Demodulation and decoding method based on 16APSK waveform

Abstract

The invention discloses a demodulation and decoding method based on a 16APSK waveform, and belongs to the field of communication signal processing. Low-complexity receiver construction is achieved through approximate processing during demodulation, decoding likelihood ratio information initialization and conversion of exponent and logarithm operation into simple addition and multiplication, and the method specifically comprises the steps that an original information sequence is sequentially subjected to coding and modulation processing and enters a complex channel for transmission; the demodulator receives output symbols from the AWGN channel, and initializes log-likelihood ratio information of a first bit, a second bit, a third bit and a fourth bit by using the channel receiving symbols; and an iterative decoding process is completed in the LDPC decoder, and construction of the 16APSK soft demodulation receiver based on the LDPC code is completed. According to the invention, LDPC channel coding and high-order 16 APSK modulation are combined, so that the spectrum utilization rate of the system is effectively improved; under the condition of ensuring the decoding accuracy, the complexity of engineering implementation is reduced, the engineering structure is optimized, and related engineering problems are solved.

Description

Technical Field

[0001] The present invention relates to a demodulation and decoding method based on 16APSK waveform, belonging to the field of communication signal processing. Background Art

[0002] Digital modulation and channel coding technologies are two commonly used digital signal processing technologies in wireless communication systems and are the key to improving system performance. The use of high-order modulation technology can improve the frequency band utilization rate of the system. Amplitude Phase Shift Keying (APSK) modulation combines the characteristics of amplitude modulation and phase modulation and represents digital information by changing the amplitude and phase of the signal. Low Density Parity Check (LDPC) codes have the advantages of performance close to the Shannon limit and being convenient for parallel processing. Combining 16APSK modulation and LDPC channel coding, the communication system can significantly improve the data transmission efficiency and have high reliability. In order to fully improve the channel coding gain and achieve low complexity, the LDPC decoding method usually adopts the Min-Sum (MS) algorithm in the soft decision decoding method, and it is necessary to output the soft information of each bit during the demodulation process based on the constellation diagram. Currently, the existing demodulation algorithms have a large number of multiplication, exponential, and logarithmic operations, and the implementation complexity is relatively high in an Additive White Gaussian Noise (AWGN) channel, wasting resources. Summary of the Invention

[0003] In order to solve the problem that the traditional 16APSK soft demodulation algorithm has a relatively high complexity in the AWGN channel and requires a large number of multiplication, exponential, and logarithmic operations, the purpose of the present invention is to provide a demodulation and decoding method based on 16APSK waveform. Through approximate processing during demodulation, initializing the decoding likelihood ratio information, converting exponential and logarithmic operations into simple addition and multiplication operations, a receiver structure with low complexity is realized, the processing delay is reduced, and the resource utilization rate is improved.

[0004] The purpose of the present invention is achieved through the following technical solutions:

[0005] A demodulation and decoding method based on 16APSK waveform specifically includes the following steps:

[0006] Step 1: Perform LDPC encoding and 16APSK modulation processing on the original information sequence in sequence, and transmit it through a complex AWGN channel;

[0007] The original information sequence u with a length of k bits is encoded by an LDPC encoding module to obtain an encoded sequence c = u·G, where G is the generating matrix of the LDPC code, and the length of the encoded sequence c is m;

[0008] Modulate the encoded information bits through the 16APSK constellation mapping function such that every 4 encoded and interleaved information bits b3, b2, b1, b0 are mapped to one constellation point in the space;

[0009] Among them, is the complete constellation space of the symbols after 16APSK mapping, which contains a total of 16 constellation points. The radius of the inner - circle constellation points is R1, and the radius of the outer - circle constellation points is R2;

[0010] The encoded sequence c is modulated and mapped through 16APSK to obtain the symbol sequence y = φ(c) = {y1, y2, …, y t};

[0011] Among them, is rounding up. That is, if m is not divisible by 4, then 0s are appended at the end of the sequence until it satisfies 4 bits before mapping;

[0012] The symbol sequence y is transmitted through a complex AWGN channel;

[0013] Step 2: The demodulator receives the output symbols from the AWGN channel and initializes the log - likelihood ratio information of the first bit using the received channel symbols;

[0014] The symbol sequence z = {z1, z2, …, z t} output by the channel, where z = y + n, and n is a complex additive white Gaussian noise sequence with a mean of 0 and a variance of ;

[0015] When deriving the soft - bit information using the initial symbol information of the received channel, the information source is regarded as an equiprobable distribution, that is, the probability of each initial bit being 0 or 1 is equal,

[0016] According to the sub - constellation region division diagram of the first bit b3 corresponding to the received symbol z t , the sub - constellation diagram of the first bit b3 is symmetric about the I - axis and Q - axis. Map the received symbol z t to the first quadrant as z′ t , and the log - likelihood ratio LLR(b3) of the first bit b3 is shown in Equation (1):

[0017]

[0018] Among them, i = 0, 4, 8, 12, is the i - th constellation mapping point, and ||a - b|| is the distance between points a and b on the constellation mapping diagram;

[0019] Utilize ray D1: ∠θ = π / 6 and line D2: Q = D respectively min , where determine max(P0, P4) and max(P8, P 12 ), as shown in Equation (2) and Equation (3) respectively:

[0020]

[0021] where, (x) I and (x) Q are the real part and the imaginary part of the complex signal x respectively;

[0022] Step 3: The demodulator receives the output symbol from the AWGN channel and initializes the log-likelihood ratio information of the second bit using the channel received symbol;

[0023] According to the sub-constellation region division diagram of the second bit b2 corresponding to the received symbol z t , the sub-constellation diagram of the second bit b2 is symmetric about the I-axis and the Q-axis. Map the received symbol to the first quadrant. The log-likelihood ratio LLR(b2) of the second bit b2 is as shown in Equation (4):

[0024]

[0025] Utilize ray D3: ∠θ = π / 3 and line D4: I = D respectively min , where determine max(P0, P8) and max(P4, P 12 ), as shown in Equation (5) and Equation (6) respectively:

[0026]

[0027] Step 4: The demodulator receives the output symbol from the AWGN channel and initializes the log-likelihood ratio information of the third bit using the channel received symbol;

[0028] According to the sub-constellation region division diagram of the third bit b1 corresponding to the received symbol z t , the sub-constellation diagram of the third bit b1 is completely symmetric about the Q-axis. The LLR value of the third bit b1 only depends on the real part (z t ) of the received symbol z t ) I and is independent of the imaginary part (z t ). LLR(b1) is approximately as shown in Equation (7): Q LLR(b1) = α × Amp × (z

[0029] LLR(b1) = α × Amp × (z t ) I (7)

[0030] Among them, Amp = (R1·4 + R2·12) / 16 is the average amplitude of the constellation diagram, and α is the weight factor;

[0031] Step Five: The demodulator receives the output symbol from the AWGN channel and initializes the log-likelihood ratio information of the 4th bit using the channel received symbol;

[0032] According to the received symbol z t For the sub-constellation region division diagram of the 4th bit b0 corresponding to it, the sub-constellation diagram of the 4th bit b0 is completely symmetric about the I axis. The LLR value of the 4th bit b0 only depends on the imaginary part (z t ) t of the received symbol z Q and has nothing to do with the real part (z t ) I of the received symbol z. LLR(b0) is approximately as shown in Equation (8):

[0033] LLR(b0) = α × Amp × (z t ) Q (8)

[0034] Among them, the average amplitude Amp and the weight factor α are the same as those in Step Four;

[0035] Step Six: Complete the iterative decoding process within the LDPC decoder, which specifically includes the following sub-steps:

[0036] Step 6.1 Update the information passed from the check node to the variable node, as shown in Equation (9):

[0037]

[0038] Among them, initialize Lq ij , as the log-likelihood ratio of the i-th bit obtained in Steps Two to Five, and the set R j\i is the set of positions other than the position of the i-th 1 in the j-th row of the parity-check matrix H;

[0039] Step 6.2 Update the information passed from the variable node to the check node, as shown in Equation (10):

[0040]

[0041] Among them, Lp i is the log-likelihood ratio of the i-th bit calculated in Steps Two to Five, and the set C i\j is the set of positions other than the position of the j-th 1 in the i-th row of the parity-check matrix H;

[0042] Step 6.3 Determine the posterior log-likelihood ratio of each bit, as shown in Equation (11):

[0043]

[0044] Step 6.4 Hard-decide each variable node to 0 or 1, as shown in Equation (12):

[0045]

[0046] Step 6.5 Determine the stopping condition:

[0047] If the decision result satisfies the parity-check matrix H of the LDPC code, that is or reaches the preset maximum number of iterations, stop the decoding iteration; otherwise, return to Step 6.1 to update the information again;

[0048] Through the above six steps, the construction of the 16APSK soft demodulation receiver based on LDPC code is completed;

[0049] It further includes Step Seven: Use the decision result obtained in Step Six as the demodulation and decoding result of the receiver. By judging the region where the received symbol is located in each bit sub-constellation diagram, simplify the calculation process of bit soft information, reduce the exponential processing unit and logarithmic processing unit in the traditional calculation circuit, and can reduce the complexity of engineering implementation while ensuring the decoding accuracy rate, optimize the engineering structure, and solve related engineering problems.

[0050] Beneficial effects:

[0051] 1. For a demodulation and decoding method based on 16APSK waveform of the present invention, it combines LDPC channel coding with high-order 16APSK modulation, and effectively improves the spectral efficiency of the system while ensuring the relatively low bit error rate brought by channel coding.

[0052] 2. For a demodulation and decoding method based on 16APSK waveform of the present invention, from the perspective of engineering implementation, a large number of exponential operations and logarithmic operations are transformed into simple addition operations and comparison operations. Simplify the 16APSK soft demodulation algorithm according to the symmetry characteristics of the constellation diagram, reduce the complexity of bit soft information operation, and greatly reduce the consumption of computing resources on the basis of relatively ideal bit error rate performance. Description of the Drawings

[0053] Figure 1 It is a flowchart of a demodulation and decoding method based on 16APSK waveform of the present invention;

[0054] Figure 2 It is the 16APSK constellation diagram of the embodiment;

[0055] Figure 3 Four sub - constellation region division diagrams obtained by decomposing the 16 - APSK constellation diagram of the embodiment;

[0056] Figure 4 Decision relationship diagram of the first and second bits of the embodiment in the first quadrant;

[0057] Figure 5 Bit - error rate simulation comparison curve of the reference code pattern combined with the 16 - APSK modulation receiver of the embodiment. Detailed implementation manners

[0058] To better illustrate the purpose and advantages of the present invention, the following further describes the content of the invention in combination with the drawings and examples.

[0059] Embodiment 1:

[0060] Applying a demodulation and decoding method based on 16 - APSK waveform of the present invention, simplifying the soft - demodulation algorithm according to the characteristics of the 16 - APSK constellation diagram, and using the decision region to approximately calculate the soft information required for the decoding process, as Figure 1 shown, including the following steps:

[0061] In the embodiment, the 16 - APSK mapping method is in accordance with the regulations of the DVB - S2 system, and the mapping diagram is as Figure 2 shown. The LDPC code selects the π - cyclic code type, and the main parameters are shown in the following table:

[0062] Parameter Details Code length 248 Code rate 1 / 2 Modulation method 16APSK Channel model Additive white Gaussian noise Maximum number of decoding iterations 20 Ratio of inner radius to outer radius of constellation diagram 1 / 3

[0063] Step 1: Perform LDPC encoding and 16 - APSK modulation processing on the original information sequence in turn, and enter the complex AWGN channel for transmission;

[0064] The original information sequence u with a length of 124 bits is encoded by the LDPC encoding module to obtain the encoded sequence c = u·G, where G is the generating matrix of the LDPC code, and the length of the encoded sequence c is 248;

[0065] The encoded information bits are modulated through the 16 - APSK constellation mapping function Each group of 4 encoded and interleaved information bits b3, b2, b1, b0 is mapped to 1 constellation point in the space;

[0066] Among them, is the complete constellation space of the symbols after 16 - APSK mapping, which contains 16 constellation points in total. The radius of the inner - circle constellation points is R1, and the radius of the outer - circle constellation points is R2;

[0067] The encoded sequence c is mapped through 16 - APSK modulation to obtain the symbol sequence y = φ(c) = {y1, y2, …, yt} where t = 248 / 4 = 62;

[0068] The symbol sequence y is transmitted through a complex AWGN channel;

[0069] Step 2: The demodulator receives the output symbols from the AWGN channel and initializes the log-likelihood ratio information of the first bit using the channel received symbols;

[0070] The symbol sequence z output by the channel = {z1, z2, …, z t} where z = y + n, and n is a complex additive white Gaussian noise sequence with a mean of 0 and a variance ;

[0071] When deriving the soft bit information using the initial symbol information of the received channel, the source is regarded as an equally probable distribution, that is, the probabilities of each initial bit being 0 or 1 are equal.

[0072] According to the sub-constellation region division diagram of the first bit b3 corresponding to the received symbol z t as shown in Figure 3 (a), the sub-constellation diagram of the first bit b3 is symmetric about the I-axis and the Q-axis. Map the received symbol z t to the first quadrant as z t ′. The log-likelihood ratio LLR(b3) of the first bit b3 is as shown in Equation (1):

[0073]

[0074] where i = 0, 4, 8, 12, is the i-th constellation mapping point, and ||a - b|| is the distance between points a and b on the constellation mapping diagram;

[0075] As shown in Figure 4 (a), use the ray D1: ∠θ = π / 6 and the straight line D2: Q = D min respectively, where to determine max(P0, P4) and max(P8, P 12 ), as shown in Equations (2) and (3) respectively:

[0076]

[0077] where (x) I and (x) Q are the real part and the imaginary part of the complex signal x respectively;

[0078] Step 3: The demodulator receives the output symbols from the AWGN channel and initializes the log-likelihood ratio information of the second bit using the channel received symbols;

[0079] According to the received symbol z t The sub - constellation region division diagram of the corresponding second bit b2 is as follows Figure 3 As shown in (b), the sub - constellation diagram of the second bit b2 is symmetric about the I - axis and Q - axis. Map the received symbol to the first quadrant. The log - likelihood ratio LLR(b2) of the second bit b2 is as shown in Equation (4)

[0080]

[0081] As Figure 4 shown in (b), respectively use the ray D3: ∠θ = π / 3 and the straight line D4: I = D min , where Determine max(P0, P8) and max(P4, P 12 ), as shown in Equation (5) and Equation (6) respectively:

[0082]

[0083] Step 4: The demodulator receives the output symbol from the AWGN channel and initializes the log - likelihood ratio information of the third bit using the channel - received symbol;

[0084] According to the received symbol z t The sub - constellation region division diagram of the corresponding third bit b1 is as follows Figure 3 As shown in (c), the sub - constellation diagram of the third bit b1 is completely symmetric about the Q - axis. The LLR value of the third bit b1 only depends on the real part of the received symbol z t (z t ) I and has nothing to do with the imaginary part (z t ). The LLR(b1) is approximately as shown in Equation (7): Q LLR(b1) = α × Amp × (z

[0085] LLR(b1) = α × Amp × (z t ) I (7)

[0086] where Amp = (R1·4 + R2·12) / 16 is the average amplitude of the constellation diagram, and α is the weight factor, α = 1.25;

[0087] Step 5: The demodulator receives the output symbol from the AWGN channel and initializes the log - likelihood ratio information of the fourth bit using the channel - received symbol;

[0088] According to the received symbol z t The sub - constellation region division diagram of the corresponding fourth bit b0 is as follows Figure 3As shown in (d), the sub - constellation diagram of the 4th bit b0 is completely symmetric about the I - axis. The LLR value of the 4th bit b0 is only related to the imaginary part of the received symbol z t (z t ), Q and has nothing to do with the real part (z t ). LLR(b0) is approximately as shown in Equation (8): I LLR(b0) = α×Amp×(z

[0089] (8) t ) Q where the average amplitude Amp and the weight factor α are the same as in Step 4;

[0090] Step 6. Complete the iterative decoding process within the LDPC decoder, which specifically includes the following sub - steps:

[0091] Step 6.1 Update the information passed from the check node to the variable node, as shown in Equation (9):

[0092]

[0093]

[0094] where the initial Lq ij ,, is the log - likelihood ratio of the i - th bit obtained in Steps 2 to 5, and the set R j\i is the set of positions other than the position of the i - th 1 in the j - th row of the parity - check matrix H;

[0095] Step 6.2 Update the information passed from the variable node to the check node, as shown in Equation (10):

[0096]

[0097] where Lp i is the log - likelihood ratio of the i - th bit calculated in Steps 2 to 5, and the set C i\j is the set of positions other than the position of the j - th 1 in the i - th row of the parity - check matrix H;

[0098] Step 6.3 Determine the posterior log - likelihood ratio of each bit, as shown in Equation (11):

[0099]

[0100] Step 6.4 Make a hard decision on each variable node to be 0 or 1, as shown in Equation (12):

[0101]

[0102] Step 6.5 Determine the stop condition:

[0103] If the judgment result satisfies the parity-check matrix H of the LDPC code, that is or reaches the preset maximum number of iterations, stop the decoding iteration; otherwise, return to step 6.1 to update the information again;

[0104] The embodiment combines the (248, 124) LDPC code and 16APSK modulation, improves the bandwidth utilization rate on the basis of ensuring the channel coding gain, and adopts an approximate algorithm during demodulation and decoding, reducing the receiver implementation complexity with almost no loss of system performance;

[0105] Perform a bit error rate simulation analysis for this embodiment and compare it with traditional demodulation algorithms and the Max-log-MAP algorithm. The simulation results are as Figure 5 shown. The bit error rate of the soft demodulation approximation method of the present invention has no obvious loss, but the implementation complexity is greatly reduced;

[0106] Through the above six steps, the construction of a 16APSK soft demodulation receiver based on the LDPC code is completed;

[0107] It further includes step seven: using the judgment result obtained in step six as the receiver demodulation and decoding result, by judging the region where the received symbol is located in each bit sub-constellation diagram, simplifying the calculation process of bit soft information, reducing the exponential processing unit and logarithmic processing unit in the traditional calculation circuit, being able to reduce the engineering implementation complexity while ensuring the decoding accuracy, optimizing the engineering structure, and solving related engineering problems.

[0108] The above specific description further details the purpose, technical solution, and beneficial effects of the invention. It should be understood that the above is only a specific embodiment of the present invention and is not used to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A demodulation and decoding method based on 16APSK waveform, characterized in that: By performing approximate processing during demodulation, initializing the decoding likelihood ratio information, and converting exponential and logarithmic operations into simple addition and multiplication operations, a low-complexity receiver can be constructed, processing delay can be reduced, and resource utilization can be improved. Specifically, the following steps are included: Step 1: The original information sequence is sequentially LDPC encoded and 16APSK modulated, and then transmitted through the complex AWGN channel; The original information sequence u with a length of k bits is passed through the LDPC coding module to obtain a coded sequence c = u·G, where G is the generator matrix of the LDPC code and the length of the coded sequence c is m; The encoded information bits are passed through the 16APSK constellation mapping function Modulation is performed, and every 4 coded interleaved information bits b3, b2, b1, b0 are mapped to a constellation point in a space; in, It is the complete constellation space of the 16APSK mapped symbol, which contains 16 constellation points in total. The radius of the inner circle constellation points is R1, and the radius of the outer circle constellation points is R2; The coded sequence c is mapped through 16APSK modulation to obtain the symbol sequence y = φ (c) = {y1, y2, ..., y t }; in, To round up, that is, if m cannot divide 4, 0 is added to the end of the sequence, and mapping is performed after 4 bits are satisfied; The symbol sequence y is transmitted through the complex AWGN channel; Step 2: The demodulator receives the output symbol from the AWGN channel and initializes the log-likelihood ratio information of the first bit using the channel received symbol; The symbol sequence z output by the channel is {z1,z2,…,z t }, z = y + n, n is the mean of 0 and the variance of Complex additive Gaussian white noise sequence; When the soft bit information is derived using the initial symbol information of the receiving channel, the signal source is regarded as an equal probability distribution, that is, the probability of each initial bit being 0 or 1 is equal. According to the received symbol z t The corresponding sub-constellation area division diagram of the first bit b3, the sub-constellation diagram of the first bit b3 is symmetrical about the I axis and the Q axis, and the received symbol z t Mapped to the first quadrant as z t ′, the log-likelihood ratio LLR(b3) of the first bit b3 is as shown in formula (1): in, is the i-th constellation mapping point, ||ab|| is the distance between points a and b on the constellation mapping diagram; Using ray D1:∠θ=π / 6 and straight line D2:Q=D min ,in Determine max(P0,P4) and max(P8,P 12 ), as shown in formula (2) and formula (3) respectively: Among them, (x) I and (x) Q are the real and imaginary parts of the complex signal x respectively; Step 3: The demodulator receives the output symbol from the AWGN channel and initializes the log-likelihood ratio information of the second bit using the channel received symbol; According to the received symbol z t The corresponding sub-constellation area division diagram of the second bit b2 is symmetrical about the I axis and the Q axis, and the received symbol is mapped to the first quadrant. The log-likelihood ratio LLR(b2) of the second bit b2 is shown in formula (4): Using ray D3:∠θ=π / 3 and straight line D4:I=D min ,in Determine max(P0,P8) and max(P4,P 12 ), as shown in formula (5) and formula (6) respectively: Step 4: The demodulator receives the output symbol from the AWGN channel and initializes the log-likelihood ratio information of the third bit using the channel received symbol; According to the received symbol z t The corresponding sub-constellation area division diagram of the third bit b1. The sub-constellation diagram of the third bit b1 is completely symmetrical about the Q axis. The LLR value of the third bit b1 is only related to the received symbol z t The real part (z t ) I is related to the imaginary part (z t ) Q Regardless of , LLR(b1) is approximately as shown in formula (7): LLR(b1)=α×Amp×(z t ) I (7) Among them, Amp = (R1·4 + R2·12) / 16 is the average amplitude of the constellation diagram, and α is the weight factor; Step 5: The demodulator receives the output symbol from the AWGN channel and initializes the log-likelihood ratio information of the 4th bit using the channel received symbol; According to the received symbol z t The corresponding sub-constellation area division diagram of the 4th bit b0. The sub-constellation diagram of the 4th bit b0 is completely symmetrical about the I axis. The LLR value of the 4th bit b0 is only related to the received symbol z t The imaginary part (z t ) Q related to the real part (z t ) I Regardless of , LLR(b0) is approximately as shown in equation (8): LLR(b0)=α×Amp×(z t ) Q (8) Among them, the average amplitude Amp and the weight factor α are the same as those in step 4; Step 6: The iterative decoding process is completed in the LDPC decoder, which specifically includes the following sub-steps: Step 6.1 updates the information transmitted from the check node to the variable node, as shown in formula (9): Among them, initialize Lq ij , is the log-likelihood ratio of the i-th bit obtained in steps 2 to 5, and the set R j\i is the set of positions in the j-th row of the check matrix H except the i-th 1; Step 6.2 updates the information transmitted by the variable node to the check node, as shown in formula (10): Among them, Lp i is the log-likelihood ratio of the i-th bit calculated in steps 2 to 5, and the set C i\j is the set of positions in the i-th row of the check matrix H except the j-th 1; Step 6.3 determines the posterior log-likelihood ratio of each bit, as shown in equation (11): Step 6.4 makes a hard decision on each variable node to be 0 or 1, as shown in formula (12): Step 6.5 Determine the stopping condition: If the judgment result The check matrix H that satisfies the LDPC code is Or if the preset maximum number of iterations is reached, the decoding iteration is stopped, otherwise, the process returns to step 6.1 to update the information again; Through the above six steps, the construction of 16APSK soft demodulation receiver based on LDPC code is completed.

2. A demodulation and decoding method based on 16APSK waveform as claimed in claim 1, characterized in that: Step 7: Use the judgment result obtained in step 6 as the demodulation and decoding result of the receiver. By judging the area where the received symbol is located in each bit sub-constellation diagram, the calculation process of the bit soft information is simplified, and the exponential processing unit and logarithmic processing unit in the traditional computing circuit are reduced. This can reduce the complexity of the engineering implementation, optimize the engineering structure, and solve related engineering problems while ensuring the decoding accuracy.

Images