CN100536351C - A coding method of convolution code - Google Patents

A coding method of convolution code Download PDF

Info

Publication number
CN100536351C
CN100536351C CN 200610087453 CN200610087453A CN100536351C CN 100536351 C CN100536351 C CN 100536351C CN 200610087453 CN200610087453 CN 200610087453 CN 200610087453 A CN200610087453 A CN 200610087453A CN 100536351 C CN100536351 C CN 100536351C
Authority
CN
China
Prior art keywords
path
sliding window
back
state
time
Prior art date
Application number
CN 200610087453
Other languages
Chinese (zh)
Other versions
CN1968023A (en
Inventor
徐昌平
Original Assignee
华为技术有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Priority to CN 200610087453 priority Critical patent/CN100536351C/en
Publication of CN1968023A publication Critical patent/CN1968023A/en
Application granted granted Critical
Publication of CN100536351C publication Critical patent/CN100536351C/en

Links

Abstract

一种卷积码译码方法,本方法通过在传统的PLVA中引入滑动窗口技术,将需要译码的卷积码分段在各滑动窗口中,在各滑动窗口中分别查找局部优选路径,根据各滑动窗口的局部优选路径,获取全局的优选路径,对所述的各全局优选路径译码进行差错校验,根据校验结果,输出译码结果。 One kind of a convolutional code decoding method, the present method by introducing a sliding window technique in conventional PLVA, a convolutional code segment need to be decoded in the sliding window, in the preferred path to find the local sliding window, respectively, in accordance with preferably paths of the local sliding window, acquires global preferred path, preferably for each global path decode the error check is performed, according to the check result, outputs the decoding result. 通过本发明的技术方案,既保留了PLVA的优点,又能大大减少了PLVA译码所需要占用的存储空间。 By aspect of the present invention, not only retains the advantages of PLVA, but also greatly reduces the storage space required for decoding PLVA occupied.

Description

一种巻积码译码方法 Convolving one kind code decoding method

技术领域 FIELD

本发明涉及译码方法领域,特别是涉及一种巻积码译码方法。 The present invention relates to the field decoding method, in particular, it relates to a method code decoding convolving. 背景技术 Background technique

巻积码是爱里斯(Eiias)在1955年提出的一种信道标码方法。 Convolving code love Rees (Eiias) A channel code standard method proposed in 1955. 巻积码的校验元不仅与本组的信息有关,而且与以前输入至编码器的信息组有关。 Convolving membered check codes only information about this group, and the input to the encoder with the previous information about the group. 同样, 译码过程中也要从前后相关的码组中提取信息,因此,巻积码具有比较好的信能,并且因为编码简单,编码增益高以及具有很强的纠正随机错误能力,因此在通讯系统中得到十分广泛的应用。 Similarly, the decoding process should also extract information from the code group contextually relevant, and therefore, has good code convolving channel energy, and because of the simple coding, coding gain high and has a strong ability to correct random errors, so communication system has been very widely used.

维特比算法(Viterbi Algorithms,简称VA)是一种基于码的网格图基础上的最大似然准则算法,也是在加性高斯白噪声信道下性能最佳的巻积码译码方法,同时由于具有效率高、速度快、译码简单的特点,因此被广泛地应用于各种通信系统中。 Viterbi algorithm (Viterbi Algorithms, referred to as VA) is a maximum likelihood criterion algorithm based on a trellis diagram based code, also in the additive white gaussian noise channel performance convolving the best code decoding method, and because with high efficiency, fast, simple decoding characteristic, and therefore is widely used in various communication systems.

基于VA, Texas Instruments公司在2004年9月公布了一篇题目为"维特比译码器协处理器TMS320C64x DSP参考指南,,(TMS320C64xDSP Viterbi-Decoder Coprocessor(VCP) Reference Guide")的文献(以下简称文献1 )。 Based VA, Texas Instruments company in September 2004 published an article entitled "The Viterbi decoder coprocessor TMS320C64x DSP Reference Guide ,, (TMS320C64xDSP Viterbi-Decoder Coprocessor (VCP) Reference Guide") documents (hereinafter referred to as Document 1). 在文献中,提出了一种应用VA进行译码的译码方法的解决方案。 In the literature, the application proposes a solution for VA decoding method for decoding.

其技术方案具体是,在传统的VA中引入了滑动窗口的技术,以减少译码所需的存储单元。 Particular aspect thereof, the introduction of a sliding window technique in the conventional VA in order to reduce the storage required for the decoding unit. 根据VA的特性,当译码深度d足够大的时候(比如, d=4〜5*K,其中K为巻积编码器的约束长度),所有的幸存路径会议很大的概率在d个时间单位内合并。 The characteristics of the VA when decoding depth d is large enough (for example, d = 4~5 * K, where K is convolving encoder constraint length), high probability of all survivor paths meeting time d the unit consolidation. 当帧长足够大时,把巻积帧分段在滑动窗口中。 When the frame length is sufficiently large, convolving the frame segments in the sliding window. 设滑动窗口的宽度为W=R+C,其中R段为判决输出部分,C段为相邻两个滑动窗口的重叠部分。 Provided sliding window width W = R + C, where R is a para-decision output section, C is the intersection of two segments adjacent to the sliding window. 状态度量以及回溯在整个滑动窗口内进行,仅输出R个判决。 State metric and traceback in the entire sliding window, only the R output decision. 通过这种机制,当帧长较长时,可以显著的减少路径寄存器的存储单元占用。 By this mechanism, when the frame length is longer, can significantly reduce the memory path register unit occupied. 具体是,对于一巻积编码块长度是M,状态数为N的巻积码,应用传统的VA,所需要占用的存储单元为:M*N个状态存储单元,以及M*N个度量存储单元;而应用文献1所述的技术方案,所需要的存储单元为:W + N个状态存储单元,以及\¥*^^个度量存储单元。 Specifically, for a convolving the coding block length is M, the number of states is N convolving code, the application of traditional VA, memory cells need to occupy as: of M * N state storage unit, and of M * N metric memory unit; and the application of an aspect of the document, the storage unit is required: W + N th state storage unit, and \ ¥ * ^^ metrics memory cells. 可见,文献l中所述的技术方案, 所需要占用的存储空间是传统VA的W/M倍。 Seen in the literature l technical solution, the required storage space is the traditional VA, W / M times.

同时,基于VA的扩展,Nambirajian Seshadri和Carl-Erik W.Sundberg在美国电子电机工程师学会通信学报,1994年,2月,3月,4月,第42巻,313-323 页的题目为"列表维特比译码算法以及应用,,文献(Nambirajan Seshadri and Carl-Erik W, Sundberg, List Viterbi Decoding Algorithms with Applications, IEEE Trans, on Comm., Vol. 42, No. 2/3/4, Feb,/March/April 1994, 313-323.)中提出了一种列表维特比算法(一般也称为次优路径算法(ListVA ,简称LVA))。图1为该方法的译码结构示意图,如图示,LVA在译码过程中,根据差错校验的结果从L个全局优选路径中,选择最有可能正确的路径。由于LVA相对VA 增加了搜索过程中(Ll)个优选路径,提高了找到正确译码的几率,被广泛应用于包括巻积码和校验码的译码系统中。 At the same time, based on the expansion of the VA, Nambirajian Seshadri, and Carl-Erik W.Sundberg in the United States Institute of Electrical and Electronics Engineers Communications, 1994, February, March, April, Volume 42, pages 313-323 title is "List and applying Viterbi algorithm ,, literature (Nambirajan Seshadri and Carl-Erik W, Sundberg, List Viterbi decoding algorithms with applications, IEEE Trans, on Comm., Vol. 42, No. 2/3/4, Feb, / march / April 1994, 313-323.) proposes a list Viterbi algorithm (also commonly referred to sub-optimal path algorithm (ListVA, referred LVA)). Figure 1 a schematic view of a method for decoding, as illustrated , LVA during decoding, according to the error check result from the L global preferred path, select the most likely correct path. Since VA is relatively increased LVA search process (Ll) preferred path, to find the right improved decoding probability, are widely used, including convolving code and check code decoding system.

并行次优路径算法(Parallel LVA,简称PLVA ),是LVA的一种基本实现方式。 Parallel sub-optimal path algorithm (Parallel LVA, referred PLVA), it is a basic implementation of the LVA. 图2所示为PLVA的方法示意图,如图示,PLVA的基本实现方式是: 根据VA,通过一次网格搜索,同时得到L个全局优选路径。 Figure 2 shows a schematic diagram of a method PLVA, as illustrated, is substantially PLVA implementations are: According to VA, by a grid search, at the same time to give the L Global preferred path. 由于对于网格图中每个时刻的每个状态都找到进入该状态的L个局部优选路径,PLVA在每个时刻的每个状态,计算2个分支度量和2L个路径度量,最终一次得到所需要的L个全局优选路径。 Since for each state in the trellis diagram at each time can be found to this state is the L paths local Preferably, PLVA each state at each time point, and calculates two branch metrics 2L path metrics, a final obtain It requires L global preferred path. 特别的当L=l时,就是VA。 When L = l especially when that VA.

但是,PLVA在每个时刻的每个状态均计算2个分支度量、以及2L个路径度量,同时,需要L个IS^N的矩阵分别存储每个时刻的每状态点对应的L 个优选路径,其中M为巻积编码块长度,N为状态个数,L为预设的获取的全局优选路径的路径总数。 However, PLVA are calculated at each state of each time two branch metrics and path metrics 2L, simultaneously, requires the L IS ^ N matrices are stored for each state of the L paths preferably each time point corresponding, where M is the block length encoding convolving, N is the number of states, L is the total number of paths preset overall acquired preferred path. 与VA相比,该方法需要额外计算2 (Ll)个路径度量,额外占用(Ll)个MtN矩阵存储优选路径。 Compared with the VA, this method requires additional computational 2 (Ll) path metrics, take up additional (Ll) a matrix storage MtN preferred path. PLVA的计算量以及存储空间大约是VA算法的L倍,基本与L呈增长的关系。 PLVA amount of calculation and a storage space of about L times the VA algorithm, with a substantially L shape growth relationship. 因此,PLVA虽然提高了找到正确译码的概率,但是由于其需要占用巨大的存储空间,大大限制了 Therefore, PLVA While increasing the probability of finding the correct decoding, but because of its need to occupy huge storage space, greatly limits

6PLVA的实际应用。 The practical application of 6PLVA. 发明内容 SUMMARY

本发明要解决的技术问题是提供一种巻积码译码方法,解决现有方法中PLVA需要占用过大存储空间的问题。 The present invention is to solve the technical problem of providing a method of convolving code decoding, in the conventional method to solve the problem of excessive PLVA need to occupy memory space.

为解决上述技术问题,本发明的目的是通过以下技术方案实现的: 一种巻积码译码方法,包括以下步骤: To solve the above problems, an object of the present invention are achieved by the following technical solutions: A method for convolving code decoding, comprising the steps of:

A、 预设滑动窗口,其中所述的滑动窗口由判决输出部分以及辅助部分组成,将巻积码帧分段在所述的各滑动窗口中,使所述的各滑动窗口的判决输出部分的长度的总和等于巻积码帧长,预设滑动窗口标识iw,以及需要查找的全局优选路径的总数L,初始化iw等于0; A, a predetermined sliding window, wherein the sliding window output from the decision section and an auxiliary part, convolving the code segment in the frame of the sliding window, the sliding window so that the decisions of said output portion the sum of the code length is equal to convolving the frame length, a predetermined sliding window identifier iw, and a global preferred path needs to find the total number of L, initialized iw equal to 0;

B、 使iw累加l,获取第iw滑动窗口内,每时刻的每状态点的状态与每时刻的每状态点相对于巻积码开始时刻的路径度量,根据所述的路径度量选取L个优选路径度量,并存储所述的L个优选路径度量分别对应的优选路径,以笫iw滑动窗口的最末端的时刻的一状态点为起始点开始回溯,在每时刻,根据回溯状态点的路径值,按照最优路径回溯,回溯到第iw滑动窗口的最开始端的时刻,获取第iw滑动窗口的判决输出部分的局部最优路径,并且以所述的局部最优路径在判决输出部分最末端的时刻的状态点为回溯起始点,根据回溯的状态点的优选路径值回溯,获取除局部最优路径外的其它局部优选路径, 执行步骤C; B, enable accumulating L iw, iw obtain the first sliding window, the state of each state for each time point and each time point for each state path metric with respect to the start time code convolving, preferably the L metrics based on the path selected according to path metric, and storing the path metrics corresponding to the L are preferably of a preferred path to a state at the point in time of the endmost Zi iw sliding window back to the start point, at every moment, according to the state of the path traceback point value according to the optimal path back, back to the beginning end of the first time iw sliding window, sliding window iw acquiring first partial optimal path decision output portion and the partial optimal path in the decision output section endmost state point in time back to the starting point, the value of the path traceback according to the preferred status point back acquires preferred path in addition to other local partial optimal path, step C;

C、 如果iw小于滑动窗口的总数目,返回步骤B,否则,执行步骤D; C, iw is less than if the total number of the sliding window, return to step B, and otherwise, step D;

D、 根据各滑动窗口所获取的所有局部优选路径,获取所有的局部优选路径以及各全局优选路径的译码,并对所获取的全局优选路径输出的译码数据进行差错校验,根据校验结果,输出译码结果,其中所述的全局优选路径包括全局最优路径,以及除全局最优路径外的其它全局优选路径。 D, The preferred path for each of all the local sliding window acquired, obtaining all the local decoding paths and preferably each global preferred path, and the acquired global decoded data outputted preferred path error checking, The check As a result, the output of the decoding result, wherein the global path preferably comprises a global optimum path, and in addition to other global preferred path global optimal path.

