Radio station adaptive joint coding modulation method applied to severe channel environment

Abstract

The invention provides a radio station adaptive joint coding modulation method applied to a severe channel environment, and the method supports interconnection and intercommunication of a plurality of radio stations, and comprises a transmitting end and a receiving end. Wherein the data processing process of the sending end is as follows: data to be sent and a reference signal form a data frame according to a frame format, and scrambling interleaving processing is carried out; performing polarization coding on the processed data frame; sCMA coding is carried out; self-adaptive chirp spread spectrum modulation is carried out; the data processing process of the receiving end is as follows: the receiving end radio station synchronizes the received chirp modulation signal and then performs incoherent demodulation; channel signal-to-noise ratio estimation is carried out; sCMA multi-user detection is carried out on the demodulated data information; and performing CRC (cyclic redundancy check). According to the invention, interconnection and intercommunication of a large number of radio stations can be realized by occupying less time-frequency resources in a severe channel environment.

Description

Technical Field

[0001] This application relates to the field of anti-interference communication technologies, and particularly to a radio adaptive joint coding and modulation method applied to harsh channel environments. Background Art

[0002] In the application scenarios of radios in complex electromagnetic environments, there are a large number of shadow fades, multipath effects, and strong electromagnetic interference, which can cause the transmitted signal to undergo attenuation and interference, shorten the communication distance, make information interaction difficult, and at the same time reduce the available frequency band resources, seriously affecting the reliability of communication and the data transmission efficiency. Currently, communication broadband radios do not yet have the communication ability in extremely low signal-to-noise ratio environments, and their technologies mainly have the following deficiencies:

[0003] In terms of anti-interference: Currently, radio transceivers generally adopt a medium-speed frequency hopping anti-interference system. Frequency hopping is one of the most commonly used spread spectrum methods, which refers to a communication method in which the carrier frequencies of the transmitted and received signals change discretely according to a predetermined rule. Compared with fixed-frequency communication, frequency hopping communication is more concealed and difficult to intercept; at the same time, frequency hopping communication has good anti-interference ability. Even if some frequency points are interfered, normal communication can still be carried out on other non-interfered frequency points. However, in extremely low signal-to-noise ratios, interference and noise will have a certain impact on the detection of frequency hopping signals, resulting in a sharp increase in the bit error rate at the receiving end.

[0004] In terms of channel encoding and decoding: Currently, RS codes (Reed-Solomon codes) and convolutional codes are generally used to correct errors in the transmitted data. RS codes and convolutional codes are efficient linear error correction coding technologies, and their implementation costs are simple and power consumption is low, which are widely used in various communication systems. However, the signal-to-noise ratio required to achieve an acceptable bit error rate is relatively high, and the error correction performance is poor in complex electromagnetic environment scenarios.

[0005] In terms of multiple access: Multiple access communication means that many users form a communication network, and any two users in the network can communicate, and when multiple pairs of users communicate simultaneously, they do not interfere with each other. Currently, most radios adopt a Time Division Multiple Access (TDMA) multiple access method. TDMA technology divides the channel according to time slots, that is, different access radios are assigned different time periods to share the same channel. In a TDMA system, time is divided into periodic frames, and each frame is further divided into several non-overlapping time slots. The number of time slots on each carrier is limited, which limits the number of users who can communicate simultaneously; when the data transmissions of different users are uneven, idle time slots will cause resource waste and reduce the spectrum utilization rate; fixed time slot allocation will also bring data transmission delays, making it difficult to meet the low-latency requirements of radio services. Summary of the Invention

[0006] The present application provides a radio adaptive joint coding and modulation method for harsh channel environments, which is used to solve the problem that existing radios are difficult to communicate at low signal-to-noise ratios, and can effectively improve the waste of time-frequency resources in the prior art.

[0007] The present application provides a radio adaptive joint coding and modulation method for harsh channel environments. The method supports interconnection and interoperability of multiple radios, including a sending end and a receiving end;

[0008] It is assumed that there are J user radios sending data at the sending end;

[0009] Among them, the data processing process at the sending end is as follows:

[0010] Step 11: Combine the data to be sent and the reference signal into a data frame according to the frame format, and perform scrambling and interleaving processing to increase the anti-interference ability of the data;

[0011] Step 12: Perform polar coding on the processed data frame to introduce redundancy into the data sequence;

[0012] Step 13: Through the SCMA coding technology, map the data from the bit domain to the K-dimensional complex domain codebook, and non-orthogonally superimpose the coded words of each user after SCMA coding onto K orthogonal time-frequency resources; the mapping process is realized through a pre-allocated multi-dimensional codebook, and different user data streams are distinguished according to subcarriers and codebooks;

[0013] Step 14: Select an appropriate spreading factor according to the reference signal signal-to-noise ratio that reflects the current channel quality fed back by the receiving end, perform adaptive chirp spread spectrum modulation on the subcarrier data, obtain the transmission signal for transmission, and synchronize the updated spreading factor to the receiving node at the same time;

[0014] The data processing process at the receiving end is as follows:

[0015] Step 21: After the radio at the receiving end synchronizes the received chirp modulation signal, perform non-coherent demodulation;

[0016] Step 22: Perform channel signal-to-noise ratio estimation; extract the received signal sequence of the time-frequency resource where the reference signal is located and count its power. Since the receiving end has prior information of the reference signal, the signal-to-noise ratio is obtained based on the calculated received signal power and the known reference signal power, and fed back to the sending end;

[0017] Step 23: Perform SCMA multi-user detection on the demodulated data information using MPA, and output the soft decision of each user's information;

[0018] Step 24: Decode the polar code according to the soft decision information and perform CRC check, and finally recover the original bits sent by the sending end.

[0019] Further, in step 12, polar coding is performed on the processed data frame to introduce redundancy into the data sequence, including:

[0020] After channel polarization design of N independent channels, N polarized channels are obtained. A part of the N polarized channels has a channel capacity of 1, and the other part tends to a noise channel, that is, the channel capacity is 0. The information bits to be transmitted are transmitted on K sub-channels with high reliability. Fixed bits are transmitted on the remaining N - K sub-channels, and the fixed bits are known for both encoding and decoding. The mixed information bits and fixed bits are combined to obtain u N =(u1, u2,..., u N ), representing the bit sequence to be encoded transmitted by the radio station; G N is the generating matrix. According to c = u N G N , the encoded codeword c is obtained, where the output of the j-th user is It is set that the length of the bit sequence to be encoded is 32, and the code rate supports [0.5, 0.6, 0.7, 0.8]. Different code rates correspond to different transmission rates.

[0021] Further, in step 13, the SCMA encoding process is as follows:

[0022] The SCMA encoder maps the binary bits after polar coding to multi-layer constellation point symbols, and at the same time, the constellation point symbols are dispersed to different subcarriers. The encoding mapping process is as follows:

[0023] An SCMA(J, K) system is adopted: The SCMA encoder is defined as: where M is the constellation point set, representing the types of user transmission bit combinations; K is the number of orthogonal subcarriers. Every log2(M) information bits form an information symbol, which is mapped to a K-dimensional complex domain codeword, and the mapping relationship is determined by the SCMA codebook. After mapping, the codewords of J users are non-orthogonally superimposed on K subcarriers, and K < J;

[0024] The codebook is determined by a method based on constellation rotation and interleaving. The design principle of the lattice constellation is introduced, and the compact codebook of the multi-dimensional constellation is determined by minimizing the average symbol energy of the minimum Euclidean distance between given constellation points. The codebook is a matrix with dimensions [K, M]. The codeword is generated through the following steps:

[0025] Step 131, modulation: Map the input log2(M) bits to multi-dimensional constellation modulation symbols, and perform constellation rotation and interleaving processing to generate the modulation symbol set S K ;

[0026] Step 132, sparse expansion: Expand the modulation symbol into a sparse codeword C K , C K = F K ·SK , F K is a sparse mapping matrix that defines the non - zero positions of each user on the resource nodes and represents the allocation of carrier resources for each user;

[0027] The sparse mapping matrix F K is as follows:

[0028]

[0029] It can be known from the sparse mapping matrix that for each user, that is, each column, the sub - carriers occupied, which are one row in the allocation matrix, are 2, and each sub - carrier resource is used by 3 users simultaneously.

[0030] Furthermore, in step 14, adaptive chirp spread - spectrum modulation: includes:

[0031] Before modulation, evaluate the current channel quality according to the SNR fed back by the receiving end, and dynamically select the spreading factor suitable for the current channel conditions; the very - low SNR range is below - 5 dB;

[0032] When the SNR is low, select a larger spreading factor, i.e., SF > 8, to improve the anti - interference ability; when the SNR is high, select a smaller spreading factor, i.e., SF < 8; use the SNR as the spreading - factor switching threshold, and the corresponding relationship is:

[0033]

[0034]

[0035] Before modulation, first digitize the complex sequence after SCMA coding through the Matlab function f(n)=dec2bin(abs(complex sequence), λ) into a binary bit sequence f(n), n ∈ [0, M - 1], where M = 2 λ , and λ is the spreading factor SF, that is, the number of binary bits transmitted by each modulation symbol; after processing, perform chirp spread - spectrum modulation: in the discrete - time domain, the CSS symbol composed of M sampling values is expressed as S(n)=f(n)c u (n), where f(n) is the information symbol, expressed as c u (n) is the upward chirp symbol, expressed as

[0036] Furthermore, in step 21, after the receiving - end radio synchronizes the received chirp - modulated signal, perform non - coherent demodulation, including:

[0037] Use non - coherent detection, and the transmitted spreading factor FS is estimated as:

[0038]

[0039] where y(n) is the signal received by the radio station, is the conjugate unmodulated chirp signal of the receiving radio station's local oscillator; for the FS estimation using non-coherent detection, first estimate the DFT of the product signal The estimation result is R(k); then, estimate the absolute value of R(k) and find the frequency index with the maximum peak.

[0040] Furthermore, in step 22, perform channel signal-to-noise ratio (SNR) estimation, including:

[0041] Extract the received signal sequence of the time-frequency resource where the reference signal is located and calculate the power. Since the receiving end has the prior information of the reference signal, based on the calculated received signal power and the known reference signal power, obtain the SNR and feedback it to the transmitting end.

[0042] Furthermore, in step 23, perform SCMA multi-user detection on the demodulated data information using MPA and output the soft decision of each user's information, including:

[0043] Perform multi-user detection using the MPA algorithm; from the SCMA coding, it can be seen that the mapping from SCMA users to carrier resources can be determined by the resource allocation matrix F, (F) kj = 1 indicates that user j occupies subcarrier k, conversely, (F) kj = 0 indicates that user j does not occupy subcarrier k; regard users as variable nodes, represented by circles, and carriers as function nodes, represented by squares, and connect the variable nodes and function nodes where (F) kj = 1 with lines to obtain the factor graph of the allocation matrix F;

[0044] The MPA algorithm updates the prior information by passing extrinsic information between variable nodes and function nodes; assume that initially, the probabilities of all symbols are equal; first perform the initialization of conditional probability and prior probability, and then continuously update the variable nodes and function nodes through iteration; finally, calculate the log-likelihood ratio from the transmission probability of each codeword of the variable node and input it to the channel decoder.

[0045] Furthermore, in step 24, perform polar code decoding based on the soft decision information and perform CRC check, and finally recover the original bits transmitted by the transmitting end; including:

[0046] The SCMA decoding operation outputs the initial log-likelihood ratio values of J receiving radio stations for channel decoding; in the selection of the channel decoding algorithm, select the serial list cancellation decoding algorithm;

[0047] In the SCL decoding algorithm, the received codeword undergoes multiple rounds of information update and decision-making processes. Each round utilizes the decision results of the previous round to improve the decoding performance. By continuously iterating and gradually canceling the already decoded information, the SCL algorithm gradually reduces the error probability and finally obtains the correct decoding result. The SCL algorithm process includes two steps: information update and path cancellation. In the information update stage, the algorithm updates the path metric value by calculating the soft information and hard information of the current node, and then selects a more reliable path. In the path cancellation stage, the algorithm selects a path to cancel based on the hard information of the current node and updates the metric values of its child nodes.

[0048] This application proposes an adaptive chirp spread spectrum modulation technology and integrates the PC-SCMA system based on polar codes and SCMA technology with the adaptive chirp spread spectrum modulation technology to achieve a radio adaptive joint coding modulation scheme applied to harsh channel environments. The key points of this scheme are:

[0049] An adaptive chirp spread spectrum technology is proposed. By real-time evaluating the state of the communication channel and selecting the most suitable spreading factor according to the channel quality, the performance and capacity of the system are improved.

[0050] Integrating the PC-SCMA system based on polar codes and SCMA technology with the adaptive chirp spread spectrum modulation technology realizes a radio adaptive joint coding modulation scheme applied to harsh channel environments, which can recover data bits in highly attenuated and interfered signals, adaptively increase the transmission rate when the channel quality improves, and achieve interconnection of a large number of radios while occupying less time-frequency resources. Description of the Drawings

[0051] Figure 1 It is a flowchart of the method provided by an embodiment of this application;

[0052] Figure 2 It is a flowchart of polