Nothing Special   »   [go: up one dir, main page]

CN107911195B - CVA-based tail-biting convolutional code channel decoding method - Google Patents

CVA-based tail-biting convolutional code channel decoding method Download PDF

Info

Publication number
CN107911195B
CN107911195B CN201710979858.0A CN201710979858A CN107911195B CN 107911195 B CN107911195 B CN 107911195B CN 201710979858 A CN201710979858 A CN 201710979858A CN 107911195 B CN107911195 B CN 107911195B
Authority
CN
China
Prior art keywords
decoding
state
maximum
metric value
sequence
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201710979858.0A
Other languages
Chinese (zh)
Other versions
CN107911195A (en
Inventor
于秀兰
秦利科
张祖凡
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing University of Post and Telecommunications
Original Assignee
Chongqing University of Post and Telecommunications
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 Chongqing University of Post and Telecommunications filed Critical Chongqing University of Post and Telecommunications
Priority to CN201710979858.0A priority Critical patent/CN107911195B/en
Publication of CN107911195A publication Critical patent/CN107911195A/en
Application granted granted Critical
Publication of CN107911195B publication Critical patent/CN107911195B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00Arrangements for detecting or preventing errors in the information received
    • H04L1/004Arrangements for detecting or preventing errors in the information received by using forward error control
    • H04L1/0056Systems characterized by the type of code used
    • H04L1/0059Convolutional codes
    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M13/00Coding, 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
    • H03M13/03Error detection or forward error correction by redundancy in data representation, i.e. code words containing more digits than the source words
    • H03M13/23Error detection or forward error correction by redundancy in data representation, i.e. code words containing more digits than the source words using convolutional codes, e.g. unit memory codes
    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M13/00Coding, 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
    • H03M13/37Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
    • H03M13/39Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes
    • H03M13/41Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using the Viterbi algorithm or Viterbi processors
    • H03M13/413Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using the Viterbi algorithm or Viterbi processors tail biting Viterbi decoding

Landscapes

  • Physics & Mathematics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Error Detection And Correction (AREA)

Abstract

本发明请求保护一种基于CVA的咬尾卷积码信道译码方法,涉及到移动通信技术领域。为降低译码复杂度,提高译码效率,本发明提出了一种基于循环维特比算法(Circular Viterbi Algorithm,CVA)的译码算法,根据接收到信息序列的似然比信息来确定出可靠性最高的译码起始位置,从该起始位置开始进行修正的维特比译码,通过加比选过程,删除不可能的状态位置,经过几次迭代搜索后,幸存路径会收敛到较少的幸存状态,选择最大度量值的状态作为最终译码的起始位置,根据其相对应的幸存路径估计出译码结果。该算法具有更快的收敛速度,译码效率得到了进一步提高,降低了译码时延。

Figure 201710979858

The present invention claims to protect a CVA-based tail-biting convolutional code channel decoding method, which relates to the technical field of mobile communication. In order to reduce the decoding complexity and improve the decoding efficiency, the present invention proposes a decoding algorithm based on the Circular Viterbi Algorithm (CVA), and determines the reliability according to the likelihood ratio information of the received information sequence. The highest decoding start position, from which the modified Viterbi decoding is performed, and the impossible state position is deleted through the addition and selection process. After several iterations of search, the surviving path will converge to less Surviving state, select the state with the largest metric value as the starting position of the final decoding, and estimate the decoding result according to its corresponding surviving path. The algorithm has a faster convergence speed, the decoding efficiency is further improved, and the decoding delay is reduced.

Figure 201710979858

Description

一种基于CVA的咬尾卷积码信道译码方法A Channel Decoding Method for Tail-biting Convolutional Codes Based on CVA

技术领域technical field

本发明属于移动通信的信道编译码技术领域,特别涉及到LTE中PBCH解码的咬尾卷积码译码方法。The invention belongs to the technical field of channel coding and decoding of mobile communication, and particularly relates to a tail-biting convolutional code decoding method for PBCH decoding in LTE.

背景技术Background technique

随着通信技术的发展,数字通信逐步取代模拟通信成为通信技术的主流。然而,数字信息在信道中传输时,会受到噪声的干扰,误码的产生总是不可避免的。为了在已知信噪比的情况下达到一定的误码率指标,在合理设计基带信号,选择调制、解调方式,并采用频域均衡或时域均衡措施的基础上,还应该采用差错控制编码等信道编码技术,使误码率进一步降低。卷积码和分组码是差错控制编码的两种主要形式,在编码器复杂度相同的情况下,卷积码的性能优于分组码,因此卷积码被应用在许多无线通信的标准中。With the development of communication technology, digital communication has gradually replaced analog communication as the mainstream of communication technology. However, when digital information is transmitted in the channel, it will be interfered by noise, and the generation of bit errors is always inevitable. In order to achieve a certain bit error rate index under the condition of known signal-to-noise ratio, on the basis of rationally designing baseband signals, selecting modulation and demodulation methods, and adopting frequency domain equalization or time domain equalization measures, error control should also be used. Coding and other channel coding techniques further reduce the bit error rate. Convolutional codes and block codes are the two main forms of error control coding. In the case of the same encoder complexity, the performance of convolutional codes is better than that of block codes, so convolutional codes are used in many wireless communication standards.

自卷积码被发明以来,它一直作为一种高效的信道编码技术应用在通信系统中。LTE系统作为3GPP标准化组织提出的无线空口技术演进,目前正在全球范围内得到大力的发展和部署,该系统需要实现更高的带宽、更大的容量、更高的数据传输速率、更低的传输时延、更低的运营成本、同时为满足用户对于广播及多播业务等实时业务的高速率需求,LTE系统在信道编码过程中根据不同传输信道采用了Turbo编码和咬尾卷积码编码,其中咬尾卷积编码主要用于广播信道PBCH、上下行的控制信道信息DCI、UCI编码过程,并且对于不同类型的传输信道和控制信道使用的编码方案和编码速率也有所不同。咬尾卷积码编码过程首先会用到咬尾技术,即保证格形起始和结尾在同一个状态,这就需要将被编码的数据块的最后几个比特作为寄存器的初始状态。Since the invention of convolutional codes, it has been used in communication systems as an efficient channel coding technique. As the evolution of wireless air interface technology proposed by the 3GPP standardization organization, the LTE system is currently being vigorously developed and deployed around the world. The system needs to achieve higher bandwidth, larger capacity, higher data transmission rate, and lower transmission rate. Time delay, lower operating costs, and at the same time, in order to meet users' high-speed requirements for real-time services such as broadcast and multicast services, the LTE system uses Turbo coding and tail-biting convolutional coding according to different transmission channels in the channel coding process. The tail-biting convolutional coding is mainly used in the coding process of broadcast channel PBCH, uplink and downlink control channel information DCI, and UCI, and the coding scheme and coding rate used for different types of transmission channels and control channels are also different. The tail-biting convolutional code encoding process first uses the tail-biting technique, that is, to ensure that the start and end of the trellis are in the same state, which requires the last few bits of the encoded data block as the initial state of the register.

采用咬尾方式编码的卷积码不仅消除了用已知比特初始化编码器所导致的误码率损失,同时咬尾结构可以对所有的信息比特提供相同的保护能力。正是因为咬尾卷积码的这些优点,它被广泛应用在各种通信系统中,作为控制信令的编码方式。对于较短的信息序列,咬尾编码对码率的保护是很可观的,比如LTE中广播信道,在加了循环冗余检验比特之后共有40比特,这40比特的信息序列如果不用咬尾编码技术,码率损失将达到13%。目前采用咬尾卷积码作为控制信道编码方式通信标准的系统有:EDGE、WIMAX和LTE等。The convolutional code encoded by tail-biting not only eliminates the loss of bit error rate caused by initializing the encoder with known bits, but also the tail-biting structure can provide the same protection capability for all information bits. It is precisely because of these advantages of tail-biting convolutional codes that it is widely used in various communication systems as a coding method for control signaling. For short information sequences, the protection of code rate by tail-biting coding is very considerable. For example, in the broadcast channel of LTE, there are 40 bits in total after the cyclic redundancy check bit is added. If the 40-bit information sequence does not need tail-biting coding technology, the bit rate loss will reach 13%. At present, the systems that use tail-biting convolutional codes as the control channel coding mode communication standards include EDGE, WIMAX, and LTE.

咬尾卷积码虽然有很多优点,但是对于译码器来说,由于不知道译码的起始状态和终止状态,基于维特比算法的最优译码方案实现过于复杂,而基于循环维特比译码方法在达到最大迭代次数前,可能已检测出最优咬尾路径,多余的迭代造成了译码时延的加长和资源的浪费。目前还没有实用的基于循环维特比算法的最优译码方案,基于以上问题,本发明设计出了一种基于CVA的咬尾卷积码信道译码方法,适合工程实际应用。Although the tail-biting convolutional code has many advantages, for the decoder, because it does not know the starting state and ending state of the decoding, the implementation of the optimal decoding scheme based on the Viterbi algorithm is too complicated. Before the decoding method reaches the maximum number of iterations, the optimal tail-biting path may have been detected, and the redundant iterations lead to longer decoding delay and waste of resources. At present, there is no practical optimal decoding scheme based on the cyclic Viterbi algorithm. Based on the above problems, the present invention designs a CVA-based tail-biting convolutional code channel decoding method, which is suitable for practical engineering applications.

发明内容SUMMARY OF THE INVENTION

本发明旨在解决以上现有技术的问题。提出了一种可以在低复杂度下实现咬尾卷积码最优译码的基于CVA的咬尾卷积码信道译码方法。本发明的技术方案如下:The present invention aims to solve the above problems of the prior art. A CVA-based channel decoding method for tail-biting convolutional codes is proposed, which can realize optimal decoding of tail-biting convolutional codes with low complexity. The technical scheme of the present invention is as follows:

一种基于CVA的咬尾卷积码信道译码方法,其包括以下步骤:A CVA-based tail-biting convolutional code channel decoding method, comprising the following steps:

101、输入数据流,根据接收到数据流信息序列的似然比信息,确定出可靠性最高的译码起始位置,从该起始位置开始进行修正的维特比译码;101, input the data stream, according to the likelihood ratio information of the received data stream information sequence, determine the decoding starting position with the highest reliability, and start from the starting position to carry out the modified Viterbi decoding;

102、执行第i次迭代译码,到达每个状态有2条路径,分别计算到达每个状态的2条分支之间的分支度量值,选择最优的一个,此过程成为加比选操作。通过加比选过程并更新译码参数,删除不可能的状态,开始下次迭代译码。102. Execute the i-th iterative decoding, there are two paths to each state, calculate the branch metric values between the two branches reaching each state, and select the optimal one, and this process becomes the addition and comparison operation. By adding the selection process and updating the decoding parameters, the impossible state is deleted, and the next iterative decoding is started.

103、经过几次迭代搜索后,根据迭代停止准则,选择最大度量值的状态作为最终译码的起始位置,依据最终起始位置和对应的网格路径,估计出译码结果。103. After several iterative searches, according to the iterative stop criterion, select the state with the largest metric value as the starting position of the final decoding, and estimate the decoding result according to the final starting position and the corresponding trellis path.

进一步的,所述步骤101在译码开始的时候,对接收到的信息进行处理,选择可靠性最高的位置lopt作为译码的起始位置。接收到的符号为

Figure BDA0001439141250000021
计算得到似然比信息
Figure BDA0001439141250000022
定义lopt的计算公式为Further, in the step 101, when the decoding starts, the received information is processed, and the position l opt with the highest reliability is selected as the starting position of the decoding. The received symbol is
Figure BDA0001439141250000021
Calculate the likelihood ratio information
Figure BDA0001439141250000022
Define the calculation formula of l opt as

Figure BDA0001439141250000031
Figure BDA0001439141250000031

式中(l+Q)L=(l+Q)modL,其中Q是待确定的量,在具体应用中将根据不同的码字选择合适的值,当Q为接受序列的长度,此时每个位置的可靠度完全一样,那么译码从头开始。In the formula (l+Q) L = (l+Q) modL, where Q is the quantity to be determined, in the specific application, the appropriate value will be selected according to different code words, when Q is the length of the accepted sequence, then each time The reliability of each position is exactly the same, then the decoding starts from the beginning.

进一步的,所述步骤101还包括初始化的步骤,i=0,

Figure BDA0001439141250000032
Figure BDA0001439141250000033
其中,
Figure BDA0001439141250000034
表示第i次迭代中,状态sj的度量值,j为译码器状态数,j=1,2,3...2v,v是移位寄存器的个数,
Figure BDA0001439141250000035
表示最优路径的度量值,
Figure BDA0001439141250000036
表示执行一次译码后状态sj的度量值的净增量。Further, the step 101 also includes the step of initialization, i=0,
Figure BDA0001439141250000032
Figure BDA0001439141250000033
in,
Figure BDA0001439141250000034
represents the metric value of the state s j in the ith iteration, j is the number of decoder states, j=1, 2, 3...2 v , v is the number of shift registers,
Figure BDA0001439141250000035
represents the metric value of the optimal path,
Figure BDA0001439141250000036
represents the net increment of the metric value of state s j after performing one decoding.

进一步的,所述步骤102具体包括步骤:通过加比选操作,步进每个序列位置各状态的度量值,循环一周后更新各个状态的度量值和似然路径,检测出最大度量值

Figure BDA0001439141250000037
和相应的咬尾路径
Figure BDA0001439141250000038
Further, the step 102 specifically includes the steps of: step by step the metric value of each state of each sequence position by adding a comparison operation, update the metric value and the likelihood path of each state after a cycle, and detect the maximum metric value.
Figure BDA0001439141250000037
and the corresponding tail-biting path
Figure BDA0001439141250000038

维特比Viterbi算法采用接收码元条件概率的乘积的最大值作为估计序列,采用对数似然函数来表示为

Figure BDA0001439141250000039
其中,y表示码字序列经过传输映射的结果,r表示接收端得到的序列,简化上式中的对数函数求和运算,可以定义如下码元度量
Figure BDA00014391412500000310
其中,令The Viterbi algorithm uses the maximum value of the product of the conditional probabilities of the received symbols as the estimated sequence, and uses the log-likelihood function to express as
Figure BDA0001439141250000039
Among them, y represents the result of the codeword sequence after transmission mapping, and r represents the sequence obtained by the receiving end. Simplifying the logarithmic function summation operation in the above formula, the following symbol metric can be defined
Figure BDA00014391412500000310
Among them, let

Figure BDA00014391412500000311
Figure BDA00014391412500000311

此种情况下,Viterbi算法中的码元度量值就是在编码网格上选择与接收序列r之间汉明距离最小的码字作为译码输出。In this case, the symbol metric value in the Viterbi algorithm is to select the codeword with the smallest Hamming distance from the received sequence r on the encoding grid as the decoding output.

进一步的,所述步骤102执行第i次迭代译码,如果当前迭代得出的状态净增量最大值

Figure BDA00014391412500000312
大于存储的最大度量值
Figure BDA00014391412500000313
Figure BDA00014391412500000314
则更新最大度量值
Figure BDA0001439141250000041
和相应的最大咬尾路径
Figure BDA0001439141250000042
Further, the step 102 executes the i-th iterative decoding, if the maximum value of the state net increment obtained by the current iteration is
Figure BDA00014391412500000312
Greater than the maximum stored metric
Figure BDA00014391412500000313
which is
Figure BDA00014391412500000314
then update the maximum metric
Figure BDA0001439141250000041
and the corresponding maximum tail-biting path
Figure BDA0001439141250000042

本发明的优点及有益效果如下:The advantages and beneficial effects of the present invention are as follows:

本发明设计了一种迭代停止准则,通过统计当前迭代完成时各状态度量值的净增量

Figure BDA0001439141250000043
大于存储的最大度量值
Figure BDA0001439141250000044
的数量num(i),若num(i)<=num(i-1),则停止迭代;否则继续迭代直到达到最大迭代次数。这样当译码提前完成时,减小了译码延时,提高了译码效率。The present invention designs an iterative stop criterion, which calculates the net increment of each state metric value when the current iteration is completed.
Figure BDA0001439141250000043
Greater than the maximum stored metric
Figure BDA0001439141250000044
If num(i)<=num(i-1), the iteration is stopped; otherwise, the iteration is continued until the maximum number of iterations is reached. In this way, when the decoding is completed in advance, the decoding delay is reduced, and the decoding efficiency is improved.

附图说明Description of drawings

图1是本发明的咬尾卷积码译码流程图;Fig. 1 is the tail-biting convolutional code decoding flow chart of the present invention;

图2生成多项式为(7,5)的卷积码编译码格形图。Fig. 2 generates a trellis diagram of convolutional coding with the polynomial of (7, 5).

具体实施方式Detailed ways

下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、详细地描述。所描述的实施例仅仅是本发明的一部分实施例。The technical solutions in the embodiments of the present invention will be described clearly and in detail below with reference to the accompanying drawings in the embodiments of the present invention. The described embodiments are only some of the embodiments of the invention.

本发明解决上述技术问题的技术方案是:The technical scheme that the present invention solves the above-mentioned technical problems is:

本发明所述的一种基于CVA的咬尾卷积码信道译码方法适用于现有移动通信系统(如EDGE、LTE、WIMAX),也适用于下一代移动通信系统中咬尾卷积码译码。根据接收到的信息序列,通过迭代对不可能的起始状态逐一排除,最终寻找到最优咬尾路径。本发明所述译码方法通过对循环陷阱的有效处理,加快了译码器的收敛速度,同时算法简单、易于实现,有重要应用价值。A CVA-based tail-biting convolutional code channel decoding method according to the present invention is suitable for existing mobile communication systems (such as EDGE, LTE, WIMAX), and is also suitable for tail-biting convolutional code decoding in next-generation mobile communication systems code. According to the received information sequence, the impossible initial states are eliminated one by one through iteration, and finally the optimal tail-biting path is found. The decoding method of the invention speeds up the convergence speed of the decoder by effectively processing the loop trap, and meanwhile, the algorithm is simple and easy to implement, and has important application value.

一种基于CVA的咬尾卷积码信道译码方法,其特征在于,包括以下步骤:A CVA-based tail-biting convolutional code channel decoding method, comprising the following steps:

Step1.在译码开始的时候,对接收到的信息进行处理,选择可靠性最高的位置lopt作为译码的起始位置。接收到的符号为

Figure BDA0001439141250000045
计算得到似然比信息
Figure BDA0001439141250000046
定义lopt的计算公式为Step1. At the beginning of decoding, the received information is processed, and the position l opt with the highest reliability is selected as the starting position of decoding. The received symbol is
Figure BDA0001439141250000045
Calculate the likelihood ratio information
Figure BDA0001439141250000046
Define the calculation formula of l opt as

Figure BDA0001439141250000051
Figure BDA0001439141250000051

式中(l+Q)L=(l+Q)modL,其中Q是待确定的量,在具体应用中将根据不同的码字选择合适的值。当Q值越小时,可靠度的作用越明显;由于接受序列服从正态分布,因此当Q值越大时,由式计算得到的每个位置的可靠度越接近;极限情况是Q为接受序列的长度,此时每个位置的可靠度完全一样,那么译码从头开始。In the formula (l+Q) L =(l+Q)modL, where Q is a quantity to be determined, and an appropriate value will be selected according to different codewords in specific applications. When the value of Q is smaller, the effect of reliability is more obvious; since the acceptance sequence obeys the normal distribution, when the value of Q is larger, the reliability of each position calculated by the formula is closer; the limit case is that Q is the acceptance sequence The length of , at this time the reliability of each position is exactly the same, then the decoding starts from the beginning.

Step2.初始化,i=0,

Figure BDA0001439141250000052
Step2. Initialization, i=0,
Figure BDA0001439141250000052

其中,

Figure BDA0001439141250000053
表示第i次迭代中,状态sj的度量值,j为译码器的状态数,j=1,2,3...2v,v是移位寄存器的个数。
Figure BDA0001439141250000054
表示最优路径的度量值。
Figure BDA0001439141250000055
表示执行一次译码后状态sj的度量值的净增量。in,
Figure BDA0001439141250000053
Indicates the metric value of the state s j in the ith iteration, j is the number of states of the decoder, j=1, 2, 3...2 v , and v is the number of shift registers.
Figure BDA0001439141250000054
A metric representing the optimal path.
Figure BDA0001439141250000055
represents the net increment of the metric value of state s j after performing one decoding.

Step3.执行维特比译码,通过加比选操作,步进每个序列位置各状态的度量值,循环一周后更新各个状态的度量值和似然路径,检测出最大度量值

Figure BDA0001439141250000056
和相应的咬尾路径
Figure BDA0001439141250000057
Step3. Perform Viterbi decoding, step through the metric value of each state of each sequence position through the addition selection operation, update the metric value and likelihood path of each state after one cycle, and detect the maximum metric value
Figure BDA0001439141250000056
and the corresponding tail-biting path
Figure BDA0001439141250000057

Viterbi算法采用接收码元条件概率的乘积的最大值作为估计序列,采用对数似然函数来表示为

Figure BDA0001439141250000058
其中,y表示码字序列经过传输映射的结果,r表示接收端得到的序列。简化上式中的对数函数求和运算,可以定义如下码元度量
Figure BDA0001439141250000059
其中,令The Viterbi algorithm uses the maximum value of the product of the conditional probabilities of the received symbols as the estimated sequence, and uses the log-likelihood function to express as
Figure BDA0001439141250000058
Among them, y represents the result of the codeword sequence after the transmission mapping, and r represents the sequence obtained by the receiving end. Simplifying the summation operation of the logarithmic function in the above formula, the following symbol metric can be defined
Figure BDA0001439141250000059
Among them, let

Figure BDA00014391412500000510
Figure BDA00014391412500000510

此种情况下,Viterbi算法中的码元度量值就是在编码网格上选择与接收序列r之间汉明距离最小的码字作为译码输出。In this case, the symbol metric value in the Viterbi algorithm is to select the codeword with the smallest Hamming distance from the received sequence r on the encoding grid as the decoding output.

Step4.执行第i次迭代译码,如果当前迭代得出的状态净增量最大值

Figure BDA0001439141250000061
大于存储的最大度量值
Figure BDA0001439141250000062
Figure BDA0001439141250000063
则更新最大度量值
Figure BDA0001439141250000064
和相应的最大咬尾路径
Figure BDA0001439141250000065
Step4. Execute the i-th iterative decoding, if the maximum value of the state net increment obtained by the current iteration
Figure BDA0001439141250000061
Greater than the maximum stored metric
Figure BDA0001439141250000062
which is
Figure BDA0001439141250000063
then update the maximum metric
Figure BDA0001439141250000064
and the corresponding maximum tail-biting path
Figure BDA0001439141250000065

Step5.统计第i次迭代各状态度量值净增量大于上次迭代状态净增量最大值的状态数量num(i),即

Figure BDA0001439141250000066
若num(i)<=num(i-1),则停止迭代,输出最大似然的咬尾路径
Figure BDA0001439141250000067
所对应的译码结果;否则,从净增量大于最大度量值的状态继续迭代译码,其它起始状态舍去。Step5. Count the number of states num(i) whose net increment of each state metric value of the i-th iteration is greater than the maximum value of the state net increment of the previous iteration, namely
Figure BDA0001439141250000066
If num(i)<=num(i-1), stop the iteration and output the maximum likelihood tail-biting path
Figure BDA0001439141250000067
The corresponding decoding result; otherwise, iterative decoding is continued from the state where the net increment is greater than the maximum metric value, and other initial states are discarded.

Step6.执行下一次迭代,重复Step3、4、5。Step6. Perform the next iteration and repeat Step3, 4, and 5.

咬尾卷积码是指用输入的信息序列的尾比特初始化寄存器,这样编码的初始状态和结束状态相同,通过减小冗余比特从而提高了编码效率,通常用八进制来表示编码器(n,k,m)的生成多项式,(n,k,m)表示输入k比特,输出n比特,移位寄存器个数为m。如图2所示,给出了(2,1,2)卷积码的网格图,生产多项式为(7,5),输入1比特生成2比特的编码信息,码率为1/2,格形图的每个位置有2m种状态,输入0时状态转移对应的是上支路,输入1时状态转移对应的是下支路,输出的编码结果如图2右边所示。Tail-biting convolutional code refers to initializing the register with the tail bits of the input information sequence, so that the initial state and end state of the encoding are the same, and the encoding efficiency is improved by reducing the redundant bits. The encoder is usually expressed in octal (n, k, m) generator polynomial, (n, k, m) represents input k bits, output n bits, and the number of shift registers is m. As shown in Figure 2, the trellis diagram of the (2,1,2) convolutional code is given, the production polynomial is (7,5), and 1 bit is input to generate 2 bits of encoded information, and the code rate is 1/2, Each position of the trellis diagram has 2 m states. When 0 is input, the state transition corresponds to the upper branch, and when 1 is input, the state transition corresponds to the lower branch. The output encoding result is shown on the right side of Figure 2.

实施例子:Implementation example:

S1.在译码开始的时候,对接收到的信息进行处理,由公式(1)计算得出可靠性最高的位置lopt作为译码的起始位置,即令位置lopt为译码开始位置l。接收到的符号为

Figure BDA0001439141250000068
计算得到似然比信息
Figure BDA0001439141250000069
S1. When the decoding starts, the received information is processed, and the position l opt with the highest reliability is calculated by the formula (1) as the starting position of the decoding, that is, the position l opt is the starting position l of the decoding. . The received symbol is
Figure BDA0001439141250000068
Calculate the likelihood ratio information
Figure BDA0001439141250000069

S2.当i=1时,即第一次译码,初始化所有状态的度量值

Figure BDA00014391412500000610
最优路径度量值
Figure BDA00014391412500000611
各状态的度量值净增量
Figure BDA00014391412500000612
状态集合{sj}包含所有状态。S2. When i=1, that is, the first decoding, initialize the metric values of all states
Figure BDA00014391412500000610
optimal path metric
Figure BDA00014391412500000611
Net metric increments for each state
Figure BDA00014391412500000612
The state set {s j } contains all states.

S3.开始执行维特比译码,通过步进加比选操作,每个位置的各状态选择度量值最大值所在的状态,并保存相应的路径信息,译码结束后,检查出度量值最大的咬尾路径,并将其度量值设为最优路径度量值,其相对应的路径设为最优咬尾路径,并将各状态的度量值设为下次迭代的初始度量值。S3. Start to perform Viterbi decoding. Through the step-by-step plus selection operation, each state of each position selects the state where the maximum metric value is located, and saves the corresponding path information. After the decoding is completed, check out the one with the largest metric value. The tail-biting path is set, and its metric value is set as the optimal path metric value, the corresponding path is set as the optimal tail-biting path, and the metric value of each state is set as the initial metric value of the next iteration.

S4.执行第i次迭代,译码结束后比较各状态的度量值

Figure BDA0001439141250000071
和存储的最大度量值
Figure BDA0001439141250000072
若存在
Figure BDA0001439141250000073
情况,则更新最大度量值,即
Figure BDA0001439141250000074
并将
Figure BDA0001439141250000075
的状态删除,跟新状态集合{sj},下次只从集合{sj}包含的状态开始迭代译码。S4. Execute the i-th iteration, and compare the metric values of each state after decoding
Figure BDA0001439141250000071
and the stored maximum metric
Figure BDA0001439141250000072
if exists
Figure BDA0001439141250000073
case, update the maximum metric value, that is
Figure BDA0001439141250000074
and will
Figure BDA0001439141250000075
The state of , is deleted, followed by the new state set {s j }, and the next time only the iterative decoding starts from the states contained in the set {s j }.

S5.每次迭代译码结束后,检测各状态度量值净增量大于上次迭代状态净增量最大值的状态数量num(i),即

Figure BDA0001439141250000076
的状态数。若num(i)>num(i-1),则继续进行迭代译码过程,否则,停止迭代,此时的最大度量值所在的状态即为最大咬尾路径的起始状态,根据相应的咬尾路径估计出译码序列。S5. After each iterative decoding is completed, detect the number of states num(i) whose net metric value of each state is greater than the maximum value of the state net increment of the previous iteration, namely
Figure BDA0001439141250000076
number of states. If num(i)>num(i-1), continue the iterative decoding process, otherwise, stop the iteration, the state where the maximum metric value is at this time is the starting state of the maximum tail biting path, according to the corresponding biting The tail path estimates the decoded sequence.

由于循环维特比译码可能在达到最大迭代次数前,就已经检测出了最优咬尾路径,之后的迭代都是不必要的,本发明所述的一种基于CVA的咬尾卷积码信道译码方法的迭代停止准则可以很好的避开这种循环陷阱,提高了译码效率。Because the cyclic Viterbi decoding may have detected the optimal tail-biting path before reaching the maximum number of iterations, the subsequent iterations are unnecessary. The iterative stop criterion of the decoding method can well avoid this loop trap and improve the decoding efficiency.

以上这些实施例应理解为仅用于说明本发明而不用于限制本发明的保护范围。在阅读了本发明的记载的内容之后,技术人员可以对本发明作各种改动或修改,这些等效变化和修饰同样落入本发明权利要求所限定的范围。The above embodiments should be understood as only for illustrating the present invention and not for limiting the protection scope of the present invention. After reading the contents of the description of the present invention, the skilled person can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims (1)

1. A tail-biting convolutional code channel decoding method based on CVA is characterized by comprising the following steps:
101. inputting a data stream, determining a decoding initial position with highest reliability according to likelihood ratio information of a received data stream information sequence, and performing modified Viterbi decoding from the initial position;
102. executing the ith iterative decoding, wherein 2 paths exist when the current state is reached, respectively calculating branch metric values among 2 branches reaching each state, selecting the optimal one, wherein the process is an adding and comparing operation, deleting the impossible state and starting the next iterative decoding through the adding and comparing process and updating decoding parameters;
103. after several times of iterative search, selecting the state of the maximum metric value as the initial position of the final decoding according to the iterative stop criterion, and estimating the decoding result according to the final initial position and the corresponding grid path;
step 101, when decoding starts, processes the received information and selects the position l with the highest reliabilityoptAs a starting position for decoding, a symbol r is receivedl (j)Calculating likelihood ratio information
Figure FDA0002306041780000011
Definition of loptIs calculated by the formula
Figure FDA0002306041780000012
In the formula (l + Q)LWhere Q is the quantity to be determined, and in a particular application will choose an appropriate value according to different codewords, when Q is the length of the received sequence, when the reliabilities at each position are exactly the same, then the decoding starts from the beginning;
said step 101 further comprises the step of initializing, i-0,
Figure FDA0002306041780000013
wherein,
Figure FDA0002306041780000014
representing the state s in the ith iterationjJ is the number of states of the decoder, j is 1,2,3vAnd v is the number of shift registers,
Figure FDA0002306041780000015
a metric value representing the optimal path is determined,
Figure FDA0002306041780000016
representing the state s after one iteration of decodingjA net increase in the measure of (a);
the step 102 specifically includes the steps of: through the addition and comparison operation, the metric value of each state of each sequence position is stepped, the metric value and the likelihood path of each state are updated after one cycle, and the maximum metric value is detected
Figure FDA0002306041780000021
And corresponding tail biting path
Figure FDA0002306041780000022
The Viterbi algorithm uses the maximum of the products of the conditional probabilities of the received symbols as the estimated sequence, expressed as a log-likelihood function
Figure FDA0002306041780000023
Wherein y represents the result of the transmission mapping of the codeword sequence, r represents the sequence obtained at the receiving end, and the summation operation of the logarithmic function in the above formula is simplified, and the following symbol metric can be defined
Figure FDA0002306041780000024
Wherein, it is made
Figure FDA0002306041780000025
b is-log epsilon, in which case the symbol metric in the Viterbi algorithm is the codeword selected on the coding trellis with the smallest hamming distance from the received sequence r;
the step 102 performs the ith iteration decoding, if the maximum value of the net increment of the state obtained by the current iteration is the maximum value
Figure FDA0002306041780000026
Greater than a stored maximum metric value
Figure FDA0002306041780000027
Namely, it is
Figure FDA0002306041780000028
The maximum metric value is updated
Figure FDA0002306041780000029
And corresponding maximum tail biting path
Figure FDA00023060417800000210
CN201710979858.0A 2017-10-19 2017-10-19 CVA-based tail-biting convolutional code channel decoding method Active CN107911195B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710979858.0A CN107911195B (en) 2017-10-19 2017-10-19 CVA-based tail-biting convolutional code channel decoding method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710979858.0A CN107911195B (en) 2017-10-19 2017-10-19 CVA-based tail-biting convolutional code channel decoding method

Publications (2)

Publication Number Publication Date
CN107911195A CN107911195A (en) 2018-04-13
CN107911195B true CN107911195B (en) 2020-03-17

Family

ID=61840656

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710979858.0A Active CN107911195B (en) 2017-10-19 2017-10-19 CVA-based tail-biting convolutional code channel decoding method

Country Status (1)

Country Link
CN (1) CN107911195B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110798231B (en) * 2018-08-02 2024-01-30 北京小米松果电子有限公司 Decoding method, device and storage medium of tail biting convolutional code
CN109861695B (en) * 2019-02-22 2023-06-20 北京芯盾集团有限公司 Method for decoding convolutional code by using codebook
CN110278055B (en) * 2019-06-03 2021-11-23 京信网络系统股份有限公司 Tail-biting convolutional coding processing method and device and communication equipment
CN111510160A (en) * 2020-05-13 2020-08-07 中国人民解放军军事科学院战争研究院 Truncation convolutional coding optimization construction method
CN112217609B (en) * 2020-10-14 2022-11-01 紫光展锐(重庆)科技有限公司 Communication decoding method, device, apparatus and storage medium
CN112290957B (en) * 2020-10-24 2023-06-09 西北工业大学 Orthogonal time-frequency expansion tail biting Turbo coding and decoding communication method
CN115514379A (en) * 2022-09-16 2022-12-23 深圳华海尖兵科技有限公司 Method and device for improving robustness of shortwave data transmission
CN119135190A (en) * 2024-11-14 2024-12-13 北京奥康银华科技有限公司 Energy-saving convolutional code decoding method based on path optimization

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5881075A (en) * 1996-03-18 1999-03-09 Samsung Electronics Co., Ltd. Viterbi decoder
CN1832390A (en) * 2005-03-07 2006-09-13 松下电器产业株式会社 A retransmission method based on reliability estimation for multi-antenna adaptive transmission
CN102638277A (en) * 2011-02-11 2012-08-15 联芯科技有限公司 Tail-biting convolutional code decoding method and device
CN102891690A (en) * 2011-07-19 2013-01-23 上海无线通信研究中心 Tail-biting convolution code decoding method
CN102904668A (en) * 2011-07-27 2013-01-30 杰脉通信技术(上海)有限公司 A fast PBCH decoding method for LTE

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5881075A (en) * 1996-03-18 1999-03-09 Samsung Electronics Co., Ltd. Viterbi decoder
CN1832390A (en) * 2005-03-07 2006-09-13 松下电器产业株式会社 A retransmission method based on reliability estimation for multi-antenna adaptive transmission
CN102638277A (en) * 2011-02-11 2012-08-15 联芯科技有限公司 Tail-biting convolutional code decoding method and device
CN102891690A (en) * 2011-07-19 2013-01-23 上海无线通信研究中心 Tail-biting convolution code decoding method
CN102904668A (en) * 2011-07-27 2013-01-30 杰脉通信技术(上海)有限公司 A fast PBCH decoding method for LTE

Also Published As

Publication number Publication date
CN107911195A (en) 2018-04-13

Similar Documents

Publication Publication Date Title
CN107911195B (en) CVA-based tail-biting convolutional code channel decoding method
CN106803759A (en) Polar yards of effective adaptive decoding method based on Gauss construction
EP2595321A1 (en) Tail-biting convolutional decoding apparatus and decoding method
CN106330207A (en) Joint Detection and Decoding Algorithm Based on Turbo-SCMA System
CN103354483B (en) General high-performance Radix-4SOVA decoder and interpretation method thereof
CN104579369A (en) Turbo iterative decoding method and device
CN111277277A (en) A method and device for reducing the decoding delay of polar code continuous cancellation table decoding algorithm
CN107659318B (en) Self-adaptive polar code decoding method
US10680749B2 (en) Early-termination of decoding convolutional codes
CN102891690B (en) Tail-biting convolution code decoding method
CN106254030A (en) The two-way coding and decoding method of the code of Spinal without speed
CN108134612B (en) An Iterative Decoding Method of Concatenated Codes for Correcting Synchronization and Substitution Errors
CN110730011A (en) A Recursive Block Markov Superposition Coding Method Based on Partial Superposition
WO2012163135A1 (en) Channel decoding method and decoder
CN103634015B (en) A Maximum Likelihood Decoding Algorithm for Tail-biting Codes
CN105375934A (en) Viterbi decoder aiming at tail-biting convolution code and decoding method
CN101257315A (en) Method of Iterative Decoding of Duobinary Turbo Code
CN108471341B (en) A method of convolutional encoding and decoding
CN113114278B (en) Duobinary Turbo decoding implementation method, system, equipment and application
CN108400788A (en) The hardware implementation method of Turbo decodings
CN103701475B (en) Decoding method for Turbo codes with word length of eight bits in mobile communication system
Zhu et al. An improved decoding of tail-biting convolutional codes for LTE systems
CN107342775B (en) Viterbi decoding method for punctured convolutional code
TWI569584B (en) Decoding methods using dynamic scaling factor
CN113258937B (en) Component decoder, extrinsic information storage unit, and Turbo code decoder

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant