CN105991505A - De-convolutional interleaver and de-convolutional interleaving method - Google Patents
De-convolutional interleaver and de-convolutional interleaving method Download PDFInfo
- Publication number
- CN105991505A CN105991505A CN201510088389.4A CN201510088389A CN105991505A CN 105991505 A CN105991505 A CN 105991505A CN 201510088389 A CN201510088389 A CN 201510088389A CN 105991505 A CN105991505 A CN 105991505A
- Authority
- CN
- China
- Prior art keywords
- data
- corvolutional interleaving
- corvolutional
- interleaving
- solution
- 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.)
- Pending
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention provides a de-convolutional interleaver applied to process multi-group convolutional interleaving data. The multi-group convolutional interleaving data package includes multiple convolutional interleaving data which is formed by processing multi-group non-interleaving data through convolutional interleaving processing, each non-interleaving data includes L data, and adjacent two data corresponds a delay depth gap in each non-interleaving data after the convolutional interleaving processing, wherein L is a positive integer. The de-convolutional interleaver includes: an input data buffer configured to temporarily store multi-group convolutional interleaving data; a memory controller configured to store the multi-group convolutional interleaving data temporarily stored in the input data buffer in the memory to perform de-convolutional interleaving processing, the memory address of each stored convolutional interleaving data being determined by the delayed depth, the L value and the delay depth distance corresponding to the data; and an output data buffer configured to temporarily store multi-group convolutional interleaving data read from the memory.
Description
Technical field
The present invention is with regard to deinterlacer and de-interlace method, circles round with solving especially with respect to solving Corvolutional interleaving device
Interleave method.
Background technology
Communication transfer technology (such as OFDM (OFDM) Radio Transmission Technology) is in order to strengthen letter
Number for the resistivity of interference, time interleaving (Time Interleaving) can be imposed to signal in transmission end
Process, and impose time release of an interleave (Time De-interleaving) to signal at receiving terminal and process.Time
The staggered configuration diagram processing as it is shown in figure 1, this framework 100 comprises K staggered processing unit 110,
Each staggered processing unit 110 has ncIndividual input and ncIndividual output, is respectively intended to basis signal sampling
Clock received in sequence and output signal, and the framework of each staggered processing unit 110 is as in figure 2 it is shown, comprise
ncIndividual path, is used for imposing staggered process to signal.
Above-mentioned staggered processing unit 110 can use Corvolutional interleaving (Convolutional Interleaving) to process,
One embodiment is as it is shown on figure 3, the 1st has n between input and outputcThe delay ((n of-1 unitc
-1) D)；2nd has n between input and outputcThe delay ((n of-2 unitsc-2) D)；…；
N-thcThere is the delay (D) of 1 unit between-1 input and output；And n-thcIndividual input with
Not postponing (Bypass) between output, the wherein delay of each unit can be I letter according to design requirement
Number cycle or sample of signal clock, be positive integer at this I.And process through above-mentioned Corvolutional interleaving at signal
In the case of, signal receiving end has a corresponding solution Corvolutional interleaving and processes framework to perform release of an interleave, this solution
Corvolutional interleaving processes framework and comprises K release of an interleave processing unit, and each release of an interleave processing unit is equally according to signal
Sampling clock comes received in sequence and output signal, and has the delay arrangement contrary with staggered processing unit 110,
As shown in Figure 4, the 1st of each release of an interleave processing unit 410 does not postpone between input and output
(Bypass)；1st delay (D) having 1 unit between input and output；…；N-thc-1
There is n between individual input and outputcThe delay ((n of-2 unitsc-2) D)；And n-thcIndividual input
And there is n between outputcThe delay ((n of-1 unitc-1) D).
The delay of the various unit of above-mentioned receiving terminal can pass through delay buffer (Delay Buffer) and realizes.
A kind of current technology is the function using static RAM (SRAM) to reach delay buffer,
It although SRAM access speed is fast, but due to fairly expensive, therefore is unfavorable for cost benefit；Separately there is a kind of mesh
Front technology is to use the system storage shared to realize the function of delay buffer, owing to system storage leads to
Being often Synchronous Dynamic Random Access Memory (SDRAM), price is relatively cheap, therefore can save hardware
Cost, but, this technology is to write in SDRAM (relatively early data continuously according to the reception order of data
It is memory address written to relatively early be written), the pass that this reception order is processed because of the Corvolutional interleaving of transmission end
It is and does not sorts according to the continuity of data, therefore when solving Corvolutional interleaving process, be the continuous of restoring data
Property, this technology often need to read data from the storage address of different lines (Different Rows), also
It is to say, due to the design of structure and the access mode of SDRAM, when each pen is as release of an interleave data
When data are stored in column address regions different in SDRAM, after first time, read operation completed, need
Carry out a precharge operation (precharge) to close the column address region corresponding to first time read operation
Close, then open operation (row active) by the column address corresponding to second time read operation through once row again
Region is opened, and so, when many pen data interlock and are stored in different column position regions, is additionally required
Precharge operation and row open operation will expend a large amount of access time, constantly to arrange access conversion
On (Row Accessing Change), in summary, this technology not only needs longer Xie Huixuanjiaocuochu
The reason time, also can affect the bandwidth of memory of other circuit access system memory.
Content of the invention
In view of the deficiency of prior art, a purpose of the present invention is to provide a kind of solution Corvolutional interleaving device and untwists
Return interleave method, to improve prior art.
