WO2010145078A1 - 一种lte中并行turbo译码的方法及装置 - Google Patents
一种lte中并行turbo译码的方法及装置 Download PDFInfo
- Publication number
- WO2010145078A1 WO2010145078A1 PCT/CN2009/072339 CN2009072339W WO2010145078A1 WO 2010145078 A1 WO2010145078 A1 WO 2010145078A1 CN 2009072339 W CN2009072339 W CN 2009072339W WO 2010145078 A1 WO2010145078 A1 WO 2010145078A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- sheep
- memory
- information
- bit
- component
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/37—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
- H03M13/45—Soft decoding, i.e. using symbol reliability information
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/37—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
- H03M13/39—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes
- H03M13/3972—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using sliding window techniques or parallel windows
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/29—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes combining two or more codes or code structures, e.g. product codes, generalised product codes, concatenated codes, inner and outer codes
- H03M13/2957—Turbo codes and decoding
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/37—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
- H03M13/39—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes
- H03M13/3905—Maximum a posteriori probability [MAP] decoding or approximations thereof based on trellis or lattice decoding, e.g. forward-backward algorithm, log-MAP decoding, max-log-MAP decoding
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/65—Purpose and implementation aspects
- H03M13/6522—Intended application, e.g. transmission or communication standard
- H03M13/6525—3GPP LTE including E-UTRA
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/65—Purpose and implementation aspects
- H03M13/6561—Parallelized implementations
Definitions
- T bo type which is characterized by its bit nature, is very close in bit performance.
- T Bo BCJR (or algorithm), which reduces the general og-algorithm in engineering, is the (ob north y, maximum probability) algorithm.
- 1 shows the principle of T bo .
- T bo consists of (S ) and 2, which are the same used in the device.
- the og-algorithm can be represented by the following formula, respectively.
- S is the engraved register.
- the first is S, the bit, and the register is moved to S.
- k is the first probability of 4k, the force bit, the bit, the force bit, the bit, the AWG.
- the wood to be solved by the Institute is to provide a method and device for T Bo in T, to reduce and increase the rate.
- this provides the T Bo device in T, including Yu storage, control and , among them
- the storage the control of the root control performs the following operations, respectively storing the first information of the bit-receiving storage sheep stored in each of the working system bits in the secondary component,
- a number of peer sub-components of Yu, in the sub-component, according to the algorithm will be divided into several, the branch metric Y, the forward vector vector, the likelihood ratio R, the first information, the first information, the first information Save to the storage line, complete the component several times, and send the likelihood ratio R to
- Control, control and operation household component, control, household storage control, control, and anti-storage and
- the storage includes a memory controller sheep, a first information memory sheep, a unified bit memory sheep element and a bit memory sheep, wherein
- Memory controller sheep control and control of each memory of the household
- the bit memory stores the bits, including the bite memory, the second bit memory, and the first bit, the first bit memory to the first bit to the second bit, the second bit memory the second bit to the other, the first
- the control is connected to the control controlling the first control.
- the second component and the second component respectively have a bit bit, and the second bit force includes a bit memory, a second bit, and a second bit.
- the bit memory has a direct to the second
- the second plurality of controls are connected to the control second multiplexer for the root control to control the second component to the unidirectional bit, and the second component to the contiguous bit
- the information storage sheep stores the result of the secondary component separately, including the first information memory, the second information memory, the first device, and the third plurality, the first information from the first information memory to the second second information memory.
- the second first information to the third more third control is connected to the control third
- the root control controls the first information to the second information and the first information.
- Step, the bit memory, the bit memory, and the second bit memory are respectively composed of small and small memories, each small memory address first information memory and second first information storage Each consists of a small amount of small memory, each small memory address.
- Steps including the sheep, the first multiplier and the second, the sheep receives the stored components for several times, and the completion of the results to the first, the most control is connected to the control, the first
- the control of the control is controlled by the first component first information to the first face information memory, the second component, the second first information to the second, the second second second information channel to the second prior information memory, and another path to.
- the sheep includes a number of A sheep in the finished weight , and the A sheep consists of Y sheep , sheep , memory , second Y sheep , sheep , and sheep .
- the branch metric of the sheep's force, the branch metric obtained by using the branch to measure the second Y sheep's force, and the resulting forward branch metric is used to store the obtained O sheep in the forward direction.
- Vector R sheep is like likelihood ratio and first information.
- R sheep includes group 16 3 instruments, and group 8 m sheep, second group 4 m sheep, third group 2 m sheep, and subtractor, phase 3 method group method, total 8 and value respectively
- Steps including the judgment sheep, the cut sheep and the memory controller sheep, wherein the first information received by the judgment sheep is sent to the sheep and the memory controller sheep respectively, and the conclusion of the sheep judgment is satisfied, and the control is satisfied. Or the unsatisfied anti-satisfaction cutoff control to the memory controller sheep, by the memory controller Yang Yuanyu results.
- this method provides T in the T "bO, including
- Step, the second contains the sub-component, the second component of the second component, the second component of the previous component, and the second
- the second component of the bit, the first face information obtained by the previous component, and the first component of the second bit of the second bit are first information 0.
- Step as long as the result satisfies the following, satisfies the remaining CRC result of the cutoff. Step, length 512, 512 ⁇ 1024, 2 1024 ⁇ 2048, 4 2048 ⁇ 6144, 8.
- Step in the algorithm according to a certain, the core is divided into several, of which
- the vector is 0
- the proposed segmentation og-AP algorithm T bo method and hardware device increase the rate, reduce the consumption of hardware resources with a small amount of hardware resources, and satisfy the T-combination requirements of the T system.
- Hugh, Ben has the following
- the algorithm is made up of .
- Be ede o et al. proposed a fixed algorithm.
- the og-AP algorithm will divide the sub-lengths of the sub-goals.
- the algorithm is still og-AP algorithm, and the difference is used in the tail of the meson.
- it shows that there is a huge gap between the T bo of the 1o9 hill algorithm and the 1 bbps rate specified by T.
- This method proposes a magical segmentation 1og-AP algorithm T bo .
- the device 3 including the storage, control, and
- the branch metric Y, the forward vector vector, the likelihood ratio R, the first information of the bit, the bit and the first information, the first information is saved to the storage line, the number of times is completed, and the likelihood ratio R is sent to the instance.
- the component can be divided into at least a sub-component. Wherein, the unit of the second component stored in the second, the second pre-information (ie, the result of the last sub-component in the last time) and the second-order component stored in the second bit, the first information (ie, the first component is also The result of the last component) and the second bit
- Control control and various operations, household components Control, household storage control, control, and anti-storage and or stop, in the likelihood ratio decision, whether the judgment result meets the cutoff, the control is reversed, and the result meets the deadline and the result is forceful.
- T bo device proposed by the segmentation og-AP algorithm.
- the storage includes a memory controller sheep, a first information memory sheep, a bit memory sheep, and a bit memory sheep, wherein
- the information memory stores the results of the sub-components separately, as shown in 3, the steps include the first information memory 1, the first information memory 2, 1 and more 3, and the first information of the first information memory 1 is more
- the first information memory 1 is used to store the component result first information of the second component
- the first information in the second component C2 is used to store the component result second second information of the second component C2
- the second component first information ie, the result of the last component
- the control of the root control is to the second information (in the second component) and the first information (in the second component C2).
- the bit memory is stored in the partition of the storage, as shown in 3, the step includes the bit memory, 1 and 2, the bit memory has, directly up to 2, and the other 2 , 2 more controls are connected to the control.
- the system bit memory is used to store the divided bits.
- the first information is stored in the sheep, but in others, it can be another.
- bit memory is stored in the bits, as shown in 3, and the steps include bit memory 1, bit memory 2, and more.
- Bit memory 1 Yu Bits up to 1 Bit memory 2 Yu 2 bits up to 1 other, more
- the control of 1 is connected to the control.
- Bit memory 1 is used to store the memory controller.
- the first bit memory 2 of Yang Yuanyu is used to store the second bit of the memory controller Yang Yuanyu.
- the control of the multi-control 1 controls the second component C and the second component C2 to work on the first bit and the second bit, respectively.
- the memory controller sheep, the control of the control, the control of each memory of the household, and the control points are stored ( ) to the bit memory sheep.
- the 3 memories are composed of small memories of 8 s, respectively, 8 memory addresses, and in turn, 8 small memory addresses.
- 8 Small memories are combined into one large memory, that is, system bit or bit memory 1, or bit memory 2.
- the improved swallow the memory that can be operated by the memory, that is, the capacity of the small memory supports the size required for operation, the maximum length of T is 6144, the average is divided into 8 and the size is 768, and the small memory holds 1 size and 768, and the support operation will be Small memory 7682 is 1536 city size, storage
- the width is determined by the bit or bit or the first information.
- the length of the control determines the equal division, and the different lengths can be 1, 2, 4 or 8.
- the memory controller sheep will separately store the small memory of the bit memory, and the memory will store the equal-sized number of the same size.
- the information memory 1 is similar to the above-mentioned memory of the information memory 2.
- the memory is composed of 8 small memories and supports operations.
- the size of the small memory is 7682, that is, 1536, and the width of the memory is equal to the first information.
- the first information memory 1 and the first information memory 2 support 8 small memories and addresses constituting the first information memory, and can also be trusted.
- the 8 small memories that make up the bit memory share the address and number, and the small memories are sequentially activated, and the enable signals are sequentially generated, that is, equally divided, the first small memory, the small small small memory, The second small memory, the second small, in turn, until.
- the address is generated by the base address (in memory and memory) and the offset address (in the location or memory), the address base address offset address is sequentially increased by the offset address of the small memory, and the address 0 1 is the small memory 0 address, to The lowest address of the small memory is cut off.
- the base address generation operation operation can be controlled by the memory controller, and the memory controller can operate to generate the base address of the memory, the operation time base address 0, and the operation time base address 768. , the generation of the control requires the previous execution decision, in the second component (that is, usually said), the address direct address (ie not needed) in the second component C2
- each small memory can be the same, address and number, direct address generation, base address operation control decision, operation time base address 0, operation time base address 768, offset address meson memory is the same, 0 - ( Length, aliquot), address base address offset address.
- the address is sent directly, and the address of the household.
- bit memory 1 and the bit memory 2 are the same as the bit memory, but the time is based on the direct address.
- the second component C is the same as the bit memory, but the time is based on the direct address.
- bit memory 1 To enable the bit memory 1 to teach, in the second component C2, the bit memory 2 is taught.
- the memory controller sheep is responsible for generating the control of the information memory 1, 2 .
- the result of the first component C of the a priori information memory 1 is, according to the direct address, the first information generated by the live operator, respectively, of the first memory of the information memory 1.
- the time of the information memory 1 is the address of the address, that is, the second component of the information memory 1
- the operation is the same as the bit memory.
- the result of the second component C2 of the information memory 2 when the address is the address, that is, the time, the direct address, the second component C.
- the operation is the same as the bit memory.
- the controller in 3 the control of the household, the control is mainly used to control the execution (front, direction vector)
- R can be trusted, etc.) to generate memory controller sheep control (operation control, etc.) and memory controller sheep control respectively to store and use to generate various controls near the root
- the anti-generation energy of the cut-off sheep is the control of whether or not it is controlled, and it is the control to generate other control signals.
- the result of receiving the cutoff of the sheep meets the deadline, the control result, the letter to the storage, stop, T bo, ie A
- control will be stored in anti-storage. .
- the phase of the sum of the controller does not control the sheep, and controls the multiple times.
- the controller is close to generating segmentation control, length, and sub-goats.
- the result (including the first information and the second prior information) is at most 4. More than 4 controls are connected to the control.
- the control of the multi-four control controls the second component C and the second component C2, respectively, the first information and the second information, that is, in the second component C, the fourth information, the first information, the first information, the second information, the second component, C2, First information to the 2, 2 will be the second first information to the information memory 2, the other way to the judgment sheep.
- the function of the first component C and the second component C2 shown in sheep 1 is the same as the "segmented og-A algorithm, where the C and C2 components are the same.
- the sheep's force bit the second first message, the first bit, the result direct address storage information memory 1.
- C2 the unit of the sheep's force, the first information, the second bit, the result address first information memory 2. 2 (ie C and C2), complete T bo .
- the sheep contains a number of A sheep, A sheep in the finished weight, and more A sheep can support.
- the example contains 8 sheep (shown in 5), which can support the maximum force of 8, in the case of less than 8, you can only teach the sheep.
- a sheep consists of sheep 1, sheep, memory, sheep 2, sheep, and R sheep. And respectively forward vector and vector. among them
- the branch metric obtained by the branching measure of the sheep 1 is used to measure the branch metric of the 2 sheep force with the branch, and the obtained forward metric is used to store the sheep in the vector memory.
- the depth of the memory is equal to the length of the window. , the memory equals the result, the memory
- the RA 4 is composed of 8 small memories to support 8 kinds of sheep in the forward vector R to the likelihood ratio and the first information (including the first information and the second information). Without the algorithm, the memory size of the saved result is the same as the size, and the size increases and increases.
- the method can control the size of the memory within the desired amount, and the required memory only needs to be equal to the length of the medium, not the size.
- T bo 3 shift register that is, only 8 , the front and each have 8, and the special and (capable 0, can be 1) phase, the different special, that is, the special one shown in 7, 2, the phase 8
- special branches There are 16 special (special branches), but only 5 has 4 god branch metrics, so you can calculate 4 branch metrics, respectively, in the median period.
- R sheep's hardware 8 is shown, including group 16 3 gauges, and the first group of 8 m sheep, the second group of 4 m sheep, the third group of 2 m sheep, and the subtractor, the phase of the 3 method group method, The value of 8 and the value of the m sheep group m5 in the 8 m sheep group m sheep of the first group m sheep, respectively, the results are respectively to the 4 m sheep in the second group m sheep, the second group m sheep, the phase m sheep The group m, 2 results to the subtractor, and the likelihood ratio is obtained. The likelihood ratio is subtracted from the unified information and the first information to obtain a new prior information.
- Steps 1 to 5 flow the waters, and the steps are in the middle of the cycle. It can guarantee the sheep cycle of R.
- the R result of the second component C2 of the judgment sheep is greater than the judgment 1, otherwise 0.
- the CRC (y) result T
- the CRC decision can be made as to whether or not the CRC result is closed.
- the sheep can also be cut off, and the CRC is cut off.
- the second component C2 requires a bit
- the second component in the second component C2 requires the first information
- the second first information is sent to the first information memory 2, and the judgment is judged.
- the second and second can be, because the C2 system bit memory and the first information memory 1 are the same address.
- Step the method can include the following steps
- the preamplifier works, and the memory (including the bit memory, bit memory 1, 2) can receive new, new, new and full memory, and wait. Support operations, so you can save up to the same
- step 8 If the result of the sheep is up to the end, it is the execution of step 8, otherwise, the second, 4, 5 steps
- T Bo Before T Bo , A Yang Likong, whether there is new valid in the memory, there is new, otherwise wait. Hugh, because the length of T ( ) is 40 to 6144, the length difference is large, so the difference is also large, and the length is larger.
- the examples are 512, 512 024, 2 1024 2048, 4 2048 6144, 8, which represent the length.
- this k + is not at all , equal to the k+1 , probably in any godlike , so , put . . ten. 0 force. 0, , , because the confidence may not move high, can not be used as
- the vector of the middle window is the same as the algorithm of the first window.
- the % synchronization likelihood is the same as the R, , and execution.
- the vector of the most 1 window is the most 1 (the first, 2nd, 4th, 8 8th), then the most mediated vector of the most, 0. Other
- the most methodatic method of the force value cannot be 0, that is, it needs to be prior to
- the confidence of the first media has reached a higher level, which can be used for the former, and the innermost window of the pre-teaching is obtained. Then, the vector is sequentially obtained and saved to the phase memory.
- the forward vector of the most 1 window compares the sheep, and the forward vector of the middle window is the same, and the most forward vector of the window performs the same. Synchronization likelihood ratio R.
- T bo internal memory (including system bit memory, bit memory 1, 2 force). Therefore, both the memory and the memory are empty, and the length of 20 512 is allowed, the length is 512, and only the bit memory, the memory of the memory 1, and the 1 small memory, that is, the memory of all the small memory is the same (that is, the same as the base address 0). ). Finished, the memory is valid, and the nuclear memory force does not allow Chang, wait. However, the sheep is empty, . This second
- the second length of this waiting is 1024, because the memory has the same as the sheep, the length is 1024, the same update of the work (length, ), 2, so 1 and 2 sheep, A sheep memory, first information memory Get advice A. Until the arrival or satisfaction of the cut, before. of ,
- the sheep is empty, and the memory of the same phase allows Chang.
- the length of the phase, the phase of the sheep and the phase of the memory With the A sheep inside the full water and water, wood, speed. Make possible shrinking while improving the swallowing.
- the proposed T bo method and the phase hardware device have high performance and are well suited to meet the low throughput requirements of the TE system.
- the segmentation og-AP algorithm T bo the segmentation og-AP algorithm is a sum of 1o9 small P algorithm and algorithm, can support, and reduce, increase rate, pass and, segmentation og-AP algorithm can The rate of exchange for smaller scale and storage capacity is doubled, especially for FPGA/AC hardware high-speed T"bO coder to meet T system performance requirements.
Landscapes
- Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Error Detection And Correction (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Description
Claims
Priority Applications (6)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13/258,985 US20120106683A1 (en) | 2009-06-18 | 2009-06-18 | Method and apparatus for parallel turbo decoding in long term evolution system (lte) |
CN2009801587138A CN102396158A (zh) | 2009-06-18 | 2009-06-18 | 一种lte中并行turbo译码的方法及装置 |
KR1020117027415A KR101225016B1 (ko) | 2009-06-18 | 2009-06-18 | Lte에서 병렬 터보 디코딩의 방법 및 장치 |
PCT/CN2009/072339 WO2010145078A1 (zh) | 2009-06-18 | 2009-06-18 | 一种lte中并行turbo译码的方法及装置 |
EP09845992.8A EP2429085B1 (en) | 2009-06-18 | 2009-06-18 | Method and apparatus for parallel turbo decoding in long term evolution system (lte) |
JP2012511116A JP5479580B2 (ja) | 2009-06-18 | 2009-06-18 | Lteにおける並列turboデコーディングの方法及び装置 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2009/072339 WO2010145078A1 (zh) | 2009-06-18 | 2009-06-18 | 一种lte中并行turbo译码的方法及装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2010145078A1 true WO2010145078A1 (zh) | 2010-12-23 |
Family
ID=43355683
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2009/072339 WO2010145078A1 (zh) | 2009-06-18 | 2009-06-18 | 一种lte中并行turbo译码的方法及装置 |
Country Status (6)
Country | Link |
---|---|
US (1) | US20120106683A1 (zh) |
EP (1) | EP2429085B1 (zh) |
JP (1) | JP5479580B2 (zh) |
KR (1) | KR101225016B1 (zh) |
CN (1) | CN102396158A (zh) |
WO (1) | WO2010145078A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106209324A (zh) * | 2016-09-18 | 2016-12-07 | 幻视互动(北京)科技有限公司 | Fpga实现的基于多频无线联网模组的智能头显设备 |
Families Citing this family (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9219956B2 (en) | 2008-12-23 | 2015-12-22 | Keyssa, Inc. | Contactless audio adapter, and methods |
US9960820B2 (en) | 2008-12-23 | 2018-05-01 | Keyssa, Inc. | Contactless data transfer systems and methods |
US9191263B2 (en) | 2008-12-23 | 2015-11-17 | Keyssa, Inc. | Contactless replacement for cabled standards-based interfaces |
US8811452B2 (en) * | 2009-12-08 | 2014-08-19 | Samsung Electronics Co., Ltd. | Method and apparatus for parallel processing turbo decoder |
US9003266B1 (en) * | 2011-04-15 | 2015-04-07 | Xilinx, Inc. | Pipelined turbo convolution code decoder |
US8843807B1 (en) | 2011-04-15 | 2014-09-23 | Xilinx, Inc. | Circular pipeline processing system |
EP2579468B1 (en) * | 2011-10-05 | 2020-05-06 | Telefonaktiebolaget LM Ericsson (publ) | Method and device for decoding a transport block of a communication signal |
US20130170842A1 (en) * | 2012-01-04 | 2013-07-04 | Toshiaki Koike-Akino | Method and System for Equalization and Decoding Received Signals Based on High-Order Statistics in Optical Communication Networks |
US20150033094A1 (en) * | 2013-07-23 | 2015-01-29 | Yuan Ze University | Window-stopped method for applying to turbo decoding |
US9258015B2 (en) * | 2013-12-23 | 2016-02-09 | Apple Inc. | Decoder with selective iteration scheduling |
US9648525B2 (en) * | 2014-08-12 | 2017-05-09 | Qualcomm Incorporated | System and methods for improving intra-frequency cell reselection on a wireless communication device in connected mode |
GB2529209B (en) * | 2014-08-13 | 2021-05-26 | Accelercomm Ltd | Detection circuit, receiver, communications device and method of detecting |
US9602648B2 (en) | 2015-04-30 | 2017-03-21 | Keyssa Systems, Inc. | Adapter devices for enhancing the functionality of other devices |
US10476634B2 (en) * | 2016-03-04 | 2019-11-12 | Huawei Technologies Co., Ltd. | System and method for polar encoding and decoding |
CN107453761B (zh) * | 2016-05-31 | 2021-04-06 | 展讯通信(上海)有限公司 | Turbo码译码方法及Turbo码译码器 |
US10084486B1 (en) * | 2017-09-29 | 2018-09-25 | Intel Corporation | High speed turbo decoder |
US10491244B2 (en) * | 2017-11-14 | 2019-11-26 | Nyquist Semiconductor Limited | Systems and methods for decoding error correcting codes |
US10389388B2 (en) | 2017-12-28 | 2019-08-20 | Apple Inc. | Efficient LDPC decoding with predefined iteration-dependent scheduling scheme |
WO2022186853A1 (en) * | 2021-03-03 | 2022-09-09 | Zeku, Inc. | Dynamic cyclic redundancy check update for iterative decoding |
CN113114278B (zh) * | 2021-03-07 | 2023-03-14 | 西安电子科技大学 | 双二进制Turbo译码实现方法、系统、设备及应用 |
CN116527207B (zh) * | 2023-07-04 | 2023-08-29 | 福建福大北斗通信科技有限公司 | 一种北斗三号通信基带自适应迭代次数的turbo译码方法 |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050273687A1 (en) * | 2004-05-24 | 2005-12-08 | Sung-Jin Park | Apparatus and method for turbo decoding using a variable window size |
CN1728563A (zh) * | 2004-07-28 | 2006-02-01 | 日本电气株式会社 | Turbo译码装置和Turbo译码方法 |
CN1913368A (zh) * | 2005-08-11 | 2007-02-14 | 中兴通讯股份有限公司 | 一种自适应turbo译码的方法 |
CN101026439A (zh) * | 2007-02-07 | 2007-08-29 | 重庆重邮信科股份有限公司 | 一种提高Turbo码译码速率的译码方法 |
CN101409599A (zh) * | 2007-10-11 | 2009-04-15 | 大唐移动通信设备有限公司 | 一种Turbo码译码装置及方法 |
CN101411071A (zh) * | 2006-01-27 | 2009-04-15 | 高通股份有限公司 | 具有双向滑动窗口体系结构的map译码器 |
Family Cites Families (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6754290B1 (en) * | 1999-03-31 | 2004-06-22 | Qualcomm Incorporated | Highly parallel map decoder |
US6980605B2 (en) * | 2000-01-31 | 2005-12-27 | Alan Gatherer | MAP decoding with parallelized sliding window processing |
US20010044919A1 (en) * | 2000-05-05 | 2001-11-22 | Edmonston Brian S. | Method and apparatus for improved perormance sliding window decoding |
US6662331B1 (en) * | 2000-10-27 | 2003-12-09 | Qualcomm Inc. | Space-efficient turbo decoder |
US6813742B2 (en) * | 2001-01-02 | 2004-11-02 | Icomm Technologies, Inc. | High speed turbo codes decoder for 3G using pipelined SISO log-map decoders architecture |
US7200799B2 (en) * | 2001-04-30 | 2007-04-03 | Regents Of The University Of Minnesota | Area efficient parallel turbo decoding |
US7020827B2 (en) * | 2001-06-08 | 2006-03-28 | Texas Instruments Incorporated | Cascade map decoder and method |
JP4191393B2 (ja) * | 2001-06-11 | 2008-12-03 | 富士通株式会社 | 情報記録再生装置及び方法並びに信号復号回路 |
CN1157883C (zh) * | 2001-07-11 | 2004-07-14 | 信息产业部电信传输研究所 | 实现并行滑动窗最大后验概率算法的高速Turbo码译码器 |
JP3888135B2 (ja) * | 2001-11-15 | 2007-02-28 | 日本電気株式会社 | 誤り訂正符号復号装置 |
EP1461869A1 (en) * | 2001-12-28 | 2004-09-29 | Koninklijke Philips Electronics N.V. | Method for decoding data using windows of data |
US7107509B2 (en) * | 2002-08-30 | 2006-09-12 | Lucent Technologies Inc. | Higher radix Log MAP processor |
US7055102B2 (en) * | 2002-12-06 | 2006-05-30 | Sandbridge Technologies, Inc. | Turbo decoder using parallel processing |
JP2005210238A (ja) * | 2004-01-21 | 2005-08-04 | Nec Corp | ターボ復号装置及びその方法並びにその動作プログラム |
JP4692751B2 (ja) * | 2005-11-28 | 2011-06-01 | 日本電気株式会社 | ターボ復号器及びそれを備えた通信システム |
US7929646B2 (en) * | 2006-01-27 | 2011-04-19 | Qualcomm Incorporated | Map decoder with bidirectional sliding window architecture |
US7810018B2 (en) * | 2006-10-27 | 2010-10-05 | Telefonaktiebolaget Lm Ericsson (Publ) | Sliding window method and apparatus for soft input/soft output processing |
JP4874312B2 (ja) * | 2007-09-20 | 2012-02-15 | 三菱電機株式会社 | ターボ符号復号装置、ターボ符号復号方法及び通信システム |
KR101504101B1 (ko) * | 2007-10-02 | 2015-03-19 | 삼성전자주식회사 | 적어도 두 개의 디코딩 매소드를 디코딩하기 위한 asip 아키텍처 |
US8140932B2 (en) * | 2007-11-26 | 2012-03-20 | Motorola Mobility, Inc. | Data interleaving circuit and method for vectorized turbo decoder |
-
2009
- 2009-06-18 CN CN2009801587138A patent/CN102396158A/zh active Pending
- 2009-06-18 EP EP09845992.8A patent/EP2429085B1/en active Active
- 2009-06-18 WO PCT/CN2009/072339 patent/WO2010145078A1/zh active Application Filing
- 2009-06-18 KR KR1020117027415A patent/KR101225016B1/ko active IP Right Grant
- 2009-06-18 US US13/258,985 patent/US20120106683A1/en not_active Abandoned
- 2009-06-18 JP JP2012511116A patent/JP5479580B2/ja active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050273687A1 (en) * | 2004-05-24 | 2005-12-08 | Sung-Jin Park | Apparatus and method for turbo decoding using a variable window size |
CN1728563A (zh) * | 2004-07-28 | 2006-02-01 | 日本电气株式会社 | Turbo译码装置和Turbo译码方法 |
CN1913368A (zh) * | 2005-08-11 | 2007-02-14 | 中兴通讯股份有限公司 | 一种自适应turbo译码的方法 |
CN101411071A (zh) * | 2006-01-27 | 2009-04-15 | 高通股份有限公司 | 具有双向滑动窗口体系结构的map译码器 |
CN101026439A (zh) * | 2007-02-07 | 2007-08-29 | 重庆重邮信科股份有限公司 | 一种提高Turbo码译码速率的译码方法 |
CN101409599A (zh) * | 2007-10-11 | 2009-04-15 | 大唐移动通信设备有限公司 | 一种Turbo码译码装置及方法 |
Non-Patent Citations (1)
Title |
---|
See also references of EP2429085A4 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106209324A (zh) * | 2016-09-18 | 2016-12-07 | 幻视互动(北京)科技有限公司 | Fpga实现的基于多频无线联网模组的智能头显设备 |
CN106209324B (zh) * | 2016-09-18 | 2023-05-19 | 幻视互动(北京)科技有限公司 | Fpga实现的基于多频无线联网模组的智能头显设备 |
Also Published As
Publication number | Publication date |
---|---|
EP2429085A4 (en) | 2013-04-17 |
US20120106683A1 (en) | 2012-05-03 |
JP5479580B2 (ja) | 2014-04-23 |
EP2429085A1 (en) | 2012-03-14 |
KR20120014905A (ko) | 2012-02-20 |
KR101225016B1 (ko) | 2013-01-22 |
CN102396158A (zh) | 2012-03-28 |
JP2012527790A (ja) | 2012-11-08 |
EP2429085B1 (en) | 2018-02-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2010145078A1 (zh) | 一种lte中并行turbo译码的方法及装置 | |
US9632708B2 (en) | Method, apparatus, and system for reading and writing data | |
Tarjan | A simple version of Karzanov's blocking flow algorithm | |
US20080148013A1 (en) | RDMA Method for MPI_REDUCE/MPI_ALLREDUCE on Large Vectors | |
US11252111B2 (en) | Data transmission | |
JP2007295564A (ja) | 低密度のパリティ検査コードを用いた復号器 | |
US20070180352A1 (en) | Memory system and method for use in trellis-based decoding | |
CN101447928A (zh) | 分片信息处理的方法和装置 | |
CN105474168B (zh) | 网络装置执行的数据处理方法和相关设备 | |
CN113328756B (zh) | 用于提升分层qc-ldpc译码器硬件处理性能的方法 | |
CN107612560A (zh) | 基于部分信息比特似然比的极化码早期迭代停止方法 | |
CN109768846B (zh) | 基于二核三核混合极化码的凿孔方法、系统、装置及介质 | |
CN102523076A (zh) | 通用可配置的高速率Turbo码译码系统及其方法 | |
CN1329786A (zh) | 同步跟踪方法 | |
WO2014114246A1 (zh) | 信号检测方法和装置 | |
CN103957211B (zh) | 客户端与前端页面之间的数据通信方法及装置 | |
US20130111305A1 (en) | Channel decoding method and decoder for tail-biting codes | |
CN111597139B (zh) | 一种gpu的通信方法、系统、设备以及介质 | |
CN105471770A (zh) | 一种基于多核处理器的报文处理方法及装置 | |
CN113378194B (zh) | 一种加解密运算加速方法、系统及存储介质 | |
WO2017050232A1 (zh) | 数据传输方法及系统 | |
JP6744982B2 (ja) | OvXDMシステムに適用される一種類のファストデコード方法、装置及びOvXDMシステム | |
CN116466890A (zh) | 固态硬盘利用接口协议提高服务器综合运算速率的方法 | |
WO2017012090A1 (zh) | 一种配置下行解调参考信号dmrs端口的方法及装置 | |
CN114330743A (zh) | 一种用于最小-最大化问题的跨设备联邦学习方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200980158713.8 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 09845992 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 20117027415 Country of ref document: KR Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2012511116 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2009845992 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13258985 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |