carrying out above-mentioned 16 point data, the data M that need to use, N, 0, P are that correspondence is calculated by data 12,13,14,15,28,29,30,31,44,45,46 and 47,60,61,62,63 and produced when carrying out radix-4 butterfly computing by 64 point datas, and these data for just having inputted radix-4 butterfly arithmetic unit need some clock cycle（Generally higher than 4 clock cycle）Operation result M, N, 0, P can just be generated, therefore, may be also not all right if reading to carry out 16 computings of base 4 of next stage immediately after each 16 number in 4 tunnels that needs of 64 butterfly computations from shift register, because data 12,28,44,60 input radix-4 butterfly arithmetic unit and have just crossed 3 clocks（3 clocks are to input radix-4 butterfly arithmetic unit by 13,14,15）, data M may now also fail to calculating and obtain and be stored in shift register, and similarly N, 0, P are not also calculated now and finished.Therefore, certain delay disposal can be carried out so that data needed for carrying out rear class butterfly computation, obtains and incorporate in shift register after corresponding prime butterfly computation.
Below still with to 256 point datas（44) carry out FFT computings exemplified by, the sequence arrangement that another butterfly provided in an embodiment of the present invention is calculated is for example shown in Fig. 7-b.
(1) the radix-4 butterfly computing of the 1st grade of 256 point datas is performed, 4 tunnels are obtained（64 point datas/road）The data of pending 2nd grade of radix-4 butterfly computing；
(2) 4 tunnels that step (1) is obtained are performed（64 point datas/road）Wherein 2 tunnels in the data of pending 2nd grade of radix-4 butterfly computing（64 point datas/road）2nd grade of radix-4 butterfly computing of data, obtains 8 tunnels（16 point datas/road）The data of pending 3rd level radix-4 butterfly computing；
(3) wherein 4 tunnels in the pending 3rd level radix-4 butterfly operational data in 8 roads obtained in step (2) are performed（16 point datas/road）The 3rd level radix-4 butterfly computing of data, obtains 16 tunnels（4 point datas/road）The data of pending 4th grade of radix-4 butterfly computing；
(4) 16 tunnels that step (3) is obtained are performed（4 point datas/road）4th grade of radix-4 butterfly computing of data, obtains 64 output datas（FFT operation results）.
Wherein, step (1) is performed 1 time, 1 step (2) (3) (4) is then performed again, 1 step (3) (4) is then performed again, 1 step (2) (3) (4) is then performed again, 1 time (3) (4) are then being performed again, and so far, all data complete computing output operation result.
32 cycle have been used in step (2), 256 point datas are first subjected to 4 circuit-switched datas that first order radix-4 butterfly computing is obtained（64 point datas/road）In wherein two-way perform its corresponding rear class radix-4 butterfly computing, then
Again 256 point datas are carried out with 4 circuit-switched datas that the 1st grade of radix-4 butterfly computing is obtained（64 point datas/road）In other two-way perform its corresponding rear class radix-4 butterfly computing.So, the execution that lower 1 grade of radix-4 butterfly computing can either be ensured just is carried out after prime operation result is stored in shift register, take full advantage of the work period of radix-4 butterfly arithmetic unit again simultaneously, reduce standby time, larger improves operation efficiency.
It is understood that the example above is mainly exemplified by carrying out radix-4 butterfly computing to data to be described, and the process for carrying out other butterfly computations to data can by that analogy, and here is omitted.The execution of such scheme of the embodiment of the present invention can be controlled using controller, the controller for example can independently be disposed and deposit area with each Slow and base y butterfly computation devices are connected, or can also be deployed directly into Slow deposit in area or base y butterfly computation devices.
Therefore, by the data of the pending z grades of bases y butterfly computations in y roads in the present embodiment, input base y butterfly computations device and carry out z grades of base y butterfly computations, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained inputs base y butterfly computations device and carries out z+1 grades of base y butterfly computations.Area is deposited due to the mechanism to base y butterfly computations device and Slow to be multiplexed, two-stage base y butterfly computations are multiplexed same base y butterfly computation devices, the time-multiplexed base y butterfly computations device of such as availability data input carries out computing, so advantageously reduce the time of butterfly computation and reduce resource consumption.
Also, by the base y butterfly computations at different levels order of ingenious arrangement data, the delay stand-by period is advantageously reduced, operation efficiency is further improved.For ease of preferably implementing the relevant apparatus of the embodiment of the present invention, the embodiment of the present invention also provides the relevant apparatus for implementing such scheme.
Referring to Fig. 8-a, the embodiment of the present invention provides a kind of data processing equipment 800, may include：
Controller 810 and base y butterfly computations device 820;
Wherein controller 810, for controlling the data of the pending z grades of bases y butterfly computations in y roads, inputs base y butterfly computations device 820 and carries out z grades of base y butterfly computations, wherein, the y and z are positive integer；In the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained inputs base y butterfly computations device 820 and carries out z+1 grades
Base y butterfly computations.
Referring to Fig. 8-b, in some embodiments of the invention, data processing equipment 800 may also include y Ge Slowcun areas 830;
Controller 810 can be further used for, the data input base y butterfly computations device 820 of the pending z grades of bases y butterfly computations in y roads is being subjected to z grades of base y butterfly computations, after the result data for obtaining the z grades of base y butterfly computations in y roads, the y roads result data Zhong Mei roads result data that the z grades of base y butterfly computations are obtained is divided into y part, and this y part is respectively written into y Ge Slowcun areas 830.
In some embodiments of the invention, controller 810 can be specifically for, the data input base y butterfly computations device 820 of the pending z grades of bases y butterfly computations in y roads is subjected to z grades of base y butterfly computations, in the case where z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that z grades of base y butterfly computations of output are obtained；In the case where z grades of base y butterfly computations are not afterbody base y butterfly computations, the y roads result data Zhong Mei roads result data that z grades of base y butterfly computations are obtained is divided into y part, and y part is respectively written into y Ge Slowcun areas 830, the y part that the Zhong Gaimei roads result data of y Ge Slowcun areas 830 is divided into will be respectively written into, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input base y butterfly computations device 820 and carry out z+1 grades of base y butterfly computations, wherein, y and z is positive integer.
In some embodiments of the invention, controller 810 can be for example controlled pending kth level formerly（Such as the 2nd or 3rd level or non-last 1 grade of other levels）The data of base y butterfly computations are divided into several parts（Such as 2 parts, 3 parts or more part）, first carry out the kth level base y butterfly computations of certain partial data therein, after the completion of remaining all grades of base y butterfly computations of certain partial data are all calculated, the kth level base y butterfly computations of another part of the data of pending kth level base y butterfly computations are carried out again, until remaining all grades of base y butterfly computations of another part data all calculate completion, if there is another partial data in pending kth level, the kth level base y butterfly computations a part of again of the data of the pending kth level base y butterfly computations are carried out again, until remaining all grades of base y butterfly computations of another partial data all calculate completion, by that analogy, it is finally completed all grades of base y butterfly computations of all data, obtain FFT result of calculations.
For example, controller 810 can be specifically for, the data input base y butterfly computations device 820 of the pending z grades of bases y butterfly computations in y roads is subjected to z grades of base y butterfly computations, in the case where z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, by this
The y roads result data Zhong Mei roads result data that z grades of base y butterfly computations are obtained is divided into y part, and this y part is respectively written into y Ge Slowcun areas 830；First by the y roads result data, it is first written to the y part that the wherein m roads result data Zhong Mei roads result data in the y Ge Slowcun areas 830 is divided into respectively, z+1 grades of base y butterfly computations are carried out to input base y butterfly computations device 820 as the data of the pending z+1 grades of bases y butterfly computations in y roads, the result data of the z+1 grades of base y butterfly computations in y*m roads is obtained；After the result data of the z+1 grades of base y butterfly computations in y*m roads is obtained, by in the y roads result data, the y part that the remaining n roads result data Zhong Mei roads result data in the y Ge Slowcun areas 830 is divided into respectively is written to afterwards, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input base y butterfly computations device 820 and carry out z+1 grades of base y butterfly computations, obtain the result data of the z+1 grades of base y butterfly computations in y*n roads, the wherein m is less than y, and the m is equal to y with the n sums.
In some embodiments of the invention, controller 810 can be specifically for, by in the y roads result data of z grades of base y butterfly computations, the y part that the remaining n roads result data in the y Ge Slowcun areas 830 is divided into respectively is written to afterwards, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input before the z+1 grades of base y butterfly computations of progress of base y butterfly computations device 820, the result data Zhong Mei roads result data of z+1 grades of Xian Jiang y*m roads base y butterfly computations is divided into y part, the data of the pending z+2 grades of bases y butterfly computations in y roads are partly used as using this y, input the base y butterfly computations device 820 and carry out z+2 grades of base y butterfly computations, obtain the result data of the z+2 grades of base y butterfly computations in y*y*m roads.
In other embodiments of the present invention, such as controller 810 can also be controlled after the data of all pending every grade of base y butterfly computations all carry out this grade of base y butterfly computation, then carries out the next stage base y butterfly computations of the data of all pending next stage base y butterfly computations.
For example, controller 810 can be additionally used in, after the result data for the z+1 grades of base y butterfly computations in y*y roads that z+1 grades of base y butterfly computations are obtained is carried out, the result data Zhong Mei roads result data of z+1 grades of y*y roads base y butterfly computations is divided into y part, using this y part as the data of the pending z+2 grades of bases y butterfly computations in y roads, input base y butterfly computations device 820 and carry out z+2 grades of base y butterfly computations.
In one embodiment of the invention, y Ge Slowcun areas 830 can be located in y Slow storage respectively；Or, y Ge Slowcun areas 830 are located at least one Slow storage.
It is appreciated that controller 810 can be built in base y butterfly computations device 820, or Slow can be built in deposits in area 830, can also be independently arranged certainly.The function of controller 810, all or part of can also be integrated in
In the place Slow storages of 820 Huo Slowcun areas of base y butterfly computations device 830.
It is understood that the function of each functional module of the data processing equipment 800 of the present embodiment can be implemented according to the method for above method embodiment introduction, it implements process and may refer to above-mentioned implementation this is repeated no more.
It can be seen that, controller in the present embodiment data processing equipment 800 is controlled the data of the pending z grades of bases y butterfly computations in y roads, input base y butterfly computations device and carry out z grades of base y butterfly computations, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained inputs base y butterfly computations device and carries out 1 grade of base y butterfly computation of z+.Because the mechanism is multiplexed to base y butterfly computation devices, two-stage base y butterfly computations are multiplexed same base y butterfly computation devices, the time-multiplexed base y butterfly computations device of such as availability data input carries out computing, so advantageously reduce the time of butterfly computation and reduce resource consumption.
Referring to Fig. 9, the embodiment of the present invention also provides a kind of electronic equipment 900, wherein, data processing equipment 800 is deployed with the electronic equipment.Wherein, electronic equipment 900 for example can be access device（Such as base station, access point）, user terminal（Such as mobile phone, portable computer）Or other electronic equipments for needing to carry out data processing.
The embodiment of the present invention also provides a kind of computer-readable storage medium, wherein, the computer-readable storage medium can have program stored therein, the part or all of step including the data processing method described in the above method embodiment when program is performed.
It should be noted that, for foregoing each method embodiment, in order to which cylinder list is described, therefore it is all expressed as to a series of combination of actions, but those skilled in the art should know, the present invention is not limited by described sequence of movement, because according to the present invention, some steps can be carried out sequentially or simultaneously using other.Secondly, those skilled in the art should also know, embodiment described in this description belongs to preferred embodiment, and involved action and the module not necessarily present invention are necessary.
In the above-described embodiments, the description to each embodiment all emphasizes particularly on different fields, and does not have the part being described in detail in some embodiment, may refer to the associated description of other embodiment.
To sum up, by the data of the pending z grades of bases y butterfly computations in y roads in the embodiment of the present invention, input base y butterfly computations device carries out z grades of base y butterfly computations, is afterbody base y butterflies in the z grades of base y butterfly computations
In the case of shape computing, the result data that the z grades of base y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of base y butterfly computations are obtained inputs base y butterfly computations device and carries out z+1 grades of base y butterfly computations.Because the mechanism is multiplexed to base y butterfly computation devices, two-stage base y butterfly computations are multiplexed same base y butterfly computation devices, the time-multiplexed base y butterfly computations device of such as availability data input carries out computing, so advantageously reduce the time of butterfly computation and reduce resource consumption.
Also, by the base y butterfly computations at different levels order of ingenious arrangement data, the delay stand-by period is advantageously reduced, operation efficiency is further improved.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment can be by program to instruct the hardware of correlation to complete, the program can be stored in a computer-readable recording medium, and storage medium can for example include：Read-only storage, random access memory, disk or CD etc..
The explanation that a kind of data processing method and relevant apparatus provided above the embodiment of the present invention apply in detail example is only intended to the method and its core concept for helping to understand the present invention；Simultaneously for those of ordinary skill in the art, according to the thought of the present invention, it will change in specific embodiments and applications, to sum up, this specification content should not be construed as limiting the invention.
Claims (19)
- Claim1st, a kind of data processing method, it is characterised in that including：The data input base y butterfly computations device of the pending z grades of bases y butterfly computations in y roads is subjected to z grades of base y butterfly computations, wherein, the y and z are positive integer；In the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained inputs the base y butterfly computations device and carries out z+1 grades of base y butterfly computations.2nd, according to the method described in claim 1, it is characterised in thatThe data input base y butterfly computations device by the pending z grades of bases y butterfly computations in y roads is carried out after z grades of base y butterfly computations, in addition to：The y roads result data Zhong Mei roads result data that the z grades of bases y butterfly computations are obtained is divided into y part, and described y part is respectively written into y Ge Slowcun areas.3rd, method according to claim 2, it is characterised in thatThe result data that the z grades of bases y butterfly computations are obtained, inputs the base y butterfly computations device and carries out z+1 grades of base y butterfly computations, including：The y part being divided into per road result data that will be respectively written into the y Ge Slowcun areas, as the data of the pending z+1 grades of bases y butterfly computations in y roads, inputs the base y butterfly computations device and carries out z+1 grades of base y butterfly computations.4th, according to the method in claim 2 or 3, it is characterised in thatMethods described also includes：In the case where the z+1 grades of base y butterfly computations are afterbody base y butterfly computations, the result data of the z+1 grades of bases y butterfly computations is exported；In the case where the z+1 grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z+1 grades of bases y butterfly computations are obtained inputs the base y butterfly computations device and carries out z+2 grades of base y butterfly computations.5th, according to the method in claim 2 or 3, it is characterised in thatThe y part being divided into per road result data that will be respectively written into the y Ge Slowcun areas, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input the base y butterfly computations device and carry out z+1 grades of base y butterfly computations, including： First by the result data of the y roads, it is first written to the y part that the wherein m roads result data Zhong Mei roads result data in the y Ge Slowcun areas is divided into respectively, z+1 grades of base y butterfly computations are carried out to input the base y butterfly computations device as the data of the pending z+1 grades of bases y butterfly computations in y roads, the result data of the z+1 grades of base y butterfly computations in y*m roads is obtained；After the result data of the z+1 grades of base y butterfly computations in y*m roads is obtained, by in the result data of the y roads, the y part that the remaining n roads result data Zhong Mei roads result data in the y Ge Slowcun areas is divided into respectively is written to afterwards, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input the base y butterfly computations device and carry out z+1 grades of base y butterfly computations, obtain the result data of the z+1 grades of base y butterfly computations in y*n roadsWherein, the m is less than the y, and the m is equal to the y with the n sums.6th, method according to claim 5, it is characterised in thatMethods described also includes：In the result data by the y roads, the y part that the remaining n roads result data Zhong Mei roads result data in the y Ge Slowcun areas is divided into respectively is written to afterwards, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input before z+1 grades of base y butterfly computations of the base y butterfly computations device progress, the result data Zhong Mei roads result data of z+1 grades of y*m roads base y butterfly computations is divided into y part, the data of the pending z+2 grades of bases y butterfly computations in y roads are partly used as using this y, input the base y butterfly computations device and carry out z+2 grades of base y butterfly computations.7th, the method according to claim 5 or 6, it is characterised in thatThe m is less than or equal to the n.8th, according to the method in claim 2 or 3, it is characterised in thatMethods described also includes：After the result data for the z+1 grades of base y butterfly computations in y*y roads that the z+1 grades of bases y butterfly computations are obtained is carried out, the result data Zhong Mei roads result data of described z+1 grades of base y butterfly computations in y*y roads is divided into y part, using this y part as the data of the pending z+2 grades of bases y butterfly computations in y roads, input the base y butterfly computations device and carry out z+2 grades of base y butterfly computations.9th, the method according to any one of claim 2 to 8, it is characterised in thatThe y Ge Slowcun areas are located in y Slow storage respectively；Or, the y Ge Slowcun areas are located at least one Slow storage.10th, a kind of data processing equipment, it is characterised in that including： Controller and base y butterfly computation devices；Wherein, the controller, for base y butterfly computations device described in the data input of the pending z grades of bases y butterfly computations in y roads to be carried out into z grades of base y butterfly computations, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained, input the base y butterfly computations device and carry out z+1 grades of base y butterfly computations, wherein, the y and z are positive integer.11st, data processing equipment according to claim 10, it is characterised in thatThe data processing equipment also includes y Ge Slowcun areas；The controller is additionally operable to, after the data input base y butterfly computations device of the pending z grades of bases y butterfly computations in y roads to be carried out to z grades of base y butterfly computations, the y roads result data Zhong Mei roads result data that the z grades of bases y butterfly computations are obtained is divided into y part, and described y part is respectively written into y Ge Slowcun areas.12nd, data processing equipment according to claim 11, it is characterised in thatThe controller specifically for, base y butterfly computations device described in the data input of the pending z grades of bases y butterfly computations in y roads is subjected to z grades of base y butterfly computations, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the y roads result data Zhong Mei roads result data that the z grades of bases y butterfly computations are obtained is divided into y part, and described y part is respectively written into y Ge Slowcun areas, the y part being divided into per road result data that will be respectively written into the y Ge Slowcun areas, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input the base y butterfly computations device and carry out z+1 grades of base y butterfly computations, wherein, the y and z are positive integer.13rd, the data processing equipment according to claim 11 or 12, it is characterised in thatThe controller specifically for, base y butterfly computations device described in the data input of the pending z grades of bases y butterfly computations in y roads is subjected to z grades of base y butterfly computations, in the case where the z grades of base y butterfly computations are afterbody base y butterfly computations, the result data that the z grades of bases y butterfly computations are obtained is exported；In the case where the z grades of base y butterfly computations are not afterbody base y butterfly computations, the y roads result data Zhong Mei roads result data that the z grades of bases y butterfly computations are obtained is divided into y part, and will The y part is respectively written into y Ge Slowcun areas；First by the result data of the y roads, it is first written to the y part that the wherein m roads result data Zhong Mei roads result data in the y Ge Slowcun areas is divided into respectively, z+1 grades of base y butterfly computations are carried out to input the base y butterfly computations device as the data of the pending z+1 grades of bases y butterfly computations in y roads, the result data of the z+1 grades of base y butterfly computations in y*m roads is obtained；After the result data of the z+1 grades of base y butterfly computations in y*m roads is obtained, by in the result data of the y roads, the y part that the remaining n roads result data Zhong Mei roads result data in the y Ge Slowcun areas is divided into respectively is written to afterwards, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input the base y butterfly computations device and carry out z+1 grades of base y butterfly computations, obtain the result data of the z+1 grades of base y butterfly computations in y*n roads, wherein, the m is less than the y, and the m is equal to the y with the n sums.14th, data processing equipment according to claim 13, it is characterised in thatThe controller is additionally operable to, by in the result data of the y roads, the y part that the remaining n roads result data in the y Ge Slowcun areas is divided into respectively is written to afterwards, it is used as the data of the pending z+1 grades of bases y butterfly computations in y roads, input before z+1 grades of base y butterfly computations of the base y butterfly computations device progress, the result data Zhong Mei roads result data of z+1 grades of y*m roads base y butterfly computations is divided into y part, the data of the pending z+2 grades of bases y butterfly computations in y roads are partly used as using this y, input the base y butterfly computations device and carry out z+2 grades of base y butterfly computations, obtain the result data of the z+2 grades of base y butterfly computations in y*y*m roads.15th, the data processing equipment according to claim 11 or 12, it is characterised in thatThe controller is additionally operable to, after the result data for the z+1 grades of base y butterfly computations in y*y roads that the z+1 grades of bases y butterfly computations are obtained is carried out, the result data Zhong Mei roads result data of described z+1 grades of base y butterfly computations in y*y roads is divided into y part, using this y part as the data of the pending z+2 grades of bases y butterfly computations in y roads, input the base y butterfly computations device and carry out z+2 grades of base y butterfly computations.16th, the data processing equipment according to any one of claim 11 to 15, it is characterised in that the y Ge Slowcun areas are located in y Slow storage respectively；Or, the y Ge Slowcun areas are located at least one Slow storage.17th, a kind of access device, it is characterised in that disposed in the access device just like the data processing equipment described in any one of claim 11 to 17.18th, a kind of user equipment, it is characterised in that disposed in the user equipment just like claim 11 To the data processing equipment described in 17 any one.19th, a kind of computer-readable storage medium, it is characterised in thatThe computer-readable storage medium has program stored therein, including the step as described in any one of claim 1 to 9 when described program is performed.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2012/074911 WO2013159361A1 (en) | 2012-04-28 | 2012-04-28 | Data processing method and related device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103493039A true CN103493039A (en) | 2014-01-01 |
| CN103493039B CN103493039B (en) | 2016-06-29 |
Family
ID=49482181
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201280000317.4A Active CN103493039B (en) | 2012-04-28 | 2012-04-28 | Data processing method, data processing equipment, access device and subscriber equipment |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN103493039B (en) |
| WO (1) | WO2013159361A1 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106970895A (en) * | 2016-01-14 | 2017-07-21 | 普天信息技术有限公司 | FFT device and methods based on FPGA |
| CN109117188A (en) * | 2018-08-06 | 2019-01-01 | 合肥工业大学 | A kind of restructural butterfly computation device of multichannel hybrid base FFT |
| CN112051446A (en) * | 2020-08-18 | 2020-12-08 | 许继集团有限公司 | Mixed base FFT implementation method and device for broadband measurement of power system |
| CN113570612A (en) * | 2021-09-23 | 2021-10-29 | 苏州浪潮智能科技有限公司 | Image processing method, device and equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101300572A (en) * | 2005-03-11 | 2008-11-05 | 高通股份有限公司 | Fast fourier transform twiddle multiplication |
| KR101036873B1 (en) * | 2010-09-13 | 2011-05-25 | 심흥섭 | Flag based low-complexity, expandable split radix fft system |
| JP2011133957A (en) * | 2009-12-22 | 2011-07-07 | Hitachi Kokusai Electric Inc | Arithmetic device |
| US20120041996A1 (en) * | 2010-08-16 | 2012-02-16 | Leanics Corporation | Parallel pipelined systems for computing the fast fourier transform |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101072218B (en) * | 2007-03-01 | 2011-11-30 | 华为技术有限公司 | FFT/IFFI paired processing system, device and method |
| CN101833540B (en) * | 2010-04-07 | 2012-06-06 | 华为技术有限公司 | Signal processing method and device |
-
2012
- 2012-04-28 WO PCT/CN2012/074911 patent/WO2013159361A1/en active Application Filing
- 2012-04-28 CN CN201280000317.4A patent/CN103493039B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101300572A (en) * | 2005-03-11 | 2008-11-05 | 高通股份有限公司 | Fast fourier transform twiddle multiplication |
| JP2011133957A (en) * | 2009-12-22 | 2011-07-07 | Hitachi Kokusai Electric Inc | Arithmetic device |
| US20120041996A1 (en) * | 2010-08-16 | 2012-02-16 | Leanics Corporation | Parallel pipelined systems for computing the fast fourier transform |
| KR101036873B1 (en) * | 2010-09-13 | 2011-05-25 | 심흥섭 | Flag based low-complexity, expandable split radix fft system |
Non-Patent Citations (1)
| Title |
|---|
| BYUNG S.SON ET AL: "A HIGH-SPEED FFT PROCESSOR FOR OFDM SYSTEMS", 《ISCAS 2002》, 29 May 2002 (2002-05-29), pages 281 - 284 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106970895A (en) * | 2016-01-14 | 2017-07-21 | 普天信息技术有限公司 | FFT device and methods based on FPGA |
| CN106970895B (en) * | 2016-01-14 | 2023-10-03 | 普天信息技术有限公司 | FFT device and method based on FPGA |
| CN109117188A (en) * | 2018-08-06 | 2019-01-01 | 合肥工业大学 | A kind of restructural butterfly computation device of multichannel hybrid base FFT |
| CN112051446A (en) * | 2020-08-18 | 2020-12-08 | 许继集团有限公司 | Mixed base FFT implementation method and device for broadband measurement of power system |
| CN113570612A (en) * | 2021-09-23 | 2021-10-29 | 苏州浪潮智能科技有限公司 | Image processing method, device and equipment |
| CN113570612B (en) * | 2021-09-23 | 2021-12-17 | 苏州浪潮智能科技有限公司 | Image processing method, device and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103493039B (en) | 2016-06-29 |
| WO2013159361A1 (en) | 2013-10-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106445471B (en) | Processor and the method for performing matrix multiplication on a processor | |
| CN107797962B (en) | Neural network based computational array | |
| CN103955447B (en) | FFT accelerator based on DSP chip | |
| CN100435090C (en) | Extensible high-radix Montgomery's modular multiplication algorithm and circuit structure thereof | |
| CN101290613B (en) | FFT processor data storage system and method | |
| CN101763338B (en) | Mixed base FFT/IFFT realization device with changeable points and method thereof | |
| CN103970720B (en) | Based on extensive coarseness imbedded reconfigurable system and its processing method | |
| CN102999313B (en) | A kind of data processing method based on montgomery modulo multiplication | |
| CN112865954A (en) | Accelerator, chip and system for Paillier decryption | |
| CN103984560A (en) | Embedded reconfigurable system based on large-scale coarseness and processing method thereof | |
| CN101847986A (en) | Circuit and method for realizing FFT/IFFT conversion | |
| CN103493039A (en) | Data processing method and related device | |
| CN103984677A (en) | Embedded reconfigurable system based on large-scale coarseness and processing method thereof | |
| US9262378B2 (en) | Methods and devices for multi-granularity parallel FFT butterfly computation | |
| CN101894096A (en) | FFT computing circuit structure applied to CMMB and DVB-H/T | |
| CN103544111B (en) | A kind of hybrid base FFT method based on real-time process | |
| JP4160564B2 (en) | Fast Fourier transform apparatus with improved processing speed and processing method thereof | |
| CN101667984A (en) | 3780-point fast Fourier transform processor and computing control method thereof | |
| CN104050148B (en) | Fast Fourier Transform (FFT) accelerator | |
| CN103034621B (en) | The address mapping method of base 2 × K parallel FFT framework and system | |
| CN102129419B (en) | Based on the processor of fast fourier transform | |
| US7693926B2 (en) | Modular multiplication acceleration circuit and method for data encryption/decryption | |
| US9268744B2 (en) | Parallel bit reversal devices and methods | |
| CN105893326B (en) | The device and method for realizing 65536 point FFT based on FPGA | |
| CN109308327A (en) | Figure calculation method device medium apparatus based on the compatible dot center's model of subgraph model |
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 |