The invention discloses that a kind of being applied to processes the solution Corvolutional interleaving device organizing Corvolutional interleaving data more, these many groups are returned
Rotation intercrossed data comprises many Corvolutional interleaving data, and this many Corvolutional interleaving data are not interlock by many groups
Data impose a Corvolutional interleaving process and form, and often organize non-intercrossed data and comprise L pen data, through this Corvolutional interleaving
The corresponding delay depth difference of adjacent two pen data in non-intercrossed data are often organized away from, this L positive integer after process, should
The embodiment solving Corvolutional interleaving device comprises: an Input Data Buffer, is used for keeping in this many groups Corvolutional interleaving
Data；One Memory Controller, by will be temporarily stored into this many groups Corvolutional interleaving number of this Input Data Buffer
According to be accessed in a memory with carry out solve Corvolutional interleaving process, depositing of every this Corvolutional interleaving data being stored
Memory address be based on corresponding to this pen data postponed the degree of depth, this L value and this delay depth difference away from
Determine；And an output data buffer, it is used for the temporary many groups solution Corvolutional interleaving number reading from this memory
According to.
The present invention separately discloses and a kind of is applied to process the solution Corvolutional interleaving device organizing Corvolutional interleaving data more, these many groups
Corvolutional interleaving packet contains many Corvolutional interleaving data, and an embodiment of this solution Corvolutional interleaving device comprises: one is defeated
Enter data buffer, be used for keeping in this many groups Corvolutional interleaving data；One Memory Controller, temporary by inciting somebody to action
In this many groups Corvolutional interleaving data access of this Input Data Buffer in a memory to carry out a solution convolution
Staggered process, and obtain organizing solution Corvolutional interleaving data more, this Memory Controller is by this many Corvolutional interleaving numbers
According to, the many pen data that should organize the same group of solution Corvolutional interleaving data that solve in Corvolutional interleaving data are stored more
Multiple same column chunks in this memory；And an output data buffer, it is used for temporary depositing from this
This many groups solution Corvolutional interleaving data that reservoir reads.
The present invention also discloses a kind of solution Corvolutional interleaving method, for processing many group Corvolutional interleaving data, and these many groups
Corvolutional interleaving packet contains many Corvolutional interleaving data, and an embodiment of this solution Corvolutional interleaving method comprises following
Step: by this many groups Corvolutional interleaving data access in a memory to carry out solving Corvolutional interleaving process, obtain many
Group solves Corvolutional interleaving data；Wherein this many Corvolutional interleaving data solve in Corvolutional interleaving data to organizing more
Multiple same column chunks of being stored in this memory of many pen data of same group of solution Corvolutional interleaving data.
Feature for the present invention, implementation and effect, hereby coordinate graphic make preferred embodiment describe in detail as follows.
Brief description
Fig. 1 is the configuration diagram that time interleaving is processed；
Fig. 2 is the configuration diagram of the staggered processing unit of Fig. 1；
The configuration diagram of the staggered processing unit that Fig. 3 is processed for using Corvolutional interleaving；
Fig. 4 solves the configuration diagram of the release of an interleave processing unit that Corvolutional interleaving is processed for using；
Fig. 5 is the schematic diagram of an embodiment of the solution Corvolutional interleaving device of the present invention；
Fig. 6 a is that the non-intercrossed data of Corvolutional interleaving processing unit processes of prior art is to export Corvolutional interleaving number
According to schematic diagram；
Fig. 6 b is that the solution Corvolutional interleaving device of the present invention processes Corvolutional interleaving data to export solution Corvolutional interleaving data
Schematic diagram；
Fig. 7 is that data are stored in showing of dynamic random access memory by the memory access controller of the present invention
It is intended to；
Fig. 8 is that data are stored in dynamic random access memory by the memory access controller of prior art
Schematic diagram；
Fig. 9 is the schematic diagram of an embodiment of the memory access control of the present invention；
Figure 10 is the schematic diagram of the memory access control of prior art；
Figure 11 is the schematic diagram of an embodiment of the solution Corvolutional interleaving method of the present invention；And
Figure 12 is the schematic diagram of another embodiment of the solution Corvolutional interleaving method of the present invention.
Symbol description
The framework that 100 time interleavings are processed
110 interlock processing unit
410 release of an interleave processing units
500 solution Corvolutional interleaving devices
510 Input Data Buffers
520 memory access controllers
530 dynamic random access memory
540 output data buffers
600 cycles interlock processing unit
610～640 transmission paths
S1110, S1120, S1210, S1220 step
Detailed description of the invention
Disclosed herein a kind of Corvolutional interleaving device (Convolutional De-Interleaver) that solves to circle round with solution
Interleave method, can use dynamic random access memory carrying out release of an interleave saving hardware cost, and energy
Reduce the number of times of memory column access mapping by the suitable storage address selecting data access, thus change
Kind usefulness.
Fig. 5 is the schematic diagram of an embodiment of the solution Corvolutional interleaving device of the present invention, as it can be seen, the present embodiment
Solution Corvolutional interleaving device 500 comprise: an Input Data Buffer 510；One memory access controller 520
It is used for controlling a dynamic random access memory 530 (such as Synchronous Dynamic Random Access Memory (SDRAM))
Data access；And an output data buffer 540.This Input Data Buffer 510 is e.g. advanced
First go out buffer (FIFO) or its equivalent device, be used for receiving and export many group Corvolutional interleaving data to dynamically
Random access memory 530, to carry out input during release of an interleave computing as memory access controller 520
Data buffering, this many groups Corvolutional interleaving data are by imposing a Corvolutional interleaving process to organizing non-intercrossed data more
Form, often organize non-intercrossed data and comprise L pen data, that is this organizes J group data of non-intercrossed data more
Comprise continuous L pen data, this L pen data corresponding L kind different delay degree of depth (or saying time delay) respectively,
Wherein the delay depth difference corresponding to adjacent two pen data in same group of non-intercrossed data is away from for I operation
Clock (such as data sample clock or data signal cycles etc.), this L, this J and this I are positive integer,
And corresponding above-mentioned packet transmission sequence (or the above-mentioned many group Corvolutional interleaving data organizing non-intercrossed data of this J more
Packet reception order), wherein L kind postpones the maximum in the degree of depth is [(L-1) × I+Q] individual operating clock, should
Q is the integer not less than 0, represents this L kind and postpones reckling in the degree of depth.Output data buffer 540 can
It to be first-in first-out buffer or its equivalent device, is used for receiving from dynamic random access memory 530
Many groups solve Corvolutional interleaving (Convolutional De-interleaved) data, with as memory access control
Device 520 carries out data output buffer during release of an interleave computing.
For example, referring to Fig. 6 a, organize non-intercrossed data according to transmission sequence is (A more0,B0,C0,D0)、
(A1,B1,C1,D1)、(A2,B2,C2,D2)、(A3,B3,C3,D3)、(A4,B4,C4,D4)、(A5,B5,C5,D5)、
(A6,B6,C6,D6) etc., often organize packet and contain 4 (i.e. L) pen data, wherein subscript represents data group number
(i.e. J), the different pen data of A, B, C, D respectively corresponding 3rd, the 2nd, the 1st, delay degree of depth of 0 unit
With different transmission path, e.g. 1 operating clock of the delay degree of depth of each unit, and same group do not interlock
Adjacent 2 pen data (such as A in dataJWith BJDelay depth difference corresponding to) is away from individual for 1 (i.e. I)
Operating clock, the 3rd, those non-intercrossed datas (comprise the 2nd, the 1st, 0 list through a Corvolutional interleaving processing unit 600
Positions etc. four kinds postpone four transmission paths of the degree of depth (3D, 2D, 1D, Bypass) the 610th, the 620th, the 630th,
640) after processing, sequentially become many group Corvolutional interleaving data (×, ×, ×, D0) (× refer to reception order earlier
Data), (×, ×, C0,D1)、(×,B0,C1,D2)、(A0,B1,C2,D3)、(A1,B2,C3,D4)、(A2,
B3,C4,D5)、(A3,B4,C5,D6) etc., that is the convolution friendship received by Input Data Buffer 510
Wrong data, it is further inputted to dynamic random access memory 530 to carry out release of an interleave computing.
Hold above-mentioned, be that the continuity of restoring data (also will become (the solution friendship that do not interlocks by Corvolutional interleaving data convert
Wrong) data), memory access controller 520 can be in response to Corvolutional interleaving processing unit 600 (or its equivalence
Unit) corresponding to the 3rd, the 2nd, the 1st, 0 unit etc. four kinds postpone the degree of depth and perform release of an interleave, refer to figure
6b, the solution Corvolutional interleaving device 500 which is the present invention processes Corvolutional interleaving data to export solution Corvolutional interleaving data
Schematic diagram, wherein going through the Corvolutional interleaving data of 3 unit delay degree of depth is directly output (Bypassed)
(i.e. without postpone output, premise be minimum delay degree of depth Q be 0, if Q is not 0, all Corvolutional interleaving numbers
According to all can postpone output), other Corvolutional interleaving data then postpone export (Non-bypassed), now deposit
Access to store controller 520 can determine the Corvolutional interleaving data postponing output in dynamic random access memory
Storage address in 530, the storage address of the Corvolutional interleaving data that every is stored and this pen data institute
Corresponding the degree of depth (being the delay degree of depth of the 3rd, 2 or 1 units in this example), aforementioned L (often group are postponed
Data stroke count or say postpone the degree of depth aspect sum) value (being 4 in this example), aforementioned J (non-intercrossed data
Data group number) value (in this example, the subscript with this pen data represents) and aforementioned I (postpone the degree of depth
Gap) value (being 1 in this example) be related to.In more detail, as shown in Fig. 5～6b, in transmission end convolution
Many groups Corvolutional interleaving data that staggered processing unit 600 is exported (×, ×, ×, D0)、(×,×,C0,D1)、(×,B0,
C1,D2)、(A0,B1,C2,D3)、(A1,B2,C3,D4)、(A2,B3,C4,D5)、(A3,B4,C5,D6) etc.
It is sequentially transfused to solve Corvolutional interleaving device 500 in receiving terminal, and be stored by access controller 520 and accessed,
Often organize data the 1st (in this example, be in Fig. 6 b ×, ×, ×, A0、A1、A2A3Portion
Point, corresponding is the part that the minimum delay degree of depth is 0 unit at receiving terminal, corresponding in transmission end be maximum
Postpone the part of the degree of depth) it is either directly output to output data buffer 540, remainder data can sequentially be accessed
In (L-1) × (L-1) the individual data storage unit of dynamic random access memory 530 (or L × L number
According in the individual unit of (L-1) × (L-1) storing in unit, each unit can store Corvolutional interleaving data, its
Middle subtrahend 1 represents and with directly exporting in group intercrossed data, without storing the data of unit, is in this example
(L-1) × (L-1)=3 × 3).
Refer to Fig. 7, in very first time unit (e.g. a period of time or comprise multiple time point) T1
When, 3 × 3 data store the first row (column) data (,,) in unit, and (refers to storage
Deposit order data earlier) read after, with the first stroke count in first group of intercrossed data in input data
According to "×" alignment, jointly as first group of solution convolution data (×,,,) with output；In input data
First group of intercrossed data other pen data (×, ×, D0) be written into, and corresponding B, C and D stroke count
Row according to predetermined storage location.
When the second chronomere T2,3 × 3 data store the second row data in units (×,,)
After reading, align with the first stroke data "×" in second group of intercrossed data in input data, jointly
As second group of solution Corvolutional interleaving data (×, ×,,) with output；Second group in input data interlocks
Other pen data of data (×, C0,D1) be written into, and the storage that corresponding B, C and D pen data is predetermined
Deposit the row of position.
When the 3rd chronomere T3,3 × 3 data store the third line data in units (×, ×)
After reading, align with the first stroke data "×" in the 3rd group of intercrossed data in input data, jointly
As the 3rd group of solution Corvolutional interleaving data (×, ×, ×,) with output；The 3rd group in input data interlocks
Other pen data (B of data0,C1,D2) be written into, and corresponding B, C and D pen data is predetermined
The row of storage location.
When the 4th chronomere T4,3 × 3 data store the first row data (B in unit0,C0,D0)
After reading, with the first stroke data " A in the 4th group of intercrossed data in input data0" alignment, jointly
As the 4th group of solution Corvolutional interleaving data (A0,B0,C0,D0) with output；The 4th group of staggered number in input data
According to other pen data (B1,C2,D3) be written into, and the storage that corresponding B, C and D pen data is predetermined
Deposit row.The access of remaining the 5th to the 7th chronomere T5～T7 and subsequent chronomere is stated on can depending on
Bright analogize.
Hold above-mentioned, corresponding 3 data A having postponed the degree of depthJCan be either directly output, corresponding 2 postpone
Data B of the degree of depthJStorage address be positioned at that 3 × 3 data store in units the 1st arrange { [J mod (L-1)]
× I+1} position (mod represents complementation), corresponding 1 data C having postponed the degree of depthJStorage place
Location is positioned at 3 × 3 data and stores the 2nd in units and arrange { [J mod (L-1)] × I+1} position, correspondence 0
Individual data D having postponed the degree of depthJStorage address be in 3 × 3 data and store the 3rd in units and arrange the
{ [J mod (L-1)] × I+1} position, wherein every row residing for pen data is the numerical value of the maximum delay degree of depth
(being 3 in this example) deducts the numerical value postponing the degree of depth of this pen data, and according to above-mentioned, the present embodiment order is every
The data of individual access address are just read before being written, and through staggered reading address, write ground
Location and extra reserved storages unit are with the behavior of analogue delay buffer element and make reading address align in together
The address of row access.Note that above-mentioned term " OK ", " row ", " position " etc. describe term is to use
Represent the relativeness between access address, non-limiting in the entity circuit relationships of memory 530.
Note that aforementioned memory access controller 520 (is for example schemed in storage one first Corvolutional interleaving data
Data C of 72Before), read one second Corvolutional interleaving data (data C of such as Fig. 70), this is first years old
With the second Corvolutional interleaving data belong to this many Corvolutional interleaving data postpone output data, corresponding as
The degree of depth and the different address that is separately stored in this dynamic random access memory 530 are postponed (such as Fig. 7
Shown in), and this memory access controller 520 should with write after reading this second Corvolutional interleaving data
Before first Corvolutional interleaving data, this Corvolutional interleaving data (data D of such as Fig. 7 of read-out n pen0,
Now first and second Corvolutional interleaving data are C2、C0The delay degree of depth corresponding to) with this first and second
The delay degree of depth corresponding to Corvolutional interleaving data is different, and this n is the integer not less than zero.
For ease of understanding the difference of the present invention and prior art, the memory that Fig. 8 depicts in prior art is deposited
Take mode, many group Corvolutional interleaving data identical with Fig. 7 (×, ×, ×, D0)、(×,×,C0,D1)、(×,B0,C1,
D2)、(A0,B1,C2,D3)、(A1,B2,C3,D4)、(A2,B3,C4,D5)、(A3,B4,C5,D6A in)J
Data can be either directly output and remainder data can the continuity of interdependent memory address be accessed, in short, according to
The embodiment of known technology, the data of each access address unit at the same time (such as Fig. 8 when
Between unit T1～T7 arbitrary) when, it is known that technology can carry out the operation of write-after-read to same access site,
Use the behavior of analogue delay buffer element；Therefore the embodiment according to known technology, all of reading ground
The position of row and row is changed together with writing address in location.Owing to this partly belongs to known skill, describe in detail
Give memorandum at this.
In reality application, often organize Corvolutional interleaving/solution Corvolutional interleaving data and (or say J group Corvolutional interleaving/solution
Corvolutional interleaving data) stroke count be typically larger than the stroke count of previous example, as a example by lifting Fig. 9, when often organizing data
Stroke count is 32 (i.e. L=32), the minimum delay being respectively 0 and 31 unit with the maximum delay degree of depth, even
The delay depth difference of continuous two pen data, away from the delay (i.e. I=1) for 1 unit, is used for storing up according to the present invention
The data of the SDRAM of deposit data store unit can for (L-1) × (L-1)=31 × 31 unit (or say L ×
31 × 31 units in the unit of L=32 × 32), under above-mentioned setting, the present invention is in write J group
(L-1)=31 Corvolutional interleaving data (being labeled as w in Fig. 9) are front, can directly export (Bypass) 1
(being labeled as b in Fig. 9, it is simultaneously belonging to J group and solves in pen J group Corvolutional interleaving data one
Corvolutional interleaving data) and read (L-1)=31 Corvolutional interleaving data (being labeled as r in Fig. 9), mat
This 32 pen data is exported by this as one group of solution convolution data, if the same column of SDRAM accesses in this example
(Same Row Accessing) depth capacity be equal to L=32 (that is have 32 pen data can in same arrange into
Line access, it is not necessary to enter ranks access conversion (Row Access Change)), and 32 stroke counts of same row
According to being to be contained in a same column access/memory unit (referred to as Tile comprises 4 × 8=32 data and stores unit)
In, then Fig. 9 need 4 × 8 Tile (Tile the 11st, the Tile being defined such as the thick wire of Fig. 9 the 12nd ...,
Tile the 83rd, Tile 84) carry out data access, now export J group and solve the row needed for Corvolutional interleaving data
Access conversion times be 4 (i.e. comprise the Tiles of r the 11st, the 12nd, 13 and 14 number), in addition, write
(L-1)=31 Corvolutional interleaving data are handed over (L-1) the pen solution convolution as follow-up other the different groups being read
The row access conversion times of wrong data is 9 (i.e. to comprise the Tiles of w the 11st, the 21st, the 32nd, the 42nd, the 53rd, the 63rd,
74th, the number of 84 and 14), therefore the present invention is producing the accessing operation of J group solution Corvolutional interleaving data
In, total row access conversion times is 4+9=13 time.Each same column access/memory unit above-mentioned stores J group
Solve a part for Corvolutional interleaving data corresponding to Corvolutional interleaving data, and the Corvolutional interleaving data of every this part
Transmission sequence be connected to continue.
Note that in example, L × L=32 × 32 data store unit and constitute the individual Tile of M (4 × 8=32),
And each Tile can store (L × L/M)=32 × 32/32=32 Corvolutional interleaving data, wherein N=32/8=4
Individual Tile is stored by access controller (controller 520 of such as Fig. 5) and is used for reading J group (L-1)=31
Corvolutional interleaving data, and before this 31 pen data is read, in those data
[(L/N)-1]=[(32/4)-1]=7 is stored in (such as the Tile 11 of Fig. 9) in a same column access/memory unit,
Additionally (N-1) L/N=(4-1) 32/4=24 pen be averagely stored in the same column access/memory unit of (N-1)=3 (as
The Tiles of Fig. 9 the 12nd, 13 and 14).Separately note that please continue to refer to Fig. 9, suitable by the reading of data
For sequence, the 1st of above-mentioned J group L=32 pen solution Corvolutional interleaving data is either directly output, remaining (L-1)=31
The kth pen data of pen data is from (xr,yr) individual Tile is read, wherein k is 1 to arrive (L-1) integer
One of them, xrLateral attitude (that is the x of Tile when representing digital independentwIt is 1 to L/Tc (in this example
L/TC=32/4=8) one of them of integer), yrWhen representing digital independent Tile lengthwise position (that is
yrIt is 1 to L/TR(L/T in this exampleR=32/8=4) one of them of integer), or rather, xr=int{{ [J
mod(L-1)]×I}/TC+ 1, yr=int (k/TR)+1, int refers to round numbers computing；In addition, by data
For reception order, the 1st of above-mentioned J group L=32 Corvolutional interleaving data (is also J group simultaneously
Solve the 1st of Corvolutional interleaving data) it is either directly output, the kth pen data of remaining (L-1)=31 pen data is
It is stored in (xw,yw) in individual Tile, wherein xwLateral attitude (that is the x of Tile when representing data writew
It is 1 to L/TCOne of them of=32/4=8 integer), ywWhen representing data write, the lengthwise position of Tile is (also
I.e. ywIt is 1 to L/TROne of them of=32/8=4 integer), or rather, xw=int{ [k+{ [J mod
(L-1)]×I}]/TC+ 1, yw=int (k/TR)+1.It is noted that above-mentioned J, L, k, I, TR、TCDeng
The division of the numerical value example of parameter and Tile and starting point etc. are for understanding that the present invention is used, this area personage
The value of those parameters and the arrangement planning of Tile can be determined according to its demand.
Hold above-mentioned, the stroke count of often organizing data same, postpone the degree of depth with postpone depth difference away from setting under,
As shown in Figure 10, as seen from the figure, prior art utilizes 16 Tile (sizes to the access context of prior art
Being all 32 data and storing unit, shape is only signal) enter line access, export the solution convolution of J group and hand over
Wrong row access conversion times needed for data was 15 (i.e. comprising the Tile number of r), write Corvolutional interleaving number
The row access conversion times solving Corvolutional interleaving data as follow-up other different groups being read according to this is all 15
(i.e. comprising the Tile number of w), therefore prior art is producing the access of J group solution Corvolutional interleaving data
In operation, total row access conversion times is 15+15=30 time, and far above 13 times of the present invention, this means
The access usefulness of the present invention is much better than prior art.Owing to the data acess method shown in Figure 10 is known,
Its details is omitted at this.
In addition to said apparatus invention, the present invention also discloses a kind of solution Corvolutional interleaving method, by a dynamic random
M same column access/memory unit of access memory processes many group Corvolutional interleaving data, these many groups convolutions
Intercrossed data be equally by organize more non-intercrossed data impose one Corvolutional interleaving process form, often organize and do not interlock
Packet pen data containing L, this J group organizing non-intercrossed data comprises continuous L pen data more, this J group
The different degree of depth that postpones of L pen data corresponding L kind respectively, this L and J is positive integer, and the numerical values recited pair of J
Should organize the packet transmission sequence of non-intercrossed data, under above-mentioned setting, an enforcement of this method is for example schemed more
Shown in 11, comprise the steps of
Step S1110: data write one dynamic random postponing to receive of this many groups Corvolutional interleaving data is deposited
In M same column access/memory unit of access to memory, wherein continuous two are written of this Corvolutional interleaving number
According to storage address discontinuous, (L-1) in L Corvolutional interleaving data of this many groups Corvolutional interleaving data
Pen is written in Nw same column access/memory unit in this M same column access/memory unit, this M
For the integer more than 1, and this Nw is for being not more than the positive integer of this M.In this example, above-mentioned (L-1) pen returns
Rotation intercrossed data dispersion is stored in Nw same column access/memory unit, wherein each same column access/memory list
J group Corvolutional interleaving data stroke count stored by Yuan is not more than [int ((L-1)/Nw)+1], and int represents round numbers；
In addition, a delay depth difference corresponding to continuous two Corvolutional interleaving data is away from for I operating clock, and often
The value that the storage address of Corvolutional interleaving data is based on this L, J, I determines.
Step S1120: export in this L Corvolutional interleaving data with as one group of solution Corvolutional interleaving number
According to one, and read in Nr same column access/memory unit through at most Nr time row access conversion
This Corvolutional interleaving data of (L-1) pen with as (L-1) pen in this group solution Corvolutional interleaving data, wherein this Nr
Individual same column access/memory unit is contained in this M same column access/memory unit, and this Nr is for being not more than this M
Positive integer.In this example, Nw is more than Nr.
Another embodiment of the solution Corvolutional interleaving method of the present invention as shown in figure 12, is deposited by a dynamic random
M same column access/memory unit of access to memory processes many group Corvolutional interleaving data, and these many groups convolutions are handed over
Wrong data are by imposing a Corvolutional interleaving and process and form to organizing non-intercrossed data more, often organize non-intercrossed data bag
Pen data containing L, the corresponding L kind of this many groups Corvolutional interleaving data is different postpones the degree of depth, and this L is positive integer,
Under the premise of above-mentioned, the present embodiment comprises the steps of
Step S1210: by data write one dynamic random postponing to receive of above-mentioned many group Corvolutional interleaving data
In M same column access/memory unit of access memory, wherein continuous two are written of this Corvolutional interleaving
The storage address of data is discontinuous, and this M is the integer more than 1.In this example, continuous two convolutions are handed over
Wrong delay depth difference corresponding to data is away from for I operating clock, and this packet organizing non-intercrossed data is transmitted more
Order is J (or saying that often group/every Corvolutional interleaving data corresponding are grouped reception order J), and every convolution
The value that the storage address of intercrossed data is based on this L, J, I determines, wherein I and J is positive integer.
Step S1220: read this M same column access/memory unit through at most Nr time row access conversion
Stored (L-1) Corvolutional interleaving data, with as (L-1) pen data in one group of solution Corvolutional interleaving data, are somebody's turn to do
The positive integer that Nr is no more than M.In above-mentioned (L-1) pen data, [(L/Nr)-1] pen data is had to be stored in together
In row access/memory unit, other (Nr-1) L/Nr pen data are averagely stored in (Nr-1) individual same column access/memory
In unit.
In sum, the solution Corvolutional interleaving device of the present invention and method can use dynamic random access memory
Carry out release of an interleave with cost-effective, and storage can be reduced by the storage address of suitable determination data access
The number of times of device row access mapping, thus improve usefulness.
Although embodiments of the invention are as described above, but those embodiments are not for limiting the present invention, originally
Skilled person can impose change according to the technical characteristic to the present invention for the content expressed or imply of the present invention
Changing, all this kind change all may belong to the patent protection category sought by the present invention, in other words, the present invention
Scope of patent protection must be as the criterion depending on the as defined in claim of this specification.
Claims (18)
1. it is applied to process the solution Corvolutional interleaving device organizing Corvolutional interleaving data, this many groups Corvolutional interleaving more
Packet contains many Corvolutional interleaving data, and this many Corvolutional interleaving data are by executing to organizing non-intercrossed data more
Process with a Corvolutional interleaving and form, often organize non-intercrossed data and comprise L pen data, after processing through this Corvolutional interleaving
Often organize the corresponding delay depth difference of adjacent two pen data in non-intercrossed data away from this L positive integer, this solution is circled round
Interleaver comprises:
One Input Data Buffer, is used for keeping in this many groups Corvolutional interleaving data；
One Memory Controller, by will be temporarily stored into this many groups Corvolutional interleaving data of this Input Data Buffer
It is accessed in a memory to carry out solving Corvolutional interleaving process, the storage of every this Corvolutional interleaving data being stored
Device address is based on corresponding to this pen data and has postponed the degree of depth, the value of this L with this delay depth difference away from certainly
Fixed；And
One output data buffer, is used for the temporary many groups solution Corvolutional interleaving data reading from this memory.
2. the solution Corvolutional interleaving device as described in claim 1, it is characterised in that often organize non-intercrossed data
A corresponding packet transmission sequence J, and the storage address more foundation of every this Corvolutional interleaving data being stored
This J value determines.
3. the solution Corvolutional interleaving device as described in claim 1, it is characterised in that postpone depth difference away from for I
Individual operating clock, this I is positive integer, and in this Corvolutional interleaving data delay degree of depth, the maximum is [(L-1) × I+Q]
Individual operating clock, this Q is the integer not less than 0, represents and postpones reckling in the degree of depth.
4. the solution Corvolutional interleaving device as described in claim 1, it is characterised in that this Memory Controller
The individual data of at least (L-1) × (L-1) are utilized in this memory to store unit to access this many groups Corvolutional interleaving data,
Each data stores unit can store this Corvolutional interleaving data.
5. the solution Corvolutional interleaving device as described in claim 4, it is characterised in that this L × L data storage
Deposit receipt position constitutes M same column chunk, and often this same column chunk can store (L × L/M) this time of pen
Rotation intercrossed data, this M is the positive integer more than 1.
6. the solution Corvolutional interleaving device as described in claim 5, it is characterised in that this Memory Controller
This same column chunk N number of is utilized to read this Corvolutional interleaving data of (L-1) pen with as the convolution of this many groups solution
(L-1) pen data of a group of intercrossed data, and be somebody's turn to do in (L-1) Corvolutional interleaving data, there is [(L/N)-1] pen
This Corvolutional interleaving data are stored in the one of this N number of same column chunk, separately have this convolution of (N-1) L/N pen
Intercrossed data is stored in this N number of same column and takes in (N-1) the individual same column chunk depositing chunk, and this N is
It is not more than the positive integer of this M.
7. the solution Corvolutional interleaving device as described in claim 1, it is characterised in that this Memory Controller
The N number of same column chunk in this memory is utilized to read this Corvolutional interleaving data of (L-1) pen with as this
Many groups solve (L-1) pen data of a group of Corvolutional interleaving data, and are somebody's turn to do in (L-1) Corvolutional interleaving data, have
[(L/N)-1] Corvolutional interleaving data are stored in the one of this N number of same column chunk, separately have (N-1) L/N pen
Corvolutional interleaving data are stored in (N-1) individual same column chunk of this same column chunk N number of, and this N is
Positive integer.
8. it is applied to process the solution Corvolutional interleaving device organizing Corvolutional interleaving data, this many groups Corvolutional interleaving more
Packet contains many Corvolutional interleaving data, and this solution Corvolutional interleaving device comprises:
One Input Data Buffer, is used for keeping in this many groups Corvolutional interleaving data；
One Memory Controller, by will be temporarily stored into this many groups Corvolutional interleaving data of this Input Data Buffer
Being accessed in a memory carrying out a solution Corvolutional interleaving process, and obtaining organizing solution Corvolutional interleaving data more, this is deposited
Memory controller by this many Corvolutional interleaving data to same group solving in Corvolutional interleaving data should be organized more
Solve multiple same column chunks that many pen data of Corvolutional interleaving data are stored in this memory；And
One output data buffer, is used for temporary this many groups solution Corvolutional interleaving data reading from this memory.
9. the solution Corvolutional interleaving device as described in claim 8, it is characterised in that this many Corvolutional interleavings
Data be by organize more non-intercrossed data impose one Corvolutional interleaving process form, often organize non-intercrossed data and comprise
L pen data, often organizes adjacent two pen data corresponding in non-intercrossed data and postpones deep after processing through this Corvolutional interleaving
Degree gap, this delay depth difference is away from for I operating clock, and this L and this I is positive integer, these many convolutions
Postponing degree of depth the maximum in intercrossed data is [(L-1) × I+Q] individual operating clock, whole for not less than 0 of this Q
Number, represents reckling in this delay degree of depth.
10. the solution Corvolutional interleaving device as described in claim 8, it is characterised in that this Memory Controller
The individual data of at least (L-1) × (L-1) in this memory are utilized to store unit to access this many groups Corvolutional interleaving number
According to each data stores unit can store this Corvolutional interleaving data.
The 11. solution Corvolutional interleaving devices as described in claim 8, it is characterised in that these the many groups numbers that do not interlock
According to packet transmission sequence be J, the storage address of every this Corvolutional interleaving data is based on this pen data institute
Corresponding one has postponed the degree of depth, the value of this L, this delay depth difference away from the value decision with this J.
12. 1 kinds solve Corvolutional interleaving method, are used for processing many group Corvolutional interleaving data, this many groups Corvolutional interleaving
Packet contains many Corvolutional interleaving data, and this solution Corvolutional interleaving method comprises the steps of
By this many groups Corvolutional interleaving data access in a memory to carry out solving Corvolutional interleaving process, obtain many groups
Solve Corvolutional interleaving data；
Wherein the same group of solution should organized in solution Corvolutional interleaving data is returned by this many Corvolutional interleaving data more
Multiple same column chunks that many pen data of rotation intercrossed data are stored in this memory.
The 13. solution Corvolutional interleaving methods as described in claim 12, it is characterised in that this many groups solution is returned
In rotation intercrossed data, often group solves Corvolutional interleaving data to should the L pen convolution in many Corvolutional interleaving data hand over
Wrong data, and (L-1) the Corvolutional interleaving data in this L Corvolutional interleaving data are stored in this memory
N number of same column chunk, and it is somebody's turn to do this Corvolutional interleaving data of [(L/N)-1] pen in (L-1) Corvolutional interleaving data
It is stored in the one of this N number of same column chunk, be somebody's turn to do remaining part in (L-1) Corvolutional interleaving data and divide
Dissipating and being stored in this N number of same column and take (N-1) the individual same column chunk depositing chunk, this L and N is just whole
Number.
The 14. solution Corvolutional interleaving methods as described in claim 12, it is characterised in that these many convolutions
Intercrossed data be by organize more non-intercrossed data impose one Corvolutional interleaving process form, often organize non-intercrossed data
Comprise L pen data, often organize adjacent two pen data correspondences one in non-intercrossed data after processing through this Corvolutional interleaving and prolong
Slow depth difference is away from, this delay depth difference away from for I operating clock, and this L and this I is positive integer, these many
Postponing degree of depth the maximum in Corvolutional interleaving data is [(L-1) × I+Q] individual operating clock, and this Q is for not less than 0
Integer and represent reckling in this delay degree of depth.
The 15. solution Corvolutional interleaving methods as described in claim 14, it is characterised in that these many groups are returned
Rotation intercrossed data be accessed in this memory with carry out solve Corvolutional interleaving process step in, every be stored should
The storage address of Corvolutional interleaving data according to corresponding to this pen data postponed the degree of depth, this L value with
The value of this I determines.
The 16. solution Corvolutional interleaving methods as described in claim 15, it is characterised in that often organize and do not interlock
The corresponding packet transmission sequence J of data, and the storage address of every this Corvolutional interleaving data being stored is more
Determine according to the value of this J corresponding to this pen data.
The 17. solution Corvolutional interleaving methods as described in claim 12, it is characterised in that this many groups solution is returned
In rotation intercrossed data, often group solves Corvolutional interleaving data to the L pen convolution friendship should organized in Corvolutional interleaving data more
Wrong data, and by this many groups Corvolutional interleaving data access in this memory to carry out solving the step that Corvolutional interleaving is processed
In Zhou, the individual data of at least (L-1) × (L-1) in this memory are utilized to store unit.
The 18. solution Corvolutional interleaving methods as described in claim 12, it is characterised in that this many groups solution is returned
In rotation intercrossed data, often group solves Corvolutional interleaving data to the L pen convolution friendship should organized in Corvolutional interleaving data more
Wrong data, and by this many groups Corvolutional interleaving data access in this memory to carry out solving the step that Corvolutional interleaving is processed
In Zhou, M same column chunk in this memory is utilized to store, and this M same column chunk
Stored unit by L × L data to be constituted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510088389.4A CN105991505A (en) | 2015-02-26 | 2015-02-26 | De-convolutional interleaver and de-convolutional interleaving method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510088389.4A CN105991505A (en) | 2015-02-26 | 2015-02-26 | De-convolutional interleaver and de-convolutional interleaving method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105991505A true CN105991505A (en) | 2016-10-05 |
Family
ID=57038672
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510088389.4A Pending CN105991505A (en) | 2015-02-26 | 2015-02-26 | De-convolutional interleaver and de-convolutional interleaving method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105991505A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1707957A (en) * | 2004-06-07 | 2005-12-14 | 联发科技股份有限公司 | Controller for rotating interleaved or inverse interleaved data symbol |
| CN101267212A (en) * | 2008-02-14 | 2008-09-17 | 威盛电子股份有限公司 | Group bit interleaver and its method |
| CN101904102A (en) * | 2007-12-21 | 2010-12-01 | 高通股份有限公司 | Efficient address generation for pruned interleavers and de-interleavers |
| CN102318249A (en) * | 2011-07-29 | 2012-01-11 | 华为技术有限公司 | A kind of interweaving conciliate interweaving method, interleaver and deinterleaver |
| CN103856433A (en) * | 2012-12-04 | 2014-06-11 | 晨星软件研发(深圳)有限公司 | Convolutional deinterleaving device and method |
-
2015
- 2015-02-26 CN CN201510088389.4A patent/CN105991505A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1707957A (en) * | 2004-06-07 | 2005-12-14 | 联发科技股份有限公司 | Controller for rotating interleaved or inverse interleaved data symbol |
| CN101904102A (en) * | 2007-12-21 | 2010-12-01 | 高通股份有限公司 | Efficient address generation for pruned interleavers and de-interleavers |
| CN101267212A (en) * | 2008-02-14 | 2008-09-17 | 威盛电子股份有限公司 | Group bit interleaver and its method |
| CN102318249A (en) * | 2011-07-29 | 2012-01-11 | 华为技术有限公司 | A kind of interweaving conciliate interweaving method, interleaver and deinterleaver |
| CN103856433A (en) * | 2012-12-04 | 2014-06-11 | 晨星软件研发(深圳)有限公司 | Convolutional deinterleaving device and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI569587B (en) | Convolutional de-interleaver | |
| CN102208005A (en) | 2-dimensional (2-D) convolver | |
| CN107800644A (en) | Dynamically configurable pipelined token bucket speed limiting method and device | |
| CN105490776A (en) | Interleaving method and interleaver | |
| CN102520902B (en) | Parallel write-in multi-FIFO (first in, first out) implementation method based on single chip block RAM (random access memory) | |
| CN101847986A (en) | Circuit and method for realizing FFT/IFFT conversion | |
| CN101923522B (en) | Storage controller and processing method of data operation commands | |
| CN109726154A (en) | It executes the inverted method of multiple data/address bus and executes the storage component part of this method | |
| CN104579240B (en) | A kind of wave filter, electronic equipment and the filtering method of the configurable coefficient based on FPGA | |
| CN102541769B (en) | Memory interface access control method and device | |
| CN101697491A (en) | Method for realizing convolution interleaving and de-interleaving of time domain by using SDRAM | |
| CN106330741A (en) | Message transmission method and device | |
| CN103544111B (en) | A kind of hybrid base FFT method based on real-time process | |
| CN100568382C (en) | Push-up storage | |
| CN1201234C (en) | Multichannel FILO data buffer storage devices | |
| CN105991505A (en) | De-convolutional interleaver and de-convolutional interleaving method | |
| CN102118304B (en) | Cell switching method and cell switching device | |
| CN103490855B (en) | The method and apparatus of transmission information | |
| US8923315B2 (en) | Packet router having a hierarchical buffer structure | |
| CN102611667B (en) | Stochastic accessing detection FFT/IFFT treatment process and device | |
| CN103186476B (en) | A kind of data cache method and device for multithread | |
| CN1157514A (en) | Device and method for multiplexing cells of asynchronous transmission mode | |
| CN103916140A (en) | Method and device for achieving convolution interleave/de-interleave | |
| TWI617139B (en) | Convolutional de-interleaver and convolutional de-interleaving method | |
| CN106445836B (en) | Asynchronous data reading/writing method, system and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20161005 |