A marking code transmission method using data punching

Abstract

The present invention discloses a marking code transmission method using data puncturing, comprising: a binary information sequence u is passed through a binary LDPC code encoder to generate a length N L The coding sequence

Description

Technical Field

[0001] The present invention relates to the field of digital communication error control coding, and in particular to a marking code transmission method using data puncturing. Background Art

[0002] In order to solve the synchronization error caused by the unstable sampling clock rate of the communication system, that is, the insertion / deletion error, the researchers designed a synchronization error correction code. Insertion / deletion and substitution errors will lead to catastrophic consequences, that is, the length of the transmitted and received sequences is inconsistent and there are a large number of sudden substitution errors, which will reduce the system performance; at the same time, the previous error control coding technology designed for memoryless channels and substitution errors cannot be directly adopted.

[0003] The traditional concatenated coding scheme proposed by Davey and MacKay, which can correct multiple insertion / deletion and substitution errors, is referred to as DM code for short. It uses watermark-LDPC (low-density parity check) concatenated code to correct insertion / deletion errors, and has excellent performance. In order to further improve the error correction capability of the DM scheme and the flexibility of the system, researchers proposed a marker code transmission method based on weighted edit distance (WED). This method inserts the marker code as the inner code evenly into the outer code binary LDPC code, and proposes a symbol-level synchronization algorithm to restore the length of the received sequence. This method has achieved excellent error correction capability, and the decoding module has strong portability. It is not limited to the concatenation method between the inner and outer codes, and can be flexibly used in marker code or watermark code systems. However, the WED-based marker code transmission method still needs to be improved in terms of data transmission rate.

[0004] Therefore, how to design an insertion / deletion error coding algorithm with a higher transmission rate is a problem that needs long-term research in this field. Summary of the invention

[0005] The present invention provides a method for transmitting a marking code using data puncturing. The present invention improves the transmission rate of the system while ensuring that the transmission performance of the system is not lost. The details are described below:

[0006] A marking code transmission method using data puncturing, the method comprising:

[0007] (1) The binary information sequence u is generated by the outer code, that is, the encoder of the binary LDPC code, with a length of N. L The coding sequence f ;

[0008] (2) Insert the marker code ω uniformly into the coding sequence f In the example, the generated length is N c The concatenated code g is obtained and the concatenated code g is output;

[0009] (3) Puncturing the σ-th bit in each symbol of the concatenated code g to generate a transmitted sequence g′;

[0010] (4) The transmitted sequence g′ is passed through an insertion / deletion-replacement channel to generate a received sequence h;

[0011] (5) Correct the insertion / deletion errors in the received sequence h through the WED-based symbol-level synchronization restorer, assign β to the likelihood information at the puncturing position, and finally output the likelihood ratio sequence l;

[0012] (6) Initialize the binary LDPC decoder with the likelihood ratio sequence l, decode the binary LDPC decoder using the logarithmic domain belief propagation decoding algorithm, and output an estimate of the information sequence

[0013] The marker code ω is uniformly inserted into the coding sequence. f In the example, the generated length is N c The concatenated code g is output as:

[0014] Binary LDPC Code f Divided into N symbols, each symbol has m bits, where N = N L / m; the internal encoder randomly generates a marker code ω of length N×λ, divides the marker code ω into N subsequences of length λ, and inserts each subsequence of the marker code ω into the binary LDPC code f Before each symbol of c The concatenated code g contains λ watermark bits and m LDPC code bits.

[0015] Furthermore, the σ-th bit in each symbol in the concatenated code g is punctured, and the transmission sequence g′ generated by the puncturing is:

[0016] (3.1) Let the symbol index j = 0, and a frame of concatenated code g contains N symbols;

[0017] (3.2) Let the bit subscript i in each symbol = 0;

[0018] (3.3) Determine whether i is equal to σ; when i≠σ, continue to determine whether i<σ. If so, g′ j(m-1)+i =g jm+i , execute step (3.4); otherwise i>σ, g′ j(m-1)+i-1 =g jm+i , i=i+1, execute step (3.4);

[0019] (3.4) Determine whether i=σ, if yes, i=i+1, and execute step (3.5); if no, repeat step (3.3);

[0020] (3.5) Determine whether i<m. If yes, return to step (3.3); if no, execute step (3.6);

[0021] (3.6) j = j + 1;

[0022] (3.7) Determine whether j<N. If so, return to step (3.2); if not, output the sending sequence g′ and the process ends.

[0023] In which, the likelihood information at the puncturing position is assigned β, and the final output likelihood ratio sequence l is:

[0024] (5.1) Let bit subscript k = 0, j = 0;

[0025] (5.2) If k%m=σ, then the log-likelihood ratio of the kth position is l k =β, where σ is the punching position; otherwise, l k The calculation of is as follows:

[0026]

[0027] Among them, 0<a≤q-1, 0≤k<N L , 0≤j<N, represents the jth symbol, f k represents the kth transmitted bit;

[0028]

[0029] in, Indicates that in f k =1 and The probability of receiving h under the condition of i represents the state drift at position i, F(·), B(·) and M(·) represent the forward, backward and intermediate metrics respectively;

[0030] (5.3) j = j + 1, k = k + 1;

[0031] (5.4) If k<N L , then return to step (5.2); otherwise, output the likelihood ratio sequence l and the process ends.

[0032] The beneficial effects of the technical solution provided by the present invention are:

[0033] 1. Based on the traditional marking code transmission method, the present invention adopts a data puncturing method, which reduces the number of bits required for transmission, improves transmission efficiency, and has almost no performance loss;

[0034] 2. The present invention shortens the length of the transmitted sequence code, reduces the system decoding delay and decoding calculation complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A system block diagram of a marking code transmission method using data puncturing;

[0036] Figure 2 Schematic diagram of the frame structure of the sending sequence;

[0037] Figure 3 Flowcharts to punch holes for data;

[0038] Figure 4 Flowchart for calculating likelihood ratio sequence;

[0039] Figure 5 A schematic diagram of the cumulative number of errors in received symbols;

[0040] Figure 6 Schematic diagram of performance simulation before and after drilling. DETAILED DESCRIPTION

[0041] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention are described in further detail below.

[0042] Through system simulation, the embodiment of the present invention found that in the WED-based marker code transmission method, the probability of errors in bits near specific positions of the marker code is significantly higher than that of the remaining bits. Therefore, the embodiment of the present invention adopts data puncturing technology to delete information bits that are more prone to errors at the transmitting end, and only transmit the remaining bits, so as to achieve the purpose of transmitting information with fewer bits. This method is expected to increase the transmission bit rate while ensuring that the system error correction performance is not lost.

[0043] Compared with the traditional scheme, the embodiment of the present invention has made the following modifications: for the WED-based marker code transmission scheme, the transmission sequence is punctured, the bits on the error-prone information bits are deleted, and only the unpunctured bits are transmitted. At the receiving end, the inner decoder uses the WED-based symbol-level synchronization restorer to compare with the known marker code to restore synchronization, further assigns the likelihood information of the punctured position, and then outputs the likelihood ratio sequence for the LDPC code decoder to correct the residual errors. Compared with the transmission method that does not adopt this technology, this method reduces the number of bits required to be transmitted, and transmits information by sending fewer bits to achieve the same effect, thereby improving the transmission rate without affecting the system error correction performance.

[0044] The embodiment of the present invention proposes a marker code transmission method using data puncturing. On the basis of the traditional marker code transmission method, data puncturing is performed on the transmitted information before entering the insertion / deletion-replacement channel, thereby removing information bits in the signal that are more prone to errors. The same information is transmitted by sending fewer bits. Therefore, this method effectively improves the transmission rate of the system and ensures that the performance is not deteriorated.

[0045] A marking code transmission method using data puncturing provided by an embodiment of the present invention is described in detail below in conjunction with the accompanying drawings.

[0046] like Figure 1 As shown, the embodiment of the present invention includes the following six steps:

[0047] (1) The binary information sequence u is generated by the outer code, that is, the encoder of the binary LDPC code, with a length of N. L The coding sequence f ;

[0048] (2) Insert the marker code ω uniformly into the coding sequence f In the example, the generated length is N c The concatenated code g is obtained and the concatenated code g is output;

[0049] (3) Puncturing the σ-th bit in each symbol of the concatenated code g to generate a transmitted sequence g′;

[0050] (4) The transmitted sequence g′ is passed through an insertion / deletion-replacement channel to generate a received sequence h;

[0051] (5) Correct the insertion / deletion errors in the received sequence h through the WED-based symbol-level synchronization restorer, assign β to the likelihood information at the puncturing position, and finally output the likelihood ratio sequence l;

[0052] (6) Initialize the binary LDPC decoder with the likelihood ratio sequence l, decode the binary LDPC decoder using the logarithmic domain belief propagation decoding algorithm, and output an estimate of the information sequence

[0053] The following are the specific implementation steps of the above six steps, see the following description for details:

[0054] The encoder of the binary LDPC code in step (1) is a professional term in this field and will not be described in detail in the embodiment of the present invention.

[0055] like Figure 2 As shown, step (2) inserts the marker code ω uniformly into the coding sequence f In the example, the generated length is N c The concatenated code g is output, and the output concatenated code g includes:

[0056] (2.1) Binary LDPC code f Divided into N symbols, each symbol has m bits, where N = N L / m;

[0057] (2.2) The internal encoder randomly generates a marker code ω of length N×λ, divides the marker code ω into N subsequences of length λ, and inserts each subsequence of the marker code ω into the binary LDPC code f Before each symbol of c The concatenated code g is a scalar code, where each symbol contains λ watermark bits and m LDPC code bits.

[0058] like Figure 3 As shown, step (3) punctures the σ-th bit in each symbol in the concatenated code g, and the puncturing generates a transmission sequence g′ including:

[0059] (3.1) Let the symbol index j = 0, and a frame of concatenated code g contains N symbols;

[0060] (3.2) Let the bit subscript i in each symbol = 0;

[0061] (3.3) Determine whether i is equal to σ; when i≠σ, continue to determine whether i<σ. If so, g′ j(m-1)+i =g jm+i , execute step (3.4) otherwise i>σ, g′ j(m-1)+i-1 =g jm+i , i=i+1, execute step (3.4);

[0062] (3.4) Determine whether i=σ, if yes, i=i+1, and execute step (3.5); if no, repeat step (3.3);

[0063] (3.5) Determine whether i<m. If yes, return to step (3.3); if no, execute step (3.6);

[0064] (3.6) j = j + 1;

[0065] (3.7) Determine whether j<N. If so, return to step (3.2); if not, output the sending sequence g′ and the process ends.

[0066] Step (4) sending sequence g′ to generate receiving sequence h through insertion / deletion-replacement channel includes:

[0067] The sequence g′ is sent through an insertion / deletion-replacement channel with parameter P i , P d and P s represent the probability of insertion, deletion and substitution respectively. Transmission probability Pt =1-P i -P d .

[0068] The above-mentioned insertion / deletion-replacement channel is a professional term in this field, and will not be elaborated in detail in the embodiment of the present invention.

[0069] like Figure 4 As shown, step (5) corrects the insertion / deletion errors in the received sequence h through a symbol-level synchronization restorer based on WED, and assigns β to the likelihood information at the puncturing position, and finally outputs the likelihood ratio sequence l including:

[0070] (5.1) Let bit subscript k = 0, symbol subscript j = 0;

[0071] (5.2) If k%m=σ, then the log-likelihood ratio of the kth position is l k =β, where σ is the punching position; otherwise, l k The calculation of is as follows:

[0072]

[0073] Among them, 0<a≤q-1, 0≤k<N L , 0≤j<N, represents the jth symbol, f k represents the kth transmitted bit.

[0074]

[0075] in, Indicates that in f k =1 and The probability of receiving h under the condition of i represents the state drift at position i, and F(·), B(·) and M(·) represent the forward, backward and intermediate metrics, respectively.

[0076] (5.3) j = j + 1, k = k + 1;

[0077] (5.4) If k<N L , then return to step (5.2); otherwise, output the likelihood ratio sequence l and the process ends.

[0078] Step (6) The binary LDPC decoder uses the likelihood ratio sequence l to perform iterative decoding and outputs an estimate of the information sequence The steps include:

[0079] Send the likelihood ratio sequence l to the binary LDPC decoder; decode it using the logarithmic domain belief propagation decoding algorithm of the binary LDPC code; repeat the above decoding steps until the preset maximum number of iterations δ is reachedmax The logarithmic domain belief propagation decoding algorithm of binary LDPC codes is a professional term in this field, and the embodiments of the present invention will not be described in detail. Specific embodiments

[0081] The embodiment of the present invention compares the code without data puncturing and the code after data puncturing, and selects the code length N c The concatenated code with a code rate of 576 bits and a code rate of 0.5 is a special case. A tag code transmission method using data puncturing is introduced. In the simulation, a pseudo-random sequence is used as the inner tag code and a binary LDPC code is used as the outer code, where λ = 1, N L =720, R L =0.5, the maximum number of insertion errors per bit in the channel is 5, P i =P d ,The decoder of binary LDPC codes adopts logarithmic domain belief propagation decoding algorithm, and the maximum number of iterations is 50.

[0082] A binary pseudo-random sequence is used, where λ = 1, P d =P i The maximum number of insertion errors per bit in the channel is 5. When m = 4 and m = 6, the puncture probability P d =0.01, the probability of each information bit being wrong. Figure 5 is the error condition of each information bit in the transmission sequence, such as Figure 5 As shown, the σ bits in each symbol of the transmission codeword have a higher error probability than other bits, σ=2.

[0083] The specific method for calculating the bit rate after puncturing is:

[0084]

[0085] Among them, select N L =576, R L = 0.5 as the outer code, D C Indicates the number of bits deleted in a frame, D I Indicates the number of bits deleted in the information sequence.

[0086] Figure 6 The curves of the block error rate of the system under different replacement probabilities and the insertion / deletion probability are given, where the block error rate is equal to the number of error frames divided by the number of sent frames. It can be seen that the performance of the system has almost no deterioration before and after data puncturing.

[0087] Unless otherwise specified, the models of the components in the embodiments of the present invention are not limited, and any device that can perform the above functions may be used.

[0088] Those skilled in the art will appreciate that the accompanying drawing is only a schematic diagram of a preferred embodiment, and the serial numbers of the embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0089] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A marking code transmission method using data puncturing, characterized in that: The method includes: The binary information sequence u is passed through the binary LDPC code encoder to generate a length of N L The coding sequence f ; Insert the marker code ω uniformly into the coding sequence f In the example, the generated length is N c The concatenated code g is obtained and the concatenated code g is output; Puncturing the σ-th bit in each symbol of the concatenated code g, and the puncturing generates a transmission sequence g'; the transmission sequence g' passes through an insertion / deletion-replacement channel to generate a received sequence h; Through a symbol-level synchronization recovery device based on WED, correcting the insertion / deletion errors in the received sequence h, and assigning a likelihood information β to the puncturing position, and finally outputting a likelihood ratio sequence l; The binary LDPC decoder is initialized by the likelihood ratio sequence l, and the binary LDPC decoder is decoded using the logarithmic domain belief propagation decoding algorithm to output an estimate of the information sequence Wherein, the puncturing the σ-th bit in each symbol of the concatenated code g, and the puncturing generates a transmission sequence g' is: (3.1) Let the symbol subscript j = 0, and there are N symbols in one frame of the concatenated code g; (3.2) Let the bit subscript i = 0 in each symbol; (3.3) Determine whether i is equal to σ; when i≠σ, continue to determine whether i<σ. If so, g′ j(m-1)+i =g jm+i , execute step (3.4) otherwise i>σ, g′ j(m-1)+i-1 =g jm+i , i=i+1, execute step (3.4); (3.4) Determine whether i = σ, if so, i = i + 1, and execute step (3.5); if not, repeat step (3.3); (3.5) Determine whether i < m, if so, return to step (3.3); if not, execute step (3.6); (3.6) j = j + 1; (3.7) Determine whether j < N, if so, return to step (3.2); if not, output the transmission sequence g', and the process ends; The assigning a likelihood information β to the puncturing position, and finally outputting a likelihood ratio sequence l is: (5.1) Let the bit subscript k = 0, j = 0; (5.2) If k%m=σ, then the log-likelihood ratio of the kth position is l k =β, where σ is the punching position; otherwise, l k The calculation of is as follows: Among them, 0 <a≤q-1,0≤k<N L , 0≤j <N, represents the jth symbol, f k represents the kth transmitted bit; in, Indicates that in f k =1 and The probability of receiving h under the condition of i represents the state drift at position i, F(·), B(·) and M(·) represent the forward, backward and intermediate metrics respectively; (5.3) j = j + 1, k = k + 1; (5.4) If k <N L , then return to step (5.2); otherwise, output the likelihood ratio sequence l and the process ends.

2. A method for transmitting a marking code using data puncturing according to claim 1, characterized in that: The marker code ω is uniformly inserted into the coding sequence f In the example, the generated length is N c The concatenated code g is output as: Binary LDPC Code f Divided into N symbols, each symbol has m bits, where N = N L / m; The internal encoder randomly generates a marker code ω of length N×λ, divides the marker code ω into N subsequences of length λ, and inserts each subsequence of the marker code ω into the binary LDPC code f Before each symbol of c The concatenated code g contains λ watermark bits and m LDPC code bits.

Images