本发明所述的方法,步骤A中所述的滑动窗口通过如下方式设置:设置所述的判决输出部分的长度,以及辅助部分的长度,其中辅助部分的长度具体是,根据所迷的巻积码的译码深度设置。 The method according to the present invention, in said step A sliding window provided by: setting a length of said decision output portion, and a length of the auxiliary portion, wherein the length of the auxiliary section in particular, according to the fans convolving code decoding depth setting.

本发明所迷的方法,步骤B所述的根据所述的路径度量选取L个优选路径度量具体是:如果巻积码译码器输入是硬判决,则选择较小的L个的路径度量作为所述的优选路径度量;如果所述的巻积码译码器输入是软判决,则选择较大的L个路径度量,作为所述的优选路径度量。 The method of the present invention, the fan according to the said step B is selected path metric the path metric particularly preferred L is: If the code decoder convolving the input hard decision, selecting a smaller path metric as the L preferably said path metric; convolving if said input code decoder is a soft decision, the larger L is selected path metrics, as the preferred path metric.

本发明所述的方法,所述的L个优选路径度量分别对应的优选路径具体用一具有L个比特单元的数值表示,其中各比特单元的值分别表示各优选路径。 The method according to the present invention, the preferred L of path metrics corresponding to each preferred path with a specific numerical value unit having L bits, where each bit value of each cell represent the preferred path.

优选的,在所述步骤B中,在所述的获取第iw滑动窗口的判决输出部分的局部最优路径过程中,如果所述的巻积码译码器输入是硬判决,则选取第iw 滑动窗口的最末端的时刻的路径度量最小的状态点为起始点开始回溯;如果所 Preferably, in the step B, the process of acquiring the local optimal path decision output portion of the sliding window iw, if said input code decoder convolving a hard decision, the selection of iw path timing endmost sliding window state metric minimum point is the starting point back; if

述的巻积码译码器输入是软判决,则选取第iw滑动窗口的最末端的时刻的路径度量最大的状态点为起始点开始回溯。 Convolving said input code decoder is a soft decision, the path timing of the endmost iw sliding window is selected maximum state metric point is the starting point back.

本发明所述的方法,步骤B中所述的根据回溯的状态点的优选路径值回溯,具体包括以下步骤: The method according to the present invention, the preferred state of the path traceback point value in step B in the back, includes the following steps:

Bl、根据每时刻的每状态点对应的各优选路径,预设一优选路径回溯表, 所述的优选路径回溯表,具体反映各优选路径与所述的各优选路径值的对应关系。 BL, in accordance with various preferred path for each state for each time point corresponding to a predetermined preferred path back table, the preferred path back table, particularly preferred path to reflect the relationship between each value of the respective paths is preferable.

B2、在回溯的每时刻,获取当前回溯所在状态点对应的各优选路径,并对照所述的优选路径回溯表,根据所述的优选路径回溯表,回溯获取所述的局部优选路径。 B2, every time the backtracking, backtracking acquires each preferred path where the current state of the corresponding point, and the control table according to the preferred path back, according to the preferred path back table, the preferred path back to the local acquisition.

优选的,所述步骤B2中,如果在回溯到第iw滑动窗口的最开始端的时刻之前的每时刻的状态点,所述状态点对应的优于当前回溯的优选路径的其他所有优选路径对应的路径值与当前回溯路径的路径值均相同,则回溯到第iw 滑动窗口的最开始端的时刻,停止回溯。 Preferably, the step B2, if the status of each point in time before the first iw back to the beginning of the sliding window in time of the end, preferably all of the other paths than the currently preferred path back to the state corresponding to the point corresponding to value of the current path value path are the same path back, then back to the beginning end of the first time iw sliding window, back stop.

本发明所述的方法,在步骤B2中进一步包括以下步骤: B21、 在回溯的每时刻,获取当前回溯所在状态点对应的各优选路径值后,比较各优选路径值中的当前需要回溯获取的局部优选路径对应的优选路径值以及最优路径值是否相同,在第一次不相同后,在回溯的下一时刻以及以后的回溯的每时刻,获取回溯的状态点的状态,并比较所述状态点的状态是否与局 The method of the present invention, further comprising the following steps in Step B2: B21, each time backtracking, obtaining respective values ​​of the current preferred path back state corresponding to the point where the current preferred path comparing the value of the acquired required backtracking preferred path corresponding to the local value of the preferred path and the value is the same as the optimal path, after the first is not the same, back in time and the next time after each of the backtracking, backtracking state acquisition state points, and comparing the whether the state status points and Bureau

部最优路径在所述状态点所在时刻的状态相同,如果不同,继续回溯;否则, Time optimal path portion in a state where the state of the same point, if different, continue back; otherwise,

停止回溯, <呆存当前回溯的局部优选路径。 Back stop, <stay locally stored current preferred path backtracking.

优选的,步骤B21所述的保存当前回溯的局部优选路径后,进一步包括: 舍弃之前保存的与当前回溯的优选路径的路径号相同的其他局部优选路径。 Preferably after topical preferred path, the stored current backtracking step B21, further comprising: discarding the same previously saved current path back to the preferred number of other local path preferred path.

优选的,在所述步骤B中进一步包括以下步骤: Preferably, further comprising the step of the step B:

B3、如果iw大于2,则释放滑动窗口标识小于(iw-l)的所有的所有其他滑动窗口内的每时刻的每状态点的状态、路径度量,以及路径的存储空间。 State B3, if iw greater than 2, less than the release of the sliding window identifier (iw-l) for each of all the state at the point in time each of all the other sliding window, path metric and path memory.

本发明所述的方法,步骤D中所述的全局最优路径的获取具体是:根据所获取的各滑动窗口的局部优选路径,获取全局最优路径。 The method according to the present invention, the step D for specific global optimal path is: the preferred path of each partial sliding window acquired, obtaining the global optimal path.

本发明所述的方法,步骤D中所述的除全局最优路径之外的其它全局优选路径的获取具体是:根据保存的各局部优选路径,以及各局部最优路径,获取所述的全局优选路径。 The method according to the present invention, the step D, the other global preferred path except for specific global optimal path is: The preferred paths of each local storage, and each local optimal path, the acquired global preferred path.

本发明所述的方法,步骤D中对所获取的全局优选路径输出的译码数据进行差错校验,根据校验结果,输出译码结果,具体包括以下步骤: The method according to the present invention, preferably in step D path acquired global output decoded data for error check, according to the check result, outputs the decoding result includes the following steps:

Dl、预设一实#:变量/,初始化/等于0; DL, a preset real #: variable /, initialization / equal to 0;

D2、对全局第/优路径输出的译码数据进行差错校验,如果校验正确,执行步骤D4,否则,进一步判定/是否小于所预设的需查找的全局优选路径的总数,如果小于,执行D3;否则,执行步骤D4; D2, the decoded data Global / shortest paths outputted error check, if the check is successful, step D4, otherwise, it is further determined / preset by whether the total number of required lookup global preferred path is less than, is less than, execution D3; otherwise, step D4;

D3、使/累加1,返回D2; D3, so / incremented by 1, D2 of return;

D4、输出译码结果,结束译码 D4, the output of the decoding result, the decoding end

由以上技术方案可以看出,本发明相对于现有技术,具有以下的优点: As can be seen from the above technical solutions, the present invention relative to the prior art, has the following advantages:

首先,由于在传统的PLVA技术中引入了滑动窗口的技术,将巻积码帧分段在个滑动窗口中,在各滑动窗口内,获取局部的优选路径,再根据各滑动窗口内的局部优选路径寄存器值,获取全局的优选路径。 First, the introduction of a sliding window technique in the conventional art PLVA, convolving the code segment in a frame in the sliding window, in each sliding window, acquiring local preferred path, then preferably partially within each sliding window according to the path register value, obtaining overall preferred path. 对于帧长为M,状态 For frame length is M, the state

数为N的巻积码,需要查找的优选路径的总数是L,使用传统的PLVA,需要占用的存储空间分别是:L*N*M个状态存储空间,以及2L*N*M个度量存储空间;而使用本发明方法,设滑动窗口的长度为W,需要占用的存储空间分别是2L*N*W个状态存储空间,以及2L*N*W个度量存储空间。 Number N of convolving code, the total number of preferred paths need to find is L, using conventional PLVA, requires storage space are: L * N * M states storage space, and 2L * N * M metrics storage space; method of the present invention, provided is a sliding window of length W, required storage space are 2L * N * W states storage space, and 2L * N * W metrics storage. 显然,W<M, 因此本发明相对现有技术而言,本方法既实现了PLVA巻积码译码的优点,又能减少译码所需要的存储空间,并且,由于滑窗技术的引入,使得译码需要占用的存储空间基本不受巻积码长度的约束,实际应用性更加强。 Obviously, W <M, therefore the present invention relative to the prior art, the present method not only achieve the advantages of convolving Decoding PLVA, but also reduce the storage space required for decoding, and, due to the introduction of the sliding window technique, such that the required storage space occupied by the decoding is not substantially bound by convolving code length, practical application strengthened. 特别的,当巻积码帧长较长时,应用本发明方法的效果更加显著。 In particular, when a long frame length codes convolving, application of the method according to the present invention, the effect is more significant.

其次,本发明还^:出了一种如何在滑动窗口内查找除局部最优路径外的其他局部优选路径的方法。 Secondly, the present invention further ^: how to find out a preferred path in addition to other local partial optimal path in the sliding window method. 本方法相对于文献1中揭露的如何在滑动窗口中查找局部最优路径的方法,既在技术上对滑动窗口技术的应用方面进行了更进一步的扩展和补充,更在实际应用方面,提出了更加适合实际应用的解决方案。 The method of the present process with respect to how to find the optimal path in the local sliding window disclosed in Document 1, both on the application of the sliding window is further extended and supplemented in the art, in more practical terms, the proposed more suitable solutions for practical applications.

附图说明 BRIEF DESCRIPTION

图1是LVA译码结构示意图; 1 is a schematic structural LVA decoding;

图2是PLVA的方法示意图; FIG 2 is a schematic diagram of a method PLVA;

图3是本发明方法流程示意图; Figure 3 is a schematic flowchart of a method of the present invention;

图4是本发明的一种滑动窗口示意图; FIG 4 is a schematic view of a sliding window of the present invention;

图5是本发明的另一种滑动窗口示意图; FIG 5 is a schematic view of another sliding window according to the present invention;

图6是每时刻的每状态点对应的的路径寄存器值示意图; FIG 6 is a state for each time point for each of the register values ​​corresponding to a schematic view of the path;

图7是乙=4时的优选路径回溯表; FIG 7 is a preferred path when B = 4 back table;

图8是滑动窗口对次优路径选择影响示意图; FIG 8 is a sliding window to select suboptimal paths Effect schematic;

图9是实施例1方法的流程示意图; FIG 9 is a flow diagram of one embodiment of a method embodiment;

图IO是在第iw窗口回溯局部第2优路径的流程示意图。 FIG IO iw is a flow diagram of a window in the back of the local second-shortest paths. 具体实施方式本发明的核心思想是,把滑动窗口技术引入到PLVA中,具体通过以下的技术方案实现:首先,预设滑动窗口,其中所述的滑动窗口由判决输出部分以 DETAILED DESCRIPTION The core idea of ​​the present invention, the sliding window into PLVA, specifically achieved by the following technical solutions: first, a predetermined sliding window, wherein the sliding window by the output judgment section to

及辅助部分组成,将巻积码帧分段在所述的各滑动窗口中, -使所述的各滑动窗 And auxiliary components, convolving the code segment in the frame of the sliding window, - each of the sliding window

口的判决输出部分的长度的总和等于巻积码帧长,预设滑动窗口标识变量iw, 以及需要查找的全局优选路径的总数L,初始化iwN);然后,使iw—w+l,获取第iw滑动窗口内的每时刻的每状态点的状态,以及每时刻的每状态点相对于所述的巻积码开始时刻的路径度量,根据所述的路径度量选取L个优选路径度量,并存储所述的L个优选路径度量分别对应的优选路径,以第iw滑动窗口的最末端的时刻的一状态点为起始点开始回溯,在每时刻,根据回溯状态点的路径值,按照最优路径回溯,回溯到第iw滑动窗口的最开始端的时刻,获取第iw滑动窗口的判决输出部分的局部最优路径,并且以所述的局部最优路径在判决输出部分最末端的时刻的状态点为回溯起始点,根据回溯的状态点的优选路径值回溯,获取其它局部优选路径,如果iw小于滑动窗口的总数目, 则继续使iw力 The sum of the length portion of the decision output port is equal to convolving the frame length code, the total number of preset variables L IW sliding window identifier, and a global need to find a preferred path, the initialization IWn); then the iw-w + l, acquisition of each state of each state at the point in time in the sliding window iw, and for each state for each time point with respect to the path metric convolving the start time code, and the path metric of the selected preferred path metric L, and stores the preferred L of path metrics respectively corresponding to the preferred path, in a state at the point in time of the first endmost iw sliding window back to the start point, at every time, based on the path status point value backtracking, the optimum path according to the back, back to the first sliding window iw beginning end time, obtain a partial optimal path decision output portion of the sliding window iw, and in the partial optimal path at the decision timing endmost portion of the state of the output points starting back, according to the preferred path back status point value back, preferably for additional topical route, if iw is less than the total number of the sliding window, the force continue to iw n 1,在下一滑动窗口内查找相应的局部最优路径以及除最优路径外的其他局部优选路径;如果iw不小于滑动窗口的总数目,根据各滑动窗口所获取的所有局部优选路径,获取所有的局部优选路径以及所述各全局优选路径的译码,并对所获取的全局优选路径输出的译码数据进行差错校验,根据校验结果,输出译码结果,其中所述的局部优选路径包括局部最优路径以及次优路径。 n 1, find the corresponding partial optimal path and other path preferably partial optimal path in addition to the next sliding window; iw if the total number is not less than the sliding window, according to the preferred paths of all local sliding window acquired, acquires All partial paths and preferably each of said coded global preferred path, and the acquired global decoded data outputted preferred path error check, according to the check result, outputs the decoding result, wherein preferably the partial path includes a local optimal route and the suboptimal route.

为了让本领域的技术人员更好的理解本发明内容,下面结合附图以及实施例进行进一步的说明: In order for those skilled in the art a better understanding of the present invention, further described below with the accompanying drawings and embodiments in conjunction with:

如图3所示为本发明方法的流程示意图,如图示,本发明包括以下步骤: Schematic flow chart of the method of the present invention shown in FIG. 3, as illustrated, the present invention comprises the steps of:

步骤S301:设置滑动窗口,预设需要查找的全局优选路径总数L,以及设置窗口标识变量iw,并且初始化i『0。 Step S301: Set the sliding window, the default global need to find the total number of preferred path L, and setting the identification variable window IW, and i is initialized to "0.

图4是本发明的滑动窗口示意图,如图示,滑动窗口W由判决输出部分WR、以及辅助部分Wc组成。 FIG 4 is a schematic view of a sliding window of the present invention, as illustrated, the sliding window W output from the decision section WR, and an auxiliary portion Wc composition. 滑动窗口W的长度W等于判决输出部分WR长度R, 以及辅助部分Wc长度C之和,即W二R+C。 W is the length of the sliding window W is equal to the length of WR-decision output section R, and the auxiliary portion Wc and length C ratio, i.e. W = R + C. 首先设置判决输出部分WR,该滑动窗口为译码判决输出部分,该部分的长度R的大小具体根据实际的译码系统性能进行设置,如果译码系统性能较优 First decision output section disposed WR, the sliding window decoder decision output section, the size of the length R of the specific portion in accordance with actual performance of the coding system, coding system if better performance

时,可以设置较长的判决输出部分,即较大的R值;如果i奪码系统性能较差, 存储空间较小时,可以设置较小的R值,使判决输出部分较短。 When, the judgment may be set longer output portion, i.e., the larger the value of R; i wins if the code system performance is poor, less storage space, a smaller value of R may be provided, so that a shorter decision output section. 各滑动窗口的判决输出部分的和等于所需要i奪码的巻积码的长度。 Decision output portion of each of the sliding window and equal to the length of the code i wins convolving code required.

其次辅助部分Wc,辅助部分Wc的设置,主要是由于如果各滑动窗口均只有判决输出部分,而不设置辅助输出部分,那么根据最优路径的定义以及网格图的特性,各滑动窗口中的最优路径可能不是全局意义上的最优路径在该滑动窗口中的部分路径。 Second auxiliary portion Wc, Wc auxiliary section is provided, mainly due to the sliding window averaging only if the decision output section, without providing the auxiliary output section, and then the characteristics of the definition of the optimal path of the trellis diagram, each sliding window the best path may not be optimal path on the part of the path in a global sense of the sliding window. 例如,对于一长度为M的巻积码,假设在时刻(M-1 ) 时,所有状态点中状态点A的路径度量最小,设该最小路径度量为a,并且路径度量a对应的路径为路径1,状态点A在时刻M转移到状态点B,状态点A 到状态点B的分支度量为ab,根据网格图的特性,对于状态点B还存在另一汇入分支,设由状态点C汇入,状态点C的路径度量是c,并且状态点C到状态点B的分支度量为ac,如果aOab,并且使(c+ac)〈(a+ab),即在时刻M, 路径度量最小的路径,即全局最优路径,为状态点C汇入的路径,而不是状态点A汇入的路径。 For example, for convolving the code having a length of M, it is assumed at time (M-1), the path status of all points in the state of point A measure of the smallest, provided the minimum path metric is a, and the path metric a corresponding path path 1, the state of the point a to the state at time M transition point B, the state of point a to the point B of the branch metric is ab & state, according to the characteristics of the trellis diagram for the state of point B there is another branch of import, provided by the state import point C, the path metric state at the point C is c, and you state metric to a branch point C to point B the state ac, if AoAb, and the (c + ac) <(a + ab), i.e., at time M, minimum path metric path, i.e. the overall optimal path, the state point C imported path instead of the path to import state at the point a. 可见,从巻积码开始时刻到时刻(Ml)的最优路径并不是全局最优路径的局部,如果从时刻(M-1)为起始点,根据VA回溯最优路径时,回溯获取的局部路径为路径1,不能得到正确的最优路径。 Be seen, starting from time to time convolving code (of Ml) partial optimal path is not globally optimal path, if from the time (M-1) as the starting point, in accordance with the optimal path back when VA, acquiring local backtracking path is the path 1, you can not get the correct optimal path. 因此,针对这种情况,在各滑动窗口的判决输出部分末端设置了辅助部分。 Thus, for this situation, the decision output end of each sliding window portion is provided auxiliary portion. 根据VA的特性:当译码深度d足够大的时候,比如,d等于巻积码编码器的约束长度的4 到6倍时,所有的幸存路径会以很大的概率在d个时刻单位内合并。 The characteristics of the VA: when the decoding depth d is large enough, for example, d is equal to four to six times the constraint length code encoder convolving, all the units will be the survivor paths with great probability in time d merge. 因此,将辅助部分的长度C设置为足够大,使得所有的幸存路径会以很大的概率在时刻长度C内合并,在包括判决输出部分以及辅助部分的所有滑动窗口内,进行前向计算,并且在辅助部分的最末端的时刻为起始时刻,按照最优路径回溯。 Thus, the length of the auxiliary portion C is set to be sufficiently large, such that all of the survivor paths will merge with great probability the time length C, in the decision output section comprising the sliding window and all of the auxiliary section, performing forward calculations, and at the time of the endmost portion is the auxiliary starting time back in accordance with the optimal path. 根据辅助部分的长度C的设置,回溯将很大概率的在时判决滑动窗口内与真正的全局意义上的最优路径合并,保证在判决输出滑动窗口内回溯获取的路径是全局意义上的最优路径的部分。 Depending on the setting the length of the C auxiliary sections, the probability of a large retrospective at the time of judgment within the sliding window and the optimal path in the true sense of the global merger, to ensure that the decision output path within the sliding window on the back get is the most global sense portion of shortest paths.

根据以上对滑动窗口的设置方法,可以得知,各全局优选路径与全局最优路径相分离的长度很大概率的不超过一个滑动窗口的长度,亦即,如果全局优选路径在某滑动窗口内与全局最优路径相分离,那么很大概率的,必将在本滑动窗口内,或者,在与之相邻的另一滑动窗口内(具体是另一滑动窗口的辅助部分内)与全局最优路径相合并。 According to the above method is provided for the sliding window, that may be, preferably each of the global path and the phase separation of the global optimum path length is high probability does not exceed the length of a sliding window, i.e., if a sliding window within the preferred path in the global separate from the global optimal path, the great probability, will within the sliding window, or, in other sliding window adjacent thereto (in particular the auxiliary part of another sliding window) global most shortest paths phases were combined.

将巻积码帧分段在各滑动窗口中,使所述的各滑动窗口的判决输出部分的 Convolving the frame segments in the code in the sliding window, sliding window according to each of the decision output portion

长度的总和等于巻积码帧长,在巻积码上的各滑动窗口可以相同也可以不同, 具体根据译码方便以及巻积码的类型进行设置,对于通讯系统中通常的对整帧 The sum of the code length is equal to convolving the frame length, each of the sliding window in convolving the code may be the same or different, set according to the specific type of convenience and convolving code decoding, communication system for the entire frame generally

巻积码的译码,如图4所示,通常设定使除位于巻积码帧最末端的滑动窗口外的其他所有滑动窗口一样即各滑动窗口的R、 C参数一样,而对于最末端的滑动窗口,令该滑动窗口的判决输出部分的长度R等于余下的巻积码长度,其辅助部分的长度C等于巻积码末端的冗余码段长度(一般在巻积编码时,在编码结束后,在编码的末端多输入一个(Kl)比特的各比特值为"0"的冗余码段,用于确认编码结束)。 Convolving code decoding, as shown, is generally set so that in addition to all the other sliding window located in the endmost convolving code frame of a sliding window that is the same as R 4 each sliding window, C as parameters, and for the endmost when the sliding window, so that the length of the R portion of the decision output of the sliding window is equal to convolving the remaining code length, which is equal to the length C of auxiliary section convolving redundant code having a code length ends (usually convolving coding, coding after, in a multiple-input end of the coding (Kl) bits each bit is "0" redundant code segments, used to confirm the end of the encoding).

而如果需要对一帧巻积码帧的某段进行译码时,通常采用图5所示的滑动窗口设置,所有滑动窗口的辅助部分C参数一致,而对于除了位于所需要译码的巻积码段最末端的滑动窗口的其它滑动窗口,其判决输出部分R参数一样,而对于位于所需要"^奪码的巻积码段的最末端的滑动窗口的判决输出部分R 参数具体等于余下的巻积码部分的长度。 If the need for a certain code frame convolving decoding, usually sliding window arrangement shown in FIG. 5, same parameters for all auxiliary portion C of the sliding window, and for convolving located required in addition to the decoded other sliding window endmost sliding window code segment that decision output section R parameters, like, and for the positioned required "^ judgment endmost sliding window capture code convolving code segment output section R specific parameters equal to the remaining convolving length code portion.

步骤S302: iw=iw+l,获取第iw滑动窗口的局部最优路径以及除最优路径外的其他局部优选路径。 Step S302: iw = iw + l, obtain local optimal path of the sliding window and iw preferred path in addition to other local optimal path.

按照滑动窗口在巻积码帧的时刻从大到小的顺序,将对应的时刻最小的滑动窗口称为第一滑动窗口,其次第二滑动窗口,依次类推。 Sliding window in order convolving the time code frame descending, the corresponding minimum time sliding window called a first sliding window, sliding window followed by a second, and so on.

使iw加l,首先ifl,首先对第l滑动窗口: Plus l makes iw, first ifl, first the first sliding window l:

在第l滑动窗口内的网格图中,从滑动窗口的开始时刻(对于第一滑动窗口也即需要译码的巻积码的开始时刻),前向计算,对于每个时刻的每个状态存储该状态,并分别计算2个分支度量,并分別将前一时刻的状态点的路径度量加上分支度量,获取当前状态点的路径度量。 In the trellis diagram of l in the sliding window, the sliding window from the start timing (i.e., the first sliding window decoded convolving code required start time), prior to the computing, for each state at each time storing the status, and two branch metrics are calculated, respectively, and the path state before a time point plus the metric of the branch metric obtaining path metrics of the current state point. 如杲所需要查找的全局优选路径的总数为L,则在每一状态点计算2L个路径度量,并保存该2L个路径度量中的较小的L个路径度量,以及该所保存的L个路径度量对应的路径。 The preferred path is the total number of global lookup is required Gao L, 2L is calculated path metrics at each state point, and to save a smaller path metric for the L 2L of the paths, and the L stored path metric corresponding to the path. 按照最优、 次优、第三优一直到第L优的顺序,对应存储。 According to the optimal, suboptimal, preferably up to the third order of the optimal L, corresponding to storage.

图6是每时刻的每状态点对应的的路径寄存器值示意图,该路径寄存器值具体是存储在路径寄存器中的该状态点对应的各优选路径,假设所需要查找的全局优选路径的路径总数为L,如图示,在每个时刻的每个状态点均有相应的路径寄存器来存储该L个优选路径,在路径寄存器中保存L个比特值,每比特值分別表示到达当前状态点的的迁移时刻的各优选路径,比如,第1比特表示最优路径的路径,第2比特表示次优路径的路径,第3比特表示第3优路径, 依次,第L比特表示第L优路径的路径。 FIG 6 is a path register diagram value for each state at a point corresponding to each time of the path register values ​​particular to the preferred path of the state point is stored in a path register corresponding to the total number of paths global preferred path hypotheses need to look for L, as illustrated, each state at each time point of the path has a corresponding L register to store the preferred path, the value of L bits stored in the register paths, each of the bit values ​​represent the current state reaches point preferably the migration time of each path, for example, the first bit represents the path of optimal path, the second bit represents a sub-optimal path, the path, the third bit represents the optimal path 3, sequentially, the L-bit representation of the path of shortest paths L . 每比特的值可以是"0"或者"1", 比特值具体表示到达该状态点的前一时刻的路径,根据移动通讯中网格图的特性,对于每一时刻的每状态点,均有2个分支汇入,也有2个分支汇出,因此, 可以使用比特值"0"和"1"分别当前状态具体是由那个分支转移过来的,即当前状态是由前一时刻的两个状态点的那个状态点转移过来的。 Value of each bit may be "0" or "1", the bit represents a particular value of the path the state of the previous time point is reached, according to the characteristics of the mobile communication trellis diagram, each state for each time point, are 2 branches import, export has two branches, thus, can use the bit value "0" and "1" are transferred from the current state of that particular branch, i.e., the current state is the state from the previous two time the state point of the point transfer over. 比如,如果设置比特值为"0"时,表示当前状态是由前一时刻的某状态点转移过来的,而比特值为'T,时,则表示当前状态是由前一时刻的另一状态转移过来的。 For example, if the bit value is set to "0", it indicates the current state is a state transition from the previous time point over, and the bit value 'T, when it indicates that the current state is another state from the previous time transfer over.

前向计算到滑动窗口的最末端的时刻后,当前滑动窗口内的前向计算完毕。 Before completion of the calculation to calculate a time after the end of the sliding window of the most before the current sliding window. 此时保存了当前窗口内的所有状态点的状态、所有状态点的L个优选路径值(即路径寄存器值)。 At this time, saves the state of the current status of all points within the window, L preferred values ​​of all the state point of the path (i.e. the path register value).

回溯查找局部最优路径:选取滑动窗口中的辅助部分的最末端的时刻,任选一个状态点为起始点,根据每状态点的路径寄存器值,按照最优路径回溯, 一直回溯到该滑动窗口的开始时刻,即回溯到判决输出滑动窗口的开始时刻, 其中在判决输出部分回溯的路径便是所需要的局部最优路径,保存判决输出部分时刻范围的最优路径。 Find local optimal path backtracking: Select auxiliary time endmost portion in the sliding window, optionally starting point of a state, according to a path state register values ​​for each point, in accordance with the optimal path back until the back of the sliding window the start time, i.e. back to the start timing decision output sliding window, wherein the optimal path in the local path decision output portion of the back is needed, the decision to save the output portion of the optimum path timing range.

查找除局部最优路径之外的其他局部优选路径:选取所获取的局部最优路径在滑动窗口的判决输出部分的最末端的时刻的状态点为起始点,在回溯的每时刻,根据回溯所在状态点的路径寄存器值,回溯查找所需要的局部优选路径,如果回溯在当前窗口内一直与局部最优路径相合并,则回溯到当前窗口最开始端的时刻为止,4亭止回溯;否则, 一直回溯到与局部优选路径在合并为止。 Find the best path in addition to other local topical route is preferred: partial optimal path selection status acquired at the time point decision output endmost portion as a starting point of the sliding window, each time backtracking, backtracking is located according to path register value status point, back to find the local preferred path required, if the back has been merged with the local optimal path in the current window, then back up to the time of the current window beginning end, four booths stop backtracking; otherwise, has been preferably back to the local path until the merge.

为了方便回溯查找所需要的L个局部优选路径,可以根据如图6所示的路径寄存器值,建立一优选路径回溯表,如图7所示为L=4时的优选路径回溯表,如图示,表中的"1"表示:在回溯的下一时刻,按照最优路径回溯;"2" 表示:在回溯的下一时刻,按照次优路径回溯;"3"表示:在回溯的下一时刻, 按照第3优路径回溯;"4"表示在回溯的下一时刻,4姿照第4优路径回溯。 In order to facilitate the L back to find the required local preferred path, the path can register values ​​as shown in FIG. 6, establishes a preferred path back table, as shown for the preferred path L = 4:00 to back as shown in Table 7, FIG. shown, the table, "1" indicates that: the back at the next moment, according to the optimal path back; "2": at the next moment backtracking, backtracking according suboptimal paths; "3": under the back a time, according to the third preferred path back; "4" at the next moment backtracking, 4 according to a fourth preferred posture path backtracking. 因此该表描述了,查找各局部优选路径时,在回溯查找的每一时刻,根据该时刻回溯所在状态点的路径寄存器值,确定在回溯的下一时刻,具体应该按照所述的下一时刻的路径寄存器值中的哪个路径回溯。 Thus the table describes, when each local lookup preferred path, looking back at each moment, based on the path timing register value of the state where the back point, is determined at the next moment back, in particular should be based on the next time the path which path register value back. 例如: E.g:

如果,当前回溯查找的路径是最优路径,对照图7所示的表格,可见无论回溯的每个状态点对应的路径寄存器值是什么,在回溯的每一时刻的下一时刻均按照最优路径回溯。 If the current path back to find the optimal path control table shown in FIG. 7, showing each state regardless of the path traceback register value corresponding to the point of what is, at every moment of the next timing according to the average optimal backtracking path backtracking. 回溯到当前窗口开始端的时刻时,获取到当前行窗口的局部最优路径。 When the current time window back to the start end, to obtain local optimal path to the current row of the window.

如果,当前回溯查找的路径是次优路径(即局部第2优路径),根据图7 所示的优选路径回溯表,如果当前回溯所在状态点的路径值中次优路径与最优路径相同,即对照图7所示表格中,路径值为:"0000"、 "0001"、 "0010"、 "0011"、 "1111"、 "1110"、 "1101"、 "1100",则在回溯的下一时刻均根据所述的下一时刻的状态点的路径寄存器值,按照次优路径回溯;如果当前回溯所在状态点的路径值中次优路径与最优路径不相同,即对照图7所示表格中,路径值为: If the current path is suboptimal look back path (i.e., the second local optimal path), according to the preferred path back table shown in FIG. 7, if the current state at the point where the path back value suboptimal path same optimum path, i.e., a control table shown in FIG. 7, the path is: "0000", "0001", "0010", "0011", "1111", "1110", "1101", "1100", then the back of the the average time a path state register value the next time point, in accordance with sub-optimal path backtracking; if the current state at the point where the back path value suboptimal path and the optimal path is not the same, i.e., the control shown in FIG. 7 in the table, the path is:

"iooo"、 "ioor,、 "ioio"、 "ion"、 "oiii"、 "oiio"、 "oior、 "oioo",则 "Iooo", "ioor ,," ioio "," ion "," oiii "," oiio "," oior, "oioo", then

在回溯的下一时刻,均按照所述下一时刻的路径寄存器值中的最优路径回溯。 Back at the next moment, the average value of the register back follow the path of a next optimum path timing.

当前回溯查找的路径是第3优路径,在回溯的每一时刻,获取当前回溯所在状态点的路径寄存器值,根据该路径寄存器值,对照图7所示的表格,如果路径值的最优路径值、次优路径值、第3优路径值相等,即路径值为"0000"、 "0001"、 "1110"、 "1111",则在回溯的下一时刻,按照所述下一时刻的路径寄存器值中的第3优路径回溯;如果路径值的最优;洛径值、次优路径值相同,而第3优路径值与最优路径值、次优路径值不同,即if各径值为:"0010"、 "0011"、 "1101"、 "1100",则在 Backtracking to find the current path is the first path preferably 3, at each moment backtracking, backtracking path of the current location acquired status point register value, the register value from the path, a control table shown in FIG. 7, the value of the optimal path if the path value, suboptimal path value, a third value equal to the optimal path, i.e. the path is "0000", "0001", "1110", "1111", then the back of the next timing according to the timing of a next path the third shortest paths register value back; if the optimum path value; Luo radius value, suboptimal paths of the same value, and the third value and the optimal path value optimal path, sub-optimal paths different value, i.e., if the value of each path as: "0010", "0011", "1101", "1100", the

回溯的下一时刻,4妄照所述下一时刻的路径寄存器值中的最优路径回溯; Next time backtracking, the optimum path as 4 jump the next time the path traceback register value;

如果路径值的最优路径值、次优路径值不同,而第3优路径值与最优路径值、次优路径值其中之一相同,即路径值为:"010"、 "0101"、 "0110"、 "0111"、 "1011"、 "1010"、 "1001"、 "1110",则在回溯的下一时刻,按照所述下一时刻的路径寄存器值中的次优路径回溯; If the value of the value of the path the optimal path, sub-optimal paths are different, and the third value and the optimal path optimal path value, the same value is suboptimal paths wherein one, i.e. the value of the path: "010", "0101", " 0110 "," 0111 "," 1011 "," 1010 "," 1001 "," 1110 ", then in the next time backtracking, backtracking follow the path of the next timing register value is suboptimal paths;

如果,当前回溯查找的路径是第4优路径,在回溯的每一时刻,获取当前回溯所在状态点的路径寄存器值,根据该路径寄存器值,对照图7所示的表格, 如果路径值的最优路径值、次优路径值、第3优路径值、第4优路径值均相等, 即路径值为"0000"、 "1111",则在回溯的下一时刻,按照所述下一时刻的路径寄存器值中的第4优路径回溯; If the current path back to find the optimal path 4, at each moment backtracking, backtracking path of the current location acquired status point register value, the register value from the path, a control table shown in FIG. 7, if the best path value optimal path value, the value of sub-optimal path, a third path of optimal value, the fourth value are equal shortest paths, i.e., path value of "0000", "1111", then the back of the next timing according to the timing of the next a fourth preferred route path traceback register values;

如果路径值的最优路径值、次优路径值、第3优路径值、第4优路径值只有三个相同,其中一个与其它的3个相异,即路径值为:"0010"、 "0100"、 "0111"、 "1101"、 "1011"、 "1110",则在回溯的下一时刻,按照所述下一时刻的路径寄存器值中的第3优路径回溯; If the value of the value of the path the optimal path, the path sub-optimal value, the third value optimal route, only three identical values ​​a fourth preferred path, the other one of three different, i.e. the path is: "0010", " 0100 "," 0111 "," 1101 "," 1011 "," 1110 ", then at the next moment the back, according to the time of a next optimal path 3 path traceback register value;

如果路径值的最优路径值、次优路径值、第3优路径值、第4优路径值只有两个相同,即路径值为:"0011"、 "0101"、 "0110"、 "1100"、 "1010"、 "1001", 则在回溯的下一时刻,4姿照所述下一时刻的路径寄存器值中的次优路径回溯; If the value of the value of the path the optimal path, the path sub-optimal value, the optimal path value 3, 4, only two same values ​​optimal path, i.e. the path is: "0011", "0101", "0110", "1100" , "1010", "1001", then in the next time backtracking, 4 according to the posture of the next time path register value back suboptimal paths;

如果路径值的最优路径值、次优路径值、第3优路径值相同,而第4优路径值与最优路径值、次优路径值、第3优路径值相异,即路径值为:"0001"、 "1110",则在回溯的下一时刻,按照所述下一时刻的路径寄存器值中的最优路径回溯; If the value of the value of the path the optimal path, the path sub-optimal value, preferably the same as the third path value, and the fourth shortest paths with the best path value value value suboptimal path, a third path preferably different value, i.e., the path is : "0001", "1110", then at the next moment back, in accordance with the path of the next time register value optimum path backtracking;

在回溯的每时刻,根据图7所示的优选路径回溯表,分别回溯获取个局部优选路径。 In each time backtracking, backtracking according to the preferred path table shown in FIG. 7, respectively, a local acquisition preferred path back. 值得说明的是,在回溯过程中,如果路径值的最优路径值与当前所回溯的局部优选路径对应的路径值第一次不相同时(比如,如果当前回溯查找的路径是局部次优路径,则路径值的最优路径值与次优路径值笫一次不同;如果当前回溯查找的路径是局部第3优路径,则路径值的最优路径值与第3优路径值第一次不同。)则在以后回溯的每时刻,获取回溯所在状态点的状态,将该状态与已知的局部优选路径在对应时刻的状态是否相同,如果不同,则一直 It should be noted that, in the course of backtracking, the optimum path if the value of the path with the current value of the preferred path back to the local path corresponding to the first value are not the same (for example, if the path back to find the current path is a partial suboptimal , the value of the optimal path and the path value a sub-optimal path different from the value Zi; if the path to find the optimal path back to the current value of the local optimal path 3, the path value and the third value differs from the first route optimization. ) at each time point after backtracking, backtracking state where the state acquisition point, the state of the known state corresponding to the local time at the preferred path is the same, if different, has been

根据图7所示的优选路径回溯表回溯, 一直回溯到,回溯所在状态点的状态与 The preferred path back table shown in FIG. 7 back until the back, back to the point where the state of the state

局部优选路径在对应时刻的状态点的状态相同位置,停止回溯。 Preferably partial path corresponding to the same position at the time the state status point, stopping back.

滑动窗口的位置会对除最优路径之外的其他优选路径的回溯产生影响,以下以次优路径为例进行说明,其他的优选路径与次优路径同理。 Back position of the sliding window will produce the optimal path in addition to other preferred path effects, the suboptimal paths as an example, other preferred path suboptimal paths empathy.

如图8示,当次优路径落在滑动窗口内时,如滑动窗口3,显然,次优路径的回溯在滑动窗口3内停止,不受滑动窗口的影响。 8 shows, when a sub-optimal path falls within the sliding window, sliding window 3, it is evident suboptimal paths back stops in the sliding window 3, the influence from sliding window. 但是。 but. 当滑动窗口落在次优路径与最优路径的中间部分时,如滑动窗口4的位置,则按照传统的滑动窗口的方法在当前滑动窗口内回溯,无法回溯出完整的次优路径,此时需要在滑动窗口4以外,利用前一滑动窗口,即滑动窗口3的路径信息,继续回溯, 直到与滑动窗口3的最优路径相合并为止,才能正确获得次优路径。 When the intermediate portion of the sliding window falls suboptimal path and the optimal path, such as the 4 position of the sliding window, the sliding window in accordance with a conventional method in the current sliding window backtracking, backtracking can not be a complete sub-optimal path, and the 4 other than the required sliding window, using the previous sliding window, i.e. a sliding path information window 3 back continues until merged with the best path 3 until the sliding window, in order to obtain the correct sub-optimal path.

在当前滑动窗口内,对所有的除了最优路径外的其他所有(Ll)个优选路径回溯,并存储所获取的各局部优选路径后,执行步骤S303,判定当前的iw是否小于滑动窗口总数,如果不小于,则执行步骤S304;否则,返回步骤S302,继续使iw加l,使iw:2。 In the current sliding window, all other all (Ll) in addition to the preferred path back to the optimal path, preferably after each local path and stores the acquired execution step S303, the determination current iw is less than the total number of the sliding window, If not less than, the step S304, is executed; otherwise, return to step S302, the iw continues to increase L, so iw: 2.

在第2滑动窗口内的前向计算与在第一滑动窗口内的计算基本相同,所不同的是:在本滑动窗口中,由于第1滑动窗口的判决输出部分的最末端的时刻为第二滑动窗口的判决输出端的开始时刻,第1滑动窗口的辅助部分与第二滑 Forward calculated in the second calculation of the sliding window in a first sliding window is substantially the same, except that: In the sliding window, the time since the most distal portion of the first decision output for a second sliding window start timing decision output of the sliding window, sliding window 1 auxiliary part of the second carriage

动窗口相重合。 Coincides movable window. 相重合部分的所有状态点的状态、累积度量以及路径值,在第一滑动窗口前向计算时已经获取,因此,在对第2滑动窗口进行计算时,可以不必进行重复计算,而从第1滑动窗口的最末端的时刻为前向计算的开始时刻,设该时刻为时刻i,那么,根据iw-l时获取的时刻i的所有状态点的路径度量,累计下一时刻的分支度量,获取下一时刻的路径度量。 All points coincides with the state of the state portion, and the accumulated path metric value, the calculation has been acquired prior to the first sliding window, and therefore, when the second sliding window calculation, calculation may not necessarily be repeated, and from the first time endmost forward sliding window is calculated start time, the time set for the time i, then, in accordance with the state of all the path points acquired when iw-l measure time i, the cumulative branch metric the next time, obtaining path next time metrics. 并且根据路径度量,保存L个优选路径的路径值。 And the path metric, the path value L stored preferred path.

前向计算完毕后,与获取第l滑动窗口内同法,获取第2滑动窗口的判决输出部分的局部最优路径。 After the calculation is finished before, and obtaining the same method the l-sliding window, acquire the local optimal decision output path to the second portion of the sliding window. 选取该局部最优路径在第2滑动窗口的判决输出部分的末端的时刻的状态点为回溯起始点,与在第1滑动窗口内获取除最优路径外的优选路径同法,在第2滑动窗口内回溯查找相应的优选路径。 The local optimal path selected at the time point of the end state of the second decision output portion back to the starting point of the sliding window, and access to the same method except preferred path optimal path in a first sliding window, the second slide preferred path back to find the appropriate window. 值得i兌明的是在回溯查找除最优路径外的其他优选路径时,如果所回溯的优选路径与最优路径在当前滑动窗口内保持重合,则回溯到本滑动窗口的判决输出部分的开始 I out is worth against the back when the preferred path to find another optimal path in addition, if the preferred path back to the optimal path remains in the current sliding window coincide, then back to the start of this decision output portion sliding window

端的时刻停止;否则一直回溯到与最优路径相合并为止,并存储当前获取的优 End of time to stop; otherwise, has been merged with the back to the optimal path so far, and excellent storage currently acquired

选路径在与最优路径相分离时刻段的路径寄存器值,而舍弃在前一滑动窗口回溯获取的与当前所获取的局部优选路径的路径号相同的局部优选路径(即如果 Selected from the preferred path and local path separated optimum path timing path segment register values, and a sliding window is the same as the previous discard the acquired back to the local path of the current is preferably acquired path number (i.e., if

在第2滑动窗口中获取的第2局部优选路径与该窗口内的局部最优路径,在地2滑动窗口内部在所有时刻重合,那么舍弃在第1滑动窗口内获取的局部次优路径,对于其他局部优选路径,同理处理)。 Preferably the second partial path acquired in the sliding window and the second local optimal path within the window, inside the sliding window coincide at all times, then discarded local suboptimal path acquired in the first sliding window, for other preferred topical route, empathy process). 如果所回溯的优选路径与最优路径在当前滑动窗口内相分离后,回溯到当前滑动窗口的最开始端的时刻,仍然未与当前滑动窗口内的最优路径合并,那么继续在前一滑动窗口内进行回溯, 直到回溯的优选路径与前一滑动窗口的局部优选路径合并为止。 If the preferred path back to the optimal path are separated in the current sliding window, back to the beginning of the current time of the end of the sliding window, and remains unconsolidated optimal path within the current sliding window, the sliding window continues the previous the backtracking, backtracking until the preferred path is the preferred path to the local sliding window before a merger so far. 在第2滑动窗口内,对所有的优选路径回溯完毕,并更新了所述的各优选路径的路径寄存器值后,执行步骤S303,判定当前的iw是否小于滑动窗口总数,如果不小于, 则执行步骤S304,否则,表示当前滑动窗口不是最末端的滑动窗口,返回步骤S302,继续使iw力。 After the second sliding window, back completed for all preferred path, and the updated register values ​​for each path according to the preferred path, performing step S303, the determination current iw is less than the total number of the sliding window, if not less, is executed step S304, otherwise, it represents the most current sliding window is not the end of the sliding window, return to step S302, continues to iw force. 1 ,使iw=3 。 1, so that iw = 3.

对第3滑动窗口以及后续的其他滑动窗口的最优路径以及其他优选个路径的获取方法与第2滑动窗口的同理。 Similarly to the optimum path of the third and subsequent sliding windows and other sliding window Other preferred method for obtaining a path with the second sliding window. 在此不作赘述。 Which I will not repeat them here.

值得说明的是,根据上面的描述,可以看出,在回溯查找除全局最优路径的其他优选路径时,存在回溯的优选路径不在当前滑动窗口内与当前滑动窗口的局部最优路径合并,而需要在前面的滑动窗口内继续进行回溯,回溯在前面的滑动窗口内与该滑动窗口内的的局部最优路径相合并的可能。 It should be noted that according to the above description, it can be seen, while in others inter backtracking to find the global optimum path is the preferred path, the presence of the preferred path back out of the current path of the current combined with local optimal sliding window in the sliding window, and need to continue back in front of the sliding window, in front of the back may be a partial optimal path of the sliding window in the sliding window phases combined. 再才艮据对滑动窗口的辅助部分Wc的长度C的设置,得知,如果回溯的优选个路径在当前滑动窗口内未与最优路径合并,在前一滑动窗口内必定与最优路径合并。 It was then Gen auxiliary section is provided on a sliding window of length C Wc, that, if the back is preferably combined with the paths are not optimal path in the current sliding window, within a sliding window must be combined with the previous best path . 因此, 如杲当前的滑动窗口为第iw滑动窗口,需要,并只需要存储第(iw-l)滑动窗口内的所有状态点的状态,所有状态点的L个径度量以及相对于L个路径度量的路径值,以便于第iw滑动窗口查找该滑动窗口内的除局部最优路径外的其他局部优选路径使用。 Thus, as Gao current sliding window of iw sliding window needs, and requires only the storage section (iw-l) the status of all the state point in the sliding window, the L path metrics for all status point and with respect to L paths path metric value in order to find other topical preferred path outside the local optimal path of the sliding window in addition to the first sliding window used iw.

步骤S303:判定当前iw是否小于滑动窗口的总数,如果不小于,则表示当滑动窗口是最末端的滑动窗口,执行步骤S304;否则返回步骤S302。 Step S303: determining the total number of the current iw is less than the sliding window, if not less than, then when the sliding window is a sliding window endmost, step S304,; otherwise, returns to step S302.

步骤S304:根据各滑动窗口所获取的所有局部优选路径,获取所有的局部优选路径以及所述各全局优选路径的译码,并对所获取的全局优选路径输出的译码数据进行差错校验,根据校验结果,输出译码结果。 Step S304: The preferred paths of all local sliding window acquired, obtaining all the local decoding said preferred path and preferably each of the global path and the acquired global decoded data outputted preferred path error check, The check result, the output decoding result.

显然,由于各相邻滑动窗口的判决输出部分首尾在同一时刻点相连接,因此各滑动窗口的局部优选路径,按照时刻顺序的组合Y更是全局最优路径,而对于除全局最优路径外的其他各全局优选路径,可以根据各优选路径与全局最优经相分离的时刻段的路径,结合全局最优路径,将各优选路径与全局最优经相分离的时刻段的路径,替换全局最优路径上的相应时刻段的路径,便可获取包括所有时刻的各全局优选路径。 Obviously, since the respective adjacent end-decision output section is connected to a sliding window at the same time point, and therefore the sliding window is preferably a partial path, in order of time in accordance with a combination of Y is globally optimal path, except for the outer global optimal path other global each preferred path, each path may preferably global optimum path timing phase-separation section, in conjunction with the global optimal path, the path preferably each global optimal timing via path separate segments, replace the global path corresponding time segments on the optimal path, can obtain all the preferred path including the global time.

对各全局优选路径的译码数据,进行差错校验,如果某全局优选路径的译码差错校验正确,则输出该正确的译码,结束译码;如果对所有路径输出的译码数据分别进行差错校验均不正确,则输出译码失败的译码结果,结束译码。 Preferably the decoded data for each global path, performs error checking, if a decoding error check correct global preferred path, then outputs the decoded correctly, the decoding end; if all path outputs decoded data, respectively error-checks are incorrect, the output fails to decode the decoding result, the decoding end.

现有的可以采用的差错校验的方法具体有:CRC校验、奇偶校验、重复码校验、恒比码校验等。 Conventional error check method may be employed are specific: CRC checksum, parity, parity code repeated, constant ratio of code checking. 具体是在信道编码时,在对信道传输数据进行编码时, 在数据中插入差错校验码,使校验码与数据一起发送到信道接收端。 In particular when the channel coding, when the channel encoding the transmission data, error check code inserted in the data, a parity check code transmitted together with the data channel receiver. 信道接收端对所接收到的信号进行解调,译码后,得到译码数据,根据对信道编码所采用不同的差错校验编码方式,在对译码数据进行差错校验时,采用与信道编码 After the receiving end of the channel received signal is demodulated, decoded to obtain decoded data, different error check code based on the channel coding mode used when the decoded data, error check, using the channel coding

时相应的差错校验方式。 When a corresponding error check mode. 比如,如果在信道编码时,釆用的是CRC校验码,那么对译码进行差错校验时,相应采用CRC校验方式。 For example, if the channel coding, preclude the use of the CRC checksum, then the time for decoding error check, using the corresponding CRC mode.

当译码失败时,通信系统可以才艮据该-泽码,执行自动请求重发、前向纠错、 或混合纠错等相应的差错控制策略。 When the decoding fails, a communication system may only according to the Gen - Ze code, automatic retransmission request, to the appropriate control strategy before the error correction, error correction, or mixed.

以下将通过具体实施例对本发明方法作进一步的分析:图9所示为本实施例的方法流程示意图,如图示,本方法包括以下步骤: The following specific embodiments will be made by the method of the present invention further analysis: flowchart of a method shown in the present embodiment, as illustrated in FIG. 9, the method comprising the steps of:

步骤S901:预设滑动窗口。 Step S901: pre-sliding window.

根据巻积码帧长,设置滑动窗口的判决输出部分WR的长度R,以及与下一滑动窗口相重合的部分Wc的长度C,其中各滑动窗口的判决输出部分R的总和,等于巻积码帧长。 The frame size code convolving provided sliding window WR decision output portion of the length R, and the next sliding window coincides with the length Wc of the portion C, where the decisions of the sum of the sliding window output part R, is equal to convolving code frame length.

如图4所示,对于除位于巻积码最末端的最后一滑动窗口外的其他的滑动窗口,判决输出部分WR的长度R大小均为RO,与下一滑动窗口相重合的部分Wc的长度C大小均为C(N64K,其中K为巻积编码器的约束长度。对于帧长为M的巻积码,可知滑动窗口的个数Nw为:Nw=「M/R0,;因此,对于位于巻积码最末段的滑动窗口:该滑动窗口的判决输出部分WR的长度为: RNw=M-(Nw*R0),而对于该滑动窗口的辅助部分Wc,由于该滑动窗口位于巻积码的最末段,可以将其设置为C^-K-1。 4, the other for a sliding window located convolving except the last code endmost sliding window, the length of both the RO decision output, coinciding with the next portion of the sliding window of length R of the size of the portion Wc of WR . C size are C (N64K, convolving where K is the constraint length of the encoder for the frame length of M convolving code, the number of sliding window Nw is understood: Nw = "M / R0 ,; Thus, for located convolving sliding window of the last segment code: decision output of the sliding window length of the portion of WR: RNw = M- (Nw * R0), and for the auxiliary section of the sliding window Wc, since the sliding window at convolving code the last segment may be set to C ^ -K-1.

步骤S卯2:初始化当前滑动窗口标识i『1,以及路径号/=1。 Step d S 2: Initialize the current sliding window identifier i "1, and a path number / = 1.

设滑动窗口标识iw的范围为04w〈二Nw,初始化,使当前滑动窗口为,位于巻积码最开始码段的第l个滑动窗口。 Range provided for the sliding window identifier iw 04w <Nw two, initializing the current sliding window, located in the beginning of the code convolving the l th sliding window code segments.

步骤S卯3:设置时刻变量j,并初始化jO,以及确定时刻j的范围。 D Step S 3: Set time variable j, and initializes jO, and determining a point in time j range.

^4居第iw滑动窗口,确定时刻j的范围,即时刻j的范围为,最小时刻为当前滑动窗口的判决输出部分的最开始端对应的时刻,最大时刻j为当前滑动窗口的辅助部分的最末端对应的时刻。 ^ 4 ranking iw sliding window determining the time j of the range, i.e., point in time j range, the minimum time the output timing of the beginning portion of the end corresponding to the decision of the current sliding window, the maximum time j is an auxiliary part of the current sliding window of corresponding endmost time.

步骤S904: j=j+l,获取时刻j的分支度量、路径度量,以及L个优选路径的路径寄存器值。 Step S904: j = j + l, j is the time acquisition branch metric, path metric, and the path value L registers preferred path.

时刻j加l,获取时刻j的所有状态点的状态,并计算分支度量,以及通过计 L j is incremented in time, acquires the status of all points in time j is the state, and calculating the branch metrics, and by taking into account

算汇入该状态点的前一状态点的路径度量与分支度量的和,获取路径度量,设需要查找的全局优选路径的总数为L,显然,在每状态点需要保存最小的L个路径度量,以及该L个路径度量对应的路径,由于每时刻的前一时刻具有L个路径度量,而每时刻的后一时刻具有2个分支度量进入该时刻,可见,对于每时刻均需要计算2L个路径度量,在所得到的2L个路径度量中选取较小的L个路径度量存储,并且保存该L个路径度量对应的路径值存储在路径寄存器中。 Total and import point of the previous state path metric state point and the branch metric, path metric acquisition, need to find a global set of preferred path L, obviously, needs to be saved at each state point of minimum path metrics L , and the L-th path metric corresponding to the path, since the previous time each time having L path metrics, and after each time a time with two branch metrics entering the time, be seen that for each time are required calculation the 2L path metric, select a smaller L 2L paths in the obtained path metrics stored in metric, and holds the path metric corresponding to the L value stored in the route path register. 如图6所示每时刻的每状态点对应的的路径寄存器值示意图,在该路径寄存器中, 保存有L个比特值,从第1比特、第2比特、、、第L比特的值,分别对应最优路径、次优路径、、、第L优路径的路径信息。 Path state value of each register diagram shown in Figure 6 for each time point corresponding to, in the path register, stores a value of L bits, a first bit, a second bit value from ,,, L bits, respectively, corresponding to the best path, sub-optimal path information of the path L ,,, optimal path. 对个每比特的值可能是"0"也可能是'T,。 Value for each bit may be "0" may be 'T ,. 由于每状态点均既有两可能分支汇入,也有两可能分支汇出, 即对于每状态点,只可能由前一时刻的两个状态转移过来的。 Since each state points are both two branches may import, there are two possible export branch, that is, for every status point, can only be transferred from the two states over the previous time. 因此,如果比特 Accordingly, if the bit

值为O,则表示前一状态点具体是由前一时刻一状态点转移过来的;如果该比特值为l,则表示前一状态点具体是由前一时刻另外一状态点转移过来的。 Is O, indicates the previous state transition point is a point from the particular previous state over a time; if the bit is l, the previous state represents a particular point in time is a point of transition from the former state over another. 可见,如果两比特的值相同,则表示两路径在前一时刻的状态点为同一点;否则, 两路径在前一时刻的状态点不同。 Seen, if the two-bit value, this indicates that the previous state of the two paths for the same time point a little; otherwise, the previous state of the two paths at a time point different.

步骤S卯5:判断j是否在步骤S903所确定的时刻范围内,如果在时刻范围内,返回步骤S904,否则,执行步骤S卯6。 Step S 5 d: j is determined within the range of time determined in step S903, if the time range, return to step S904, the otherwise, perform step S 6 d.

步骤S906:取第iw滑动窗口的在最末时刻的所有状态点中路径度量最小的状态点为回溯起始点,进行路径回溯,获取判决输出部分WR的局部最优路径。 Step S906: take the first sliding window iw path state at all points of time last minimum state metric back to starting point, a route back to acquire the local optimal decision output path portion of the WR.

选取第iw滑动窗口的在最末时刻的所有状态点中路径度量最小的状态点为回溯起始点,根据前向计算所获取各状态点的路径寄存器值,按照最优路径回溯,输出第iw滑动窗口的判决输出部分WR局部最优路径,并存储所述的局部最优路径的路径寄存器值。 Iw sliding window, select the first state at all points in the last time the minimum path metric state back to starting point, the path of each state of the register values ​​of the acquired points to calculate a front, back in accordance with the optimal path, the output of the sliding iw the decision window WR local optimal path output section, and a storage path of the register values ​​locally optimal path.

本步骤中,除了选取第iw滑动窗口的在最末时刻的所有状态点中路径度量最小的状态点为回溯起始点之外,还可以选取其他的任意状态点为起始点,所不同的是:选取第iw滑动窗口的在最末时刻的所有状态点中路径度量最小的状 In this step, in addition to select the first path iw sliding window the last time point in all states in the state metric minimum starting point outside the back, can also select any other state as a starting point, except that: iw sliding window, select the first path at all points the last time state metric minimum shape

态点为回溯起始点,有利于回溯尽快与滑动窗口内的局部最优路径重合,有利于尽快回溯到第iw滑动窗口的局部最优路上。 State starting point of backtracking, backtracking is conducive to the local optimal path as soon as possible within the sliding window coincide, facilitate local optimal path back to the first sliding window iw quickly.

步骤S卯7:根据局部最优路径回溯获取其他的(L-1)个局部优选路径, 并更新相应的路径寄存器值。 Step S 7 d: acquiring additional back (L-1) th local partial optimal path preferred path, the path and update the corresponding register value.

根据已获取的第iw滑动窗口的判决输出部分的局部最优路径,选取在最优路径在判决输出部分的最末端对应的时刻的状态点为回溯起始点,根据前向计算所存储的各状态点的路径寄存器值,对照图7所示的优选路径回溯表,进行回溯,为了增加可靠性,回溯的过程中,如果优选路径与最优路径分离后,则根据相应的路径寄存器值,回溯到优选路径与最优路径重合后,便停止回溯。 Partial optimal path decision output portion of the sliding window iw acquired, selecting an optimal path in a state at the time point of the endmost portion of the decision output corresponding to the starting point for the back, each state is calculated according to the stored forward waypoints register value, the preferred path back control table shown in FIG. 7, backtracking, in order to increase the reliability, during backtracking, the optimum path if the path is preferably separated, according to the register value corresponding path, back to after the optimal path coincides with the preferred path, back stops.

以下以次优路径为例,具体说明如何根据路径寄存器值回溯获取次优路 The following suboptimal path, for example, specify how backtracking obtain suboptimal paths according to the path register value

径: path:

首先,选取在最优路径上,WR的最末时刻的状态点为起始点,在起始点时刻,即WR的最末时刻,获取起始点的路径寄存器值,并根据路径寄存器值, 对照图7所示的优选i?各径回溯表,查找获知在回溯的下一时刻应该4安照那个优选路径回溯,并且在该起始点时刻,按照次优路径回溯。 First, in selecting the optimal path, the state point is the last time WR starting, at the starting time, i.e., the last time WR acquires the start point of the path register value, the register value according to the path and the control of FIG. 7 preferably shown i? backtracking of each path table lookup according to the known safety should preferably 4 path backtracking back at the next moment, the start point and time, in accordance with sub-optimal path backtracking. 在回溯的下一时刻, 即起始点的前一时刻,根据当前时刻的路径寄存器值,按照在起始时刻的路径寄存器值所获知的在当前时刻应该4艮据哪个优选3各径回溯的优选路径回溯,并在当前时刻,根据当前时刻的路径寄存器值,对照图7所示的优选路径回溯表, 获知在下一时,应该按照那个优选路径回溯,同理对于此后回溯的每一时刻。 Back at the next moment, i.e., the starting point of the previous time, based on the path of the current time value of the register, the register value according to the path of the known starting time should preferably 4 Gen 3, according to which each path backtracking preferably at the present time back path, and at the current time, the current time according to the path of the register value, the preferred path back control table shown in FIG. 7, the next time is known, according to the preferred path back should be, the same way for each time after backtracking. 如图10所示为在第iw窗口回溯局部第2优路径的流程示意图,如图示,具体如下: Figure 10 is a schematic flow chart of the back window iw of the second partial optimal path, as illustrated, as follows:

步骤S1001:初始化时刻一te。 Step S1001: initialization time a te.

初始化时刻t等于第iw滑动窗口的判决输出部分的最末端的时刻,设该时刻为时刻te; Time initialization time t equal to the extreme end of the sliding window iw decision output portion, provided that the time for time TE;

步骤S1002:获取时刻te回溯所在状态点对应的路径寄存器值;并对照优选路径回溯表获知在时刻(te-l)应按照那个优选路径回溯,获取时刻(te-l) 时刻应回溯的路径号;在时刻te,根据路径寄存器值,按照次优路径回溯。 Step S1002: acquiring time te back where a state corresponding to the point path register value; and control preferred path back sheet known to be back in accordance with the preferred path at time (te-l), acquisition time (te-l) time to be back path number ; at time TE, the register value according to the path, in accordance with sub-optimal path backtracking.

对照图7所示的优选路径回溯表,可见,在时刻te,如果路径寄存器值为: Preferred path back control table shown in FIG. 7, seen at time TE, the register value if the path:

"oooo"、 "ooor,、 "ooio"、 "ooir,、 "nir,、 "iiio"、 "iior、 "iioo",则 "Oooo", "ooor ,," ooio "," ooir ,, "nir ,," iiio "," iior, "iioo", then

在时刻(te-l),均根据时刻(te-l)的状态点的路径寄存器值,按照次优路径回溯;否则,在时刻(te-l),根据时刻(te-l)的状态点的路径寄存器值,按照最优路径回溯。 At time (te-l), are based on the path register value status point in time (te-l), in accordance with sub-optimal path backtracking; otherwise, at time (te-l), according to the state of the time (te-l) point path register values, in accordance with the optimal path back. 同时在时刻te,根据时刻te的路径寄存器值,按照次优路径回溯。 While the time te, the register value according to the path time te, suboptimal paths according to the back. 步骤S1003:根据路径寄存器值,判别路径寄存器值中次优路径值与最优路径值是否不同,如果不同执行步骤S1004,;否则,执行步骤S1004。 Step S1003: The register value path, the path determination sub-optimal path value register value and the optimum path value is different, if the different steps S1004 ,; otherwise, to step S1004. 步骤S1004:使时刻t等于t减l,执行步骤S1005。 Step S1004: time t equal to t so Save l, step S1005.

步骤S1005:获取时刻t回溯所在状态点对应的路径寄存器值,执行步骤S1006'。 Step S1005: obtaining a time t backtrack point where the path corresponding to the state of the register value, the step S1006 '.

步骤S1006:对照优选路径回溯表获知在时刻(tl )应按照那个优选路径回溯,获取时刻(tl)时刻应回溯的路径号,并且在时刻t,根据路径寄存器值,按照在时刻(t+l)获取的在时刻t按照的优选路径回溯,并返回步骤S1007。 Step S1006: a control preferred path back sheet known to be back in accordance with the preferred path at time (TL), acquired track number time (TL) in time to be back, and at time t, based on the path register value according to (t + l at time ) acquired at time t according to the preferred path back, and returns to step S1007.

对照图7所示的优选路径回溯表,可见,在时刻t,如果路径寄存器值为: Preferred path back control table shown in FIG. 7, can be seen, at time t, the value of the register if a path:

"oooo"、 "ooor、 "ooio"、 "ooir、 "iiir、 "iiio"、 "iior、 "iioo",则 "Oooo", "ooor," ooio "," ooir, "iiir," iiio "," iior, "iioo", then

在时刻(tl),均根据时刻(tl)的状态点的路径寄存器值,按照次优路径回溯; 否则,在时刻(tl),根据时刻(tl)的状态点的路径寄存器值,按照最优路径回溯。 At time (TL), were based on the path timing register value (TL) of the status point, in accordance with sub-optimal path backtracking; otherwise, at time (TL), based on the path timing register value (TL) of the status point, according to the best path backtracking. 同时根据时刻(t+l)时,所获取的在时刻t应按照的具体优选路径回溯。 The same time at time (t + l), obtained in particular according to the preferred path should backtracking time t.

步骤S1007:判定时刻t是否大于第iw窗口的最开始端的时刻,如果大于, 则返回步骤S1003;否则,执行步骤SIOOIO。 Step S1007: determining whether the time t is greater than the time of the beginning end iw window, If so, returns to step S1003; otherwise, step SIOOIO.

步骤S1004,:使时刻t等于t减l,执行步骤S1005'。 Step S1004 ,: time t equal to t so Save l, step S1005 '.

步骤S1005,:获取时刻t回溯所在状态点对应的路径寄存器值,执行步骤S1006,。 Step S1005 ,: acquire the time point t corresponding to the state where the path back to the register value, the step S1006 ,.

步骤S1006,:对照优选路径回溯表获知在时刻(t-1 )应按照那个优选路径回溯,获取时刻(tl)时刻应回溯的路径号,并且在时刻t,根据路径寄存器值,按照在时刻(t+l)获取的在时刻t按照的优选路径回溯,执行步骤S1007。 Step S1006 ,: preferred path traceback control table to be informed according to the preferred path back in time (t-1), the path number obtaining time (TL) to be back in time, and at time t, based on the path register value according to the time ( t + l) acquired at time t according to the preferred path backtracking step S1007.

对照图7所示的优选路径回溯表,可见,在时刻t,如果路径寄存器值为: Preferred path back control table shown in FIG. 7, can be seen, at time t, the value of the register if a path:

"oooo"、 "ooor,、 "ooio,'、 "ooii"、 "iiir,、 "iiio"、 "iior、 "i雨,,,则 "Oooo", "ooor ,," ooio, ', "ooii", "iiir ,," iiio "," iior, "i ,,, the rain

在时刻(tl),均根据时刻(tl)的状态点的路径寄存器值,按照次优路径回溯; 否则,在时刻(tl),根据时刻(tl)的状态点的路径寄存器值,按照最优路径回溯。 At time (TL), were based on the path timing register value (TL) of the status point, in accordance with sub-optimal path backtracking; otherwise, at time (TL), based on the path timing register value (TL) of the status point, according to the best path backtracking. 同时根据时刻(t+l)时,所获取的在时刻t应按照的具体优选路径回溯。 The same time at time (t + l), obtained in particular according to the preferred path should backtracking time t. 步骤S1007,:存储局部次优路径路径。 Step S1007 ,: suboptimal paths local storage path.

步骤SI008,:判断时刻t回溯所在状态点的状态与局部优选路径在时刻t的状态点的状态是否相同,如果相同执行,执行步骤S1008;否则,返回步骤S1004'。 The step of determining the time t SI008 ,: backtrack point where the state of the local state preferred path is the same state at time t at state point, if the same are performed in step S1008; otherwise, the process returns to step S1004 '.

如果回溯到第iw窗的最开始端对应的时刻,回溯还没有与最优路径重合, 那么进入第(iw-l)窗,继续根据每时刻的路径寄存器值进行回溯,直到回溯 If back in time to the beginning of the first end iw window corresponding back has not coincide with the optimum path, into the first (iw-l) window, proceed according to the path traceback register values ​​each time, until the back

到与第(iw-l )滑动窗口的局部最优路径合并为止。 To merge the local optimal path section (iw-l) until the sliding window. 根据滑动窗口中的辅助部 The auxiliary sliding window

分长度c的设置,可见所有优选路径与最优路径以很大的概率在c个时刻内合并,即使得各优选路径与最优路径相分离的长度不大于一个滑动窗口长度,因 C provided partial length, preferably all the paths found to the optimal path in the high probability c in time were combined to give each of the preferred path and even if the optimum path length is not greater than the phase separation of a sliding window length, because

此在第iw滑动窗口的回溯,如果不在第iw滑动窗口内与最优路径相合并,那么回溯一定能够在第(iw-1 )窗内与最优路径相重合。 This back of the sliding window iw, if the phases are combined with the optimal path is not within the sliding window first iw, it is possible to surely back section (iw-1) within the window coincides with the optimal path.

如果回溯的局部次优路径在第iw窗口内与该窗口的局部优选路径相分离了,那么舍弃在之前滑动窗口获取的局部次优路径,存储当前回溯获取的局部次优路径。 If the local sub-optimal path back in the first window iw preferred path to the local phase separation of the window, then the discard suboptimal paths in the local sliding window acquired before, the local storage of the current sub-optimal path backtracking acquired. 根据各优选路径与最优路径只可能在一连续的时刻段内相分离,因此,可以只存储与最优路径相分离时刻段的路径,最后历遍所有滑动窗口查找后结合全局最优路径,获取该局部优选路径的全局路径。 The phase separation only possible with the optimum path for each preferred path in a continuous time period, and therefore, can store only the time path of optimal path segments separated. Finally after finding the global optimal path through all the epoch sliding window, Get the local global path preferred path.

步骤S1009,:舍弃之前获取的次优路径。 Step S1009 ,: abandon sub-optimal path taken before.

步骤S10010:停止回溯。 Step S10010: stop backtracking.

第3优、第4优路径以及其他的优选路径的回溯与地有路径回溯同理,在此不作赘述。 Preferably 3, preferably 4 and a path back to ground path has other preferred path back Similarly, this will not be repeated herein.

步骤S908: iw=iw+l。 Step S908: iw = iw + l.

使iw加1,更新滑动窗口标识iw。 So iw plus 1, sliding window update identifier iw.

步骤S909:如果iw不小于滑动窗口的总数,执行步骤S910;否则,返回步骤S簡。 Step S909: If the total number is not less than iw sliding window, to step S910; otherwise, the process returns to step S profile.

步骤S910:根据路径寄存器获取全局第/优优选路径,并对第/优路径输 Step S910: Get Global / UU selected route according to a register, and a first / output path preferably

出的译码数据进行差错校验。 The decoded data for error checking.

至此,根据所存储的路径寄存器信息,可以获取到所有需要查找的全局优选路径。 Thus, according to the path information stored in the register, you can obtain all the need to find a global preferred path. 根据寄存器回溯全局第/优选路径,获取全局第/优选路径对应的译码数据,并对译码数据进行差错校验。 The error-checks the register back Global / preferred path, acquiring Global / preferred path corresponding to the decoded data, and the decoded data. 步骤S911:差错校验结果判决。 Step S911: error checking result judgment.

如果差错校验正确,执行步骤S914;否则,执行步骤S912。 If error checking is correct, to step S914; otherwise, to step S912. 步骤S912: /=/+1。 Step S912: / = / + 1. 使路径号加l,更新路径号。 The path-number plus l, updated path number. 步骤S913:判定/是否大于L。 Step S913: determining / is greater than L.

如果/不大于所预设的需查找的全局优选路径的总数L,执行步骤S914,否则返回步骤S910。 If the total number / Global preferably not greater than the predetermined path of the required lookup L, perform step S914, the otherwise returns to step S910.

步骤S914:输出译码结果。 Step S914: decoding result output.

如果步骤S911中,差错校验正确,那么输出的译码结果是全局第/优路径的译码数据,结束译码。 If the step S911, the error check is correct, then the output of the decoding result is a decoded data Global / optimal path, the decoding end. 否则,输出的译码结果是:译码失败,结束译码。 Otherwise, the decoding result is output: decoding failure, the end of the decoding.

以上对本发明所提供的一种巻积码译码方法进行了详细介绍,本文中应用 For one or more code decoding convolving the present invention provides methods described in detail herein apply

了具体个例对本发明的原理及实施方式进行了阐述,以上实施例的说明只是用于帮助理解本发明的方法及其核心思想;同时,对于本领域的一般技术人员, 依据本发明的思想,在具体实施方式及应用范围上均会有改变之处,综上所述, 本说明书内容不应理解为对本发明的限制。 Specific examples of the principles and embodiments of the invention are set forth in description of the above embodiment is only used to help understand the method and core idea of ​​the present invention; while those of ordinary skill in the art, according to the ideas of the present invention, may make modifications to the specific embodiments and application scope of the specification shall not be construed as limiting the present invention.

Claims (13)

1、一种卷积码译码方法,其特征是,包括以下步骤: A、预设滑动窗口,其中所述的滑动窗口由判决输出部分以及辅助部分组成,将卷积码帧分段在所述的各滑动窗口中,使所述的各滑动窗口的判决输出部分的长度的总和等于卷积码帧长,预设滑动窗口标识iw,以及需要查找的全局优选路径的总数L,初始化iw等于0; B、使iw累加1,获取第iw滑动窗口内,每时刻的每状态点的状态与每时刻的每状态点相对于卷积码开始时刻的路径度量,根据所述的路径度量选取L个优选路径度量,并存储所述的L个优选路径度量分别对应的优选路径,以第iw滑动窗口的最末端的时刻的一状态点为起始点开始回溯,在每时刻,根据回溯状态点的路径值,按照最优路径回溯,回溯到第iw滑动窗口的最开始端的时刻,获取第iw滑动窗口的判决输出部分的局部最优路径,并且以所述 A convolutional code decoding method characterized by comprising the following steps: A, a predetermined sliding window, wherein the sliding window output from the decision section and an auxiliary part, the frame segmentation in the convolutional code each of said sliding window, that the sum of the length of the decision output portion of each of the sliding window frame length equal to a convolutional code, a predetermined sliding window identifier iw, and the total number L of the global need to find a preferred path, is equal to the initialization iw 0; B, so that the accumulated iw 1, obtaining the first sliding window iw, the state of each state for each time point and each time point for each state path metric with respect to the start time of a convolutional code, according to the path metrics of the selected L preferred path metrics, and storing said preferred path L preferred path metrics respectively corresponding to a state at a point of time of the endmost iw sliding window back to the start point, at every moment, according to the state of point backtracking path value, in accordance with the optimal path back, back to the beginning end of the first time iw sliding window, acquire the local optimal path decision output portion of the sliding window iw, and to the 局部最优路径在判决输出部分最末端的时刻的状态点为回溯起始点,根据回溯的状态点的优选路径值回溯,获取除局部最优路径外的其它局部优选路径,执行步骤C; C、如果iw小于滑动窗口的总数目,返回步骤B,否则,执行步骤D; D、根据各滑动窗口所获取的所有局部优选路径,获取所有的局部优选路径以及各全局优选路径的译码,并对所获取的全局优选路径输出的译码数据进行差错校验,根据校验结果,输出译码结果,其中所述的全局优选路径包括全局最优路径,以及除全局最优路径外的其它全局优选路径。 Local optimum path timing decision output state point endmost portion as a starting point back, in accordance with the value of the preferred path back status point back acquires preferred path in addition to other local partial optimal path, step C; C, If less than the total number iw sliding window, return to step B, and otherwise, step D; D, according to the preferred paths of all of the local sliding window acquired, obtaining all the local decoding paths and preferably each global preferred path, and the decoded data acquired global preferred path outputted error check, according to the check result, it outputs the decoding result, wherein the global path preferably comprises a global optimum path, and preferably in addition to other global global optimal path path.
2、 根据权利要求1所述的巻积码译码方法,其特征是,步骤A中所述的滑动窗口通过如下方式设置:设置所述的判决输出部分的长度,以及辅助部分的长度,其中辅助部分的长度具体是根据所述的巻积码的译码深度设置。 2, convolving code decoding method according to claim 1, characterized in that, in step A sliding window according to the following manner: Set the length of said decision output portion, and the length of the auxiliary portion, wherein specific length of the auxiliary portion is set according to the depth of convolving decoding said code.
3、 根据权利要求1所述的巻积码译码方法,其特征是,步骤B所述的根据所述的路径度量选取L个优选路径度量具体是:如果巻积码译码器输入是硬判决,则选择较小的L个的路径度量作为所述的优选路径度量;如果所述的巻积码译码器输入是软判决,则选择较大的L个路径度量,作为所述的优选路径度量。 3, convolving code decoding method according to claim 1, characterized in that said step B according to the selected path metric the path metric particularly preferred L is: If the input code decoder convolving hard decision, the L is selected smaller path metric as the preferred path metric; convolving if said input code decoder is a soft decision, the larger L is selected path metrics as the preferred path metric.
4、 根据权利要求3所述的巻积码译码方法,其特征是,所述的L个优选路径度量分别对应的优选路径具体用一具有L个比特单元的数值表示,其中备比特单元的值分别表示各优选路径。 4. The convolving code decoding method according to claim 3, characterized in that the preferred path is preferably the L path metrics respectively corresponding to a specific value having a unit represented by L bits, wherein the bit cells prepared values ​​represent each preferred path.
5、 根据权利要求4所述的巻积码译码方法,其特征是,在所述步骤B中,所述的获取第iw滑动窗口的判决输出部分的局部最优游4圣过程中,如果所述的巻积码译码器输入是硬判决,则选取第iw滑动窗口的最末端的时刻的路径度量最小的状态点为起始点开始回溯;如果所述的巻积码译码器输入是软判'决,则选取第iw滑动窗口的最末端的时刻的路径度量最大的状态点为起始点开始回溯。 5. The method of claim convolving code decoding according to claim 4, wherein, in the step B, the acquisition of local decision output portion iw sliding window run 4 St. optimal course, if convolving said input code decoder is a hard decision, the path timing of the endmost iw sliding window is selected minimum state metric is the starting point to point back; convolving if said input code decoder is a soft judgment 'decision, the path timing of the endmost iw sliding window is selected maximum state metric point is the starting point back.
6、 根据权利要求5所述的巻积码译码方法,其特征是,步骤B中所述的根据回溯的状态点的优选路径值回溯,具体包括以下步骤:Bl、根据每时刻的每状态点对应的各优选路径,预设一优选路径回溯表, 所述的优选路径回溯表,具体反映各优选路径与所述的各优选路径值的对应关系。 6. convolving code decoding method according to claim 5, characterized in that, in step B, the value of the path traceback according to the preferred state of the back point, includes the following steps: Bl, in accordance with each state for each time point each preferred path point corresponding to a predetermined preferred path back table, the preferred path back table, particularly preferred path to reflect the relationship between each value of the respective paths is preferable. B2、在回溯的每时刻,获取当前回溯所在状态点对应的各优选路径,并对照所述的优选路径回溯表,根据所述的优选路径回溯表,回溯获取所述的局部优选路径。 B2, every time the backtracking, backtracking acquires each preferred path where the current state of the corresponding point, and the control table according to the preferred path back, according to the preferred path back table, the preferred path back to the local acquisition.
7、 根据权利要求6所述的巻积码译码方法,其特征是:所述步骤B2中, 如果在回溯到第iw滑动窗口的最开始端的时刻之前的每时刻的状态点,所述状态点对应的优于当前回溯的优选^各径的其他所有优选^各径对应的^各径值与当前回溯路径的路径值均相同,则回溯到第iw滑动窗口的最开始端的时刻, 停止回溯。 7, according to claim convolving code decoding method according to claim 6, characterized in that: said step B2, if the status of each point in time before the first iw back to the beginning of the sliding window of time ends, said state over current backtracking points corresponding to preferably ^ all other preferably each path ^ of each path corresponding to ^ each radius value and the value of the path the current back path are identical, then back to the time the beginning end of the first iw sliding window, stopping back .
8、 根据权利要求6所述的巻积码译码方法,其特征是,在步骤B2中进一步包括以下步骤:B21、在回溯的每时刻,获取当前回溯所在状态点对应的各优选路径值后, 比较各优选路径值中的当前需要回溯获取的局部优选路径对应的优选路径值以及最优路径值是否相同,在第一次不相同后,在回溯的下一时刻以及以后的回溯的每时刻,获取回溯的状态点的状态,并比较所述状态点的状态是否与局部最优^4圣在所述状态点所在时刻的状态相同,如果不同,继续回溯;否则,停止回溯,保存当前回溯的局部优选路径。 After B21, each time at the back, acquires each preferred path back where the current state value corresponding to the point: 8, according to claim convolving code decoding method according to claim 6, characterized in that, further comprising the step of step B2 , preferred path comparing the local value of the current preferred path requires backtracking preferred path corresponding to the obtained value and the value of the optimal path are the same, are not the same after the first, and at the next moment back each time after traceback to get back the state point of the state, and comparing the state of the state is the optimal point ^ 4 in the state of San moment of the state where the same point with the local, if different, continue back; otherwise, stop back, save the current retrospective local preferred path.
9、 根据权利要求8所述的巻积码译码方法,其特征是,步骤B21所述的保存当前回溯的局部优选^各径后,进一步包括:舍弃之前保存的与当前回溯的优选路径的路径号相同的其他局部优选路径。 9. convolving code decoding method according to claim 8, characterized in that, preferably after partial ^ each path stored in said current backtracking step B21, further comprising: discarding the current preferred path back to the previously saved other preferred topical same path number.
10、 根据权利要求9所述的巻积码译码方法,其特征是,在所述步骤B 中进一步包括以下步骤:B3、如果iw大于2,则释放滑动窗口标识小于(iw-l )的所有滑动窗口内的每时刻的每状态点的状态、路径度量,以及路径的存储空间。 10, according to convolving code decoding method according to claim 9, characterized in that, further comprising the step of the step B: B3, if iw greater than 2, less than the release of the sliding window identifier (iw-l) of each state of each state at the point in time of all the sliding window, the path metric and path memory.
11、 根据权利要求1所述的巻积码译码方法,其特征是,步骤D中所述的全局最优路径的获取具体是:根据所获取的各滑动窗口的局部优选路径,获取全局最优路径。 11. convolving code decoding method according to claim 1, characterized in that said step D for specific global optimal path is: the preferred path of each partial sliding window acquired, obtaining global optimal shortest paths.
12、 根据权利要求1至11其中任意之一所述的巻积码译码方法,其特征是,步骤D中所述的除全局最优路径之外的其它全局优选路径的获取具体是: 根据保存的各局部优选路径,以及各局部最优路径,获取所述的全局优选路径。 12, 1 to 11 wherein convolving code decoding method according to any one of the preceding claims, wherein, in step D, the other global preferred path other than the path of obtaining global optimal specifically: The each path preferably stored locally, and each local optimal path, the acquired global preferred path.
13、 根据权利要求12所述的巻积码译码方法,其特征是,步骤D中对所获取的全局优选路径输出的译码数据进行差错校验,根据校验结果,输出译码结果,具体包括以下步骤:Dl、预设一实数变量/,初始化/等于0;D2、对全局第/优路径输出的译码数据进行差错校验,如果校验正确,执行步骤D4,否则,进一步判定/是否小于所预设的需查找的全局优选路径的总数,如果小于,执行D3;否则,执行步骤D4;D3、使/累加1,返回D2; D4、输出译码结果,结束译码。 13. The method of claim convolving code decoding according to claim 12, wherein, in step D the acquired decoded data output from the preferred path global error-checks, according to the check result, outputs the decoding result, comprises the steps of: Dl, a pre-set real number variable /, initialization / equal to 0; D2, the decoded data global / shortest paths outputted error check, if the check is successful, step D4, otherwise, it is further determined / total number is smaller than a predetermined preferred path global need to find, if less than D3 executed; otherwise, step D4; D3, so / incremented by 1, returned D2; D4, the output of the decoding result, the decoding end.
CN 200610087453 2006-06-08 2006-06-08 A coding method of convolution code CN100536351C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 200610087453 CN100536351C (en) 2006-06-08 2006-06-08 A coding method of convolution code

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 200610087453 CN100536351C (en) 2006-06-08 2006-06-08 A coding method of convolution code

Publications (2)

Publication Number Publication Date
CN1968023A CN1968023A (en) 2007-05-23
CN100536351C true CN100536351C (en) 2009-09-02

Family

ID=38076609

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200610087453 CN100536351C (en) 2006-06-08 2006-06-08 A coding method of convolution code

Country Status (1)

Country Link
CN (1) CN100536351C (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101997553B (en) 2009-08-13 2013-03-20 中兴通讯股份有限公司 Method and device for decoding convolution code
CN105335747B (en) * 2014-08-04 2019-03-29 联想(北京)有限公司 A kind of data processing method and electronic equipment

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5390198A (en) 1993-05-26 1995-02-14 The Boeing Company Soft decision viterbi decoder for M-ary convolutional codes
US6304612B1 (en) 1998-05-26 2001-10-16 U.S. Philips Corporation Transmission system having a simplified channel decoder
CN1124691C (en) 1999-09-13 2003-10-15 华为技术有限公司 Serial/parallel caseade convolutional code decoder and method for realizing decoding
EP1471677A1 (en) 2003-04-23 2004-10-27 STMicroelectronics N.V. Method of blindly detecting a transport format of an incident convolutional encoded signal, and corresponding convolutional code decoder

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5390198A (en) 1993-05-26 1995-02-14 The Boeing Company Soft decision viterbi decoder for M-ary convolutional codes
US6304612B1 (en) 1998-05-26 2001-10-16 U.S. Philips Corporation Transmission system having a simplified channel decoder
CN1124691C (en) 1999-09-13 2003-10-15 华为技术有限公司 Serial/parallel caseade convolutional code decoder and method for realizing decoding
EP1471677A1 (en) 2003-04-23 2004-10-27 STMicroelectronics N.V. Method of blindly detecting a transport format of an incident convolutional encoded signal, and corresponding convolutional code decoder

Also Published As

Publication number Publication date
CN1968023A (en) 2007-05-23

Similar Documents

Publication Publication Date Title
Niu et al. CRC-aided decoding of polar codes
AU2005225107B2 (en) Method and apparatus for decoding low density parity check code using united node processing
CN101044688B (en) Memory efficient ldpc decoding methods and apparatus
CA2273592C (en) Processing of state histories in viterbi decoding
EP0798889B1 (en) Error control method and error control device for digital communication
JP3537183B2 (en) Reliability information signal generation method and apparatus
US7607073B1 (en) Methods, algorithms, software, circuits, receivers and systems for iteratively decoding a tailbiting convolutional code
CN100512075C (en) Digital signal processor, device and system using the digital signal processor
KR100713331B1 (en) Apparatus and method for stopping iterative decoding in a cdma mobile communications system
FI118240B (en) Method and apparatus for concatenated CRC outer code decoding
US6901117B1 (en) Soft output decoder for convolutional codes
ES2301492T3 (en) Turbo decoding with soft Viterbi decoder output.
JP3730885B2 (en) Decoder error correction turbo code
US6829313B1 (en) Sliding window turbo decoder
US6877132B1 (en) Method and apparatus for channel decoding of tail-biting convolutional codes
US20040155802A1 (en) Method of decoding a variable-length codeword sequence
CA2020899C (en) Generalized viterbi decoding algorithms
EP1022860B1 (en) Turbo decoding with variable number of iterations
CN1194475C (en) Data transmitting method, data transmitting system, receiving method and receiver
CN105227189B (en) It is segmented the polarization code coding/decoding method of CRC auxiliary
US7116732B2 (en) Method and apparatus for decoding a bit sequence
CN1104785C (en) Position diversity reception, base station main frame apparatus and processing method for receiving signal
US6148431A (en) Add compare select circuit and method implementing a viterbi algorithm
US20010007142A1 (en) Enhanced viterbi decoder for wireless applications
KR20020038556A (en) Data transmission method, data transmission system, trasmitter and receiver

Legal Events

Date Code Title Description
C06 Publication
C10 Request of examination as to substance
C14 Granted