A Multi-stage PLDPC-CSK Design Method Based on Symbol Sequence

Abstract

The invention discloses a multi-level PLDPC-CSK design method based on symbol sequence, which is applicable to the field of communication. Firstly, based on the multi-level PLDPC code and modulation constellation structure in the color shift keying visible light communication system, a multi-level PLDPC-CSK transmission scheme is designed; secondly, based on the perceived color constraint and the three-dimensional light intensity space CSK constellation, a method for sending continuous symbol sequences is introduced, and a CSK mapping scheme that minimizes the number of symbol sequences at the minimum Euclidean distance is proposed; finally, based on the multi-level PLDPC-CSK transmission system, a multi-level iterative soft decoding scheme is designed, and the receiving end is based on the channel output and prior information, and the CSK demodulator outputs a posteriori information, which is decoded as the prior information of the PLDPC decoder; the decoder outputs external information as update information and feeds it back to the demodulator for the next level of iterative decoding; the transmission method designed by the invention effectively improves the transmission quality of the visible light communication system.

Description

Technical Field

[0001] The invention belongs to the field of color shift keying coding modulation in visible light communication, and in particular relates to a multi-level PLDPC code and iterative decoding transmission method based on color shift keying. Background Art

[0002] Visible light communication technology is a communication method that uses visible light as a carrier to transmit data in a wireless channel. As an emerging technology that takes into account both lighting and communication, it has a wider spectrum and higher transmission rate, which can effectively improve the "spectrum shortage" problem faced by traditional radio frequency communications. It is the technical core that will replace radio frequency technology in the future to develop optical networking, 6G communication networks and other applications. It has important research value and broad development prospects. Visible light communication has a profound research background and urgent demand, and coding modulation, as a key technology, is of great significance for achieving high-speed and reliable visible light communication.

[0003] Color shift keying is a commonly used modulation method in visible light communication. It uses the color characteristics of visible light to transmit data, achieving high-order modulation, higher transmission rate and constant luminous intensity, thus effectively avoiding the problem of intensity flicker. The traditional CSK constellation modulation coding has a high bit error rate, and the error correction performance of the visible light communication system needs to be further improved. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a multi-level PLDPC-CSK design method based on symbol sequence, which further improves the performance gain through cascade transmission of multi-level PLDPC code and CSK modulation, reduces the bit error rate of visible light communication transmission, and improves the reliability of transmission.

[0005] Technical solution: A multi-level PLDPC-CSK design method based on symbol sequence of the present invention, the data bit sequence sent by the transmitting end is first divided into data blocks, the data blocks are PLDPC encoded and CSK modulated to obtain the transmitted symbols, and the receiving end, the signal received from the VLC channel is decoded by multi-level iterative decoding and then outputs the data bit sequence, including the following steps:

[0006] Step 1: construct a non-power-of-two CSK constellation of color intensity joint keying in a 3D light intensity space, obtain the coordinates of the CSK constellation, and form a 10-CSK constellation symbol set;

[0007] Step 2: Construct a 5-CSK modulation scheme with a data rate of R = 2 bits / symbol;

[0008] Step 3: Construct a 10-CSK modulation scheme with a data rate of R = 3 bits / symbol;

[0009] Step 4: The transmitter uses PLDPC code with a code rate of 4 / 5 and a code length of 96 for encoding, and transmits it to the channel after 5-CSK modulation and 10-CSK modulation respectively; the receiver introduces multi-level iterative soft decoding (IMSD) between the demodulator and the decoder. Based on the channel output and prior information, the CSK demodulator outputs a posteriori information, which is decoded as the prior information of the PLDPC decoder; the decoder outputs external information as update information and feeds it back to the demodulator for the next level of iterative decoding.

[0010] Further, the step 1 comprises:

[0011] By dividing the 2D constant intensity plane triangular CSK constellation by triangles, it is easy to obtain a symmetric 10-CSK constellation that meets the perceived color constraint. The 10-CSK constellation is extended from the 2D plane to the 3D multi-intensity space, and the 6 constellation points in the maximum intensity plane are optimized by maximizing the minimum Euclidean distance to obtain a 3D space 10-CSK constellation that meets the target perceived color constraint.

[0012] The target perceived color is defined as: E[Λ n ]=T c ;

[0013] Where, T c Indicates the intensity symbol corresponding to the target perceived color, T c =[1 / 3,1 / 3,1 / 3] T .

[0014] Further, the step 2 comprises:

[0015] Based on the selection criterion of the minimum number of symbol sequences at the MED, it can be described as:

[0016] min{N S″}

[0017]

[0018] d min =d S′,min =d S″,min

[0019] |S″|=64

[0020] Where, d S′,min With d S″,min Respectively represent the MED between sequence symbols in sequence sets S′ and S″, both equal to d min ; N S″ represents the number of sequence symbols at the MED based on the sequence set S″, where S″ is the selected symbol sequence set, that is, |S″|=16.

[0021] According to the selection criteria of the symbol sequence, the symbol sequence set S″ is obtained as:

[0022]

[0023] Through the analysis of S″, we can know that N S′ =4; and then the data rate R=2 bits / symbol 5-CSK modulation scheme is obtained.

[0024] Further, the step 3 comprises:

[0025] When L = 2, the corresponding symbol sequence set S′ = {s0s0, s0s1, …, s i s j ,…,s9s9}, there are many possible situations for selecting 64 groups of symbol sequences from S′ under the constraint of target perceived color; based on the fact that the MED between symbol sequences remains unchanged, a selection criterion with the minimum number of symbol sequences at the MED is proposed, which is updated as follows:

[0026] min{N S″}

[0027]

[0028] d min =d S′,min =d S″,min

[0029] |S″|=64

[0030] According to the selection criteria of the symbol sequence, the symbol sequence set S″ is obtained as:

[0031]

[0032] Through the analysis of S″, we can know that N S″ =6; and then the data rate R=3 bits / symbol 10-CSK modulation scheme is obtained.

[0033] Further, the step 4 comprises:

[0034] Under known AWGN channel conditions, the soft information output by the demapping module at the receiving end is described in the form of log-likelihood ratio. Then, the i-th data information of a transmitted symbol sequence ss is The log-likelihood ratio is:

[0035]

[0036] In the formula, and represents the subset of symbol sequences whose i-th bit of data information is 0 and 1 respectively. After simplification, we get:

[0037]

[0038] And meet

[0039] For multi-level PLDPC codes, demodulation and decoding are performed step by step, and the l-1 level PLDPC decoding output is used for l level demodulation and PLDPC decoding; the demodulator outputs soft information and transmits it to the decoder as its prior information:

[0040]

[0041] Set M c is the set of check nodes, N v A collection of variable nodes. Indicates that the variable node VN in the hth BP iterative decoding of the PLDPC code i Sent to verification node CN j Soft information; Indicates that in the hth iteration, the check node CN j Send to variable node VN i Soft information; It means that after the hth iteration, decoder l obtains the codeword of soft information.

[0042] Based on the above description, the main steps of the decoder BP decoding algorithm are as follows:

[0043] (1) Initialization,

[0044] (2) Update information obtained by the check node from the variable node:

[0045] (3) Update information obtained by variable nodes from check nodes:

[0046]

[0047] (4) The soft information of the codeword can be obtained based on the updated information of the check nodes and variable nodes:

[0048]

[0049] If the PLDPC decoder l decodes correctly, it stops decoding and outputs decision c (l) ; Otherwise, update the number of iterations h and go to step (2); if h>IT, then stop iterative decoding and output the decision c (l) .

[0050] In summary, we get the m-level PLDPC coding bit for:

[0051] Among them, the external information output by decoder 1 is:

[0052] In the formula Feedback to the demodulator as its prior information:

[0053] Multi-level iterative soft decoding reaches the maximum number of iterations N MAX When , decoding ends.

[0054] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0055] The present invention is based on a 3D multi-intensity plane CSK constellation and proposes a CSK scheme with L=2 continuous symbol sequences. The scheme improves the error correction performance by reducing the number of symbol sequences at the minimum Euclidean distance. Compared with the traditional CSK constellation structure, efficient power transmission is achieved. Based on this constellation structure, a multi-level PLDPC code and iterative decoding transmission method based on CSK is designed. The CSK demodulator uses the Log-Map algorithm to soft-demodulate the received symbol sequence, and the PLDPC decoder uses the BP algorithm for decoding. The multi-level iterative soft decoding introduced between the demodulator and the decoder effectively improves the BER performance of signal transmission. When the BER is 10 -6 At this point, the performance of BICM-ID-5CSK is about 1.3dB better than that of BICM-ID-4CSK, and the performance of BICM-ID-10CSK is about 6.1dB better than that of BICM-ID-8CSK. Therefore, at the same symbol data rate, the performance of the present invention is better than that of the standard PLDPC-CSK. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The transmission block diagram of the VLC system based on multi-level PLDPC code and CSK modulation cascade (PLDCP-CSK);

[0057] Figure 2 10-CSK constellation structure diagram in the embodiment: (a) 2D plane projection of the 10-CSK constellation; (b) 3D space projection of the 10-CSK constellation;

[0058] Figure 3 is a comparison diagram of BER curves when R=2 bits / symbol in the embodiment;

[0059] Figure 4 BER curve comparison diagram when R=3 bits / symbol in the embodiment. DETAILED DESCRIPTION

[0060] In order to realize the design method of the present invention and demonstrate the technical features in more detail, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation plans.

[0061] like Figure 1 As shown in FIG. 1 , the transmission block diagram of the visible light communication system designed by the present invention based on multi-level PLDPC code and CSK modulation cascade (PLDCP-CSK) transmission. At the transmitting end, the data bit sequence μ is first divided into m data blocks, and the length of the data block is k. After PLDPC encoding, the output codeword is of length n l = 1, ..., m; secondly, m multi-level codewords output m information bits to form the i-th array, which is mapped into symbol s after CSK modulation. i =(s r,i ,s g,i ,s b,i ) and transmit it to the channel. and Denote the mapping function and demapping function respectively, then and The receiving end receives the signal r from the VLC channel i For: r i =Hs i +n.

[0062] Where H is a 3×3 channel gain matrix; n is a 3×1 additive white Gaussian noise vector with a mean of 0 and a variance of σ 2 The receiver introduces iterative multilevel soft decoding (IMSD) between the demodulator and the decoder. The CSK demodulator uses the Log-Map algorithm to soft-demodulate the received symbol sequence and outputs As the prior information of decoder l; decoder l uses BP algorithm to decode and output Multi-level After parallel-to-serial conversion, the data information sequence is output The specific steps are as follows:

[0063] Step 1: construct a non-power-of-two CSK constellation of color intensity joint keying in a 3D light intensity space, obtain the coordinates of the CSK constellation, and form a 10-CSK constellation symbol set;

[0064] In one example, the above step 1 specifically includes:

[0065] By dividing the 2D constant intensity plane triangular CSK constellation by triangles, it is easy to obtain a symmetric 10-CSK constellation that meets the perceived color constraint. The 10-CSK constellation is extended from the 2D plane to the 3D multi-intensity space, and the 6 constellation points in the maximum intensity plane are optimized by maximizing the minimum Euclidean distance to obtain a 3D space 10-CSK constellation that meets the target perceived color constraint.

[0066] The target perceived color is defined as: E[Λ n ]=T c ;

[0067] Where, T c Indicates the intensity symbol corresponding to the target perceived color, T c =[1 / 3,1 / 3,1 / 3] T .

[0068] Step 2: Construct a 5-CSK modulation scheme with a data rate of R = 2 bits / symbol;

[0069] In one example, the above step 2 specifically includes:

[0070] Based on the selection criterion of the minimum number of symbol sequences at the MED, it can be described as:

[0071] min{N S″}

[0072]

[0073] d min =d S′,min =d S″,min

[0074] |S″|=64

[0075] Where, d S′,min With d S″,min Respectively represent the MED between sequence symbols in sequence sets S′ and S″, both equal to d min ; N S″ represents the number of sequence symbols at the MED based on the sequence set S″, where S″ is the selected symbol sequence set, that is, |S″|=16.

[0076] According to the selection criteria of the symbol sequence, the symbol sequence set S″ is obtained as:

[0077]

[0078] Through the analysis of S″, we can know that N S′ =4; and then the data rate R=2 bits / symbol 5-CSK modulation scheme is obtained.

[0079] Step 3: Construct a 10-CSK modulation scheme with a data rate of R = 3 bits / symbol;

[0080] In one example, the above step 3 specifically includes:

[0081] When L = 2, the corresponding symbol sequence set S′ = {s0s0, s0s1, …, s i s j ,…,s9s9}, there are many possible situations for selecting 64 groups of symbol sequences from S′ under the constraint of target perceived color; based on the fact that the MED between symbol sequences remains unchanged, a selection criterion with the minimum number of symbol sequences at the MED is proposed, which is updated as follows:

[0082] min{N S″}

[0083]

[0084] d min =d S′,min =d S″,min

[0085] |S″|=64

[0086] According to the selection criteria of the symbol sequence, the symbol sequence set S″ is obtained as:

[0087] Through the analysis of S″, we can know that N S″ =6; and then the data rate R=3 bits / symbol 10-CSK modulation scheme is obtained.

[0088] Step 4: The transmitter uses PLDPC code with a code rate of 4 / 5 and a code length of 96 for encoding, and transmits it to the channel after 5-CSK modulation and 10-CSK modulation respectively; the receiver introduces multi-level iterative soft decoding (IMSD) between the demodulator and the decoder. Based on the channel output and prior information, the CSK demodulator outputs a posteriori information, which is decoded as the prior information of the PLDPC decoder; the decoder outputs external information as update information and feeds it back to the demodulator for the next level of iterative decoding.

[0089] In one example, the above step 4 specifically includes:

[0090] Under known AWGN channel conditions, the soft information output by the demapping module at the receiving end is described in the form of log-likelihood ratio. Then, the i-th data information of a transmitted symbol sequence ss is The log-likelihood ratio is:

[0091]

[0092] In the formula, and represents the subset of symbol sequences whose i-th bit of data information is 0 and 1 respectively. After simplification, we get:

[0093]

[0094] And meet

[0095] For multi-level PLDPC codes, demodulation and decoding are performed step by step, and the l-1 level PLDPC decoding output is used for l level demodulation and PLDPC decoding; the demodulator outputs soft information and transmits it to the decoder as its prior information:

[0096] Set M c is the set of check nodes, N v A collection of variable nodes. Indicates that the variable node VN in the hth BP iterative decoding of the PLDPC code i Sent to verification node CN j Soft information; Indicates that in the hth iteration, the check node CN j Send to variable node VN i Soft information; It means that after the hth iteration, decoder l obtains the codeword of soft information.

[0097] Based on the above description, the main steps of the decoder BP decoding algorithm are as follows:

[0098] (1) Initialization,

[0099] (2) Update information obtained by the check node from the variable node:

[0100] (3) Update information obtained by variable nodes from check nodes:

[0101]

[0102] (4) The soft information of the codeword can be obtained based on the updated information of the check nodes and variable nodes:

[0103]

[0104] If the PLDPC decoder l decodes correctly, it stops decoding and outputs decision c (l) ; Otherwise, update the number of iterations h and go to step (2); if h>IT, then stop iterative decoding and output the decision c (l) .

[0105] In summary, we get the m-level PLDPC coding bit for:

[0106] Among them, the external information output by decoder 1 is:

[0107] In the formula Feedback to the demodulator as its prior information:

[0108] Multi-level iterative soft decoding reaches the maximum number of iterations N MAX When , decoding ends.

[0109] based on Figure 1 The block diagram of the PLDPC-CSK multi-stage iterative decoding system is shown, and its BER performance is simulated and verified. Figure 3 The comparison BER curves of the two schemes with R = 2 bits / symbol are given. It can be seen from the figure that compared with the iterative bit interleaved coded modulation and iterative decoding (BICM-ID) and the IMSD scheme designed in this paper, the latter has obvious BER performance advantages. Specifically, when the BER is 10 -6 At N, the performance of BICM-ID-5CSK is about 1.3dB better than that of BICM-ID-4CSK. MAX =1, N MAX =2, N MAX =5, N MAX =10, the performance gain of IMSD-5CSK is better than that of IMSD-4CSK by about 1.8dB, 1.85dB, 1.75dB and 1.8dB respectively.

[0110] Figure 4 The comparative BER curves of the two schemes with R = 3 bits / symbol are given. It can be seen from the figure that the IMSD-10CSK scheme has a BER performance advantage. Specifically, when the BER is 10 -6 At N, the performance of BICM-ID-10CSK is about 6.1dB better than that of BICM-ID-8CSK. MAX =1, N MAX =2, N MAX =5, N MAX =10, the performance gain of IMSD-10CSK is better than that of IMSD-8CSK by about 4.3dB, 4.5dB, 4.6dB and 4.7dB respectively.

Claims

1. A multi-stage PLDPC-CSK design method based on symbol sequence, characterized in that: The steps include: Step 1: construct a non-power-of-two CSK constellation of color intensity joint keying in a 3D light intensity space, obtain the coordinates of the CSK constellation, and form a 10-CSK constellation symbol set; Step 2: Construct a 5-CSK modulation scheme with a data rate of R = 2 bits / symbol; Step 3: Construct a 10-CSK modulation scheme with a data rate of R = 3 bits / symbol; Step 4: The transmitter uses a PLDPC code with a code rate of 4 / 5 and a code length of 96 for encoding, and transmits it to the channel after 5-CSK modulation and 10-CSK modulation respectively; the receiver introduces multi-level iterative soft decoding IMSD between the demodulator and the decoder. Based on the channel output and prior information, the CSK demodulator outputs a posteriori information, which is decoded as the prior information of the PLDPC decoder; The decoder outputs external information as update information and feeds it back to the demodulator for the next level of iterative decoding; Step 1 includes: The 2D constant intensity plane triangle CSK constellation is divided by triangles to obtain a symmetric 10-CSK constellation that meets the perceived color constraint. The 10-CSK constellation is extended from the 2D plane to the 3D multi-intensity space, and the 6 constellation points in the maximum intensity plane are optimized by maximizing the minimum Euclidean distance to obtain a 3D space 10-CSK constellation that meets the target perceived color constraint. The target perceived color is defined as: E[Λ n ]=T c Where, T c Indicates the intensity symbol corresponding to the target perceived color, T c =[1 / 3,1 / 3,1 / 3] T ; Step 2 is as follows: Based on the selection criterion of the minimum number of symbol sequences at the MED, it can be described as: Where, d S′,min With d S″,min Respectively represent the MED between sequence symbols in sequence sets S′ and S″, both equal to d min ; N S″ represents the number of sequence symbols at the MED based on the sequence set S″, where S″ is the selected symbol sequence set, that is, S″|=16; According to the selection criteria of the symbol sequence, the symbol sequence set S″ is obtained as: Through the analysis of S″, we can know that N S′ =4; and then obtain the data rate R = 2 bits / symbol 5-CSK modulation scheme; Step 3 is as follows: When L = 2, the corresponding symbol sequence set S′ = {s0s0, s0s1, …, s i s j ,…,s9s9}, based on the target perceived color constraint, there are many possible situations for selecting 64 groups of symbol sequences from S′; based on the fact that the MED between symbol sequences remains unchanged, a selection criterion with the minimum number of symbol sequences at the MED is proposed. According to the symbol sequence selection criterion, the symbol sequence set S″ is obtained as: Through the analysis of S″, we can know that N S″ =6; and then the data rate R=3 bits / symbol 10-CSK modulation scheme is obtained.

2. A multi-stage PLDPC-CSK design method based on symbol sequence according to claim 1, characterized in that: Step 1 is as follows: divide the 2D constant intensity plane triangle CSK constellation by triangles to obtain a symmetric 10-CSK constellation that satisfies the perceived color constraint; at the transmitting end, the data bit sequence μ is first divided into m data blocks, and the length of the data block is k. After PLDPC encoding, the output codeword is of length n Secondly, m multi-level codewords output m information bits to form the i-th array, which is mapped into symbol s after CSK modulation. i =(s r,i ,s g,i ,s b,i ) and transmit it to the channel; and Denote the mapping function and demapping function respectively, then and The receiving end receives the signal r from the VLC channel i For: r i =Hs i +n; where H is a 3×3 channel gain matrix; n is a 3×1 additive white Gaussian noise vector with a mean of 0 and a variance of σ 2 ; The receiving end introduces multi-level iterative soft decoding between the demodulator and the decoder; The CSK demodulator uses the Log-Map algorithm to soft-demodulate the received symbol sequence and output As the prior information of decoder l; decoder l uses BP algorithm to decode and output Multi-level After parallel-to-serial conversion, the data information sequence is output The 10-CSK constellation is expanded from the 2D plane to the 3D multi-intensity space, and the six constellation points in the maximum intensity plane are optimized to maximize the minimum Euclidean distance, thus obtaining a 3D space 10-CSK constellation that meets the target perceived color constraints.

3. The multi-stage PLDPC-CSK design method based on symbol sequence according to claim 1, characterized in that: Step 4 is as follows: Under known AWGN channel conditions, the soft information output by the demapping module at the receiving end is described in the form of log-likelihood ratio. Then, the i-th data information of a transmitted symbol sequence ss is The log-likelihood ratio is: And meet For multi-level PLDPC codes, demodulation and decoding are performed level by level, and the l-1 level PLDPC decoding output is used for l level demodulation and PLDPC decoding; The demodulator outputs soft information which is transmitted to the decoder as its prior information: Set M c is the set of check nodes, N v is a collection of variable nodes, Indicates that the variable node VN in the hth BP iterative decoding of the PLDPC code i Sent to verification node CN j Soft information; Indicates that in the hth iteration, the check node CN j Send to variable node VN i Soft information; It means that after the hth iteration, decoder l obtains the codeword of soft information.

4. The multi-stage PLDPC-CSK design method based on symbol sequence according to claim 3 is characterized in that: The steps of BP iterative decoding are as follows: (1) Initialization, (2) Update information obtained by the check node from the variable node: (3) Update information obtained by variable nodes from check nodes: (4) The soft information of the codeword can be obtained based on the updated information of the check nodes and variable nodes: If the PLDPC decoder l decodes correctly, it stops decoding and outputs decision c (l) ; Otherwise, update the number of iterations h and go to step (2); if h>IT, stop iterative decoding and output the decision c (l) ; Get the m-level PLDPC coded bits for: Among them, the external information output by decoder 1 is: In the formula Feedback to the demodulator as its prior information: Multi-level iterative soft decoding reaches the maximum number of iterations N MAX When , decoding ends.

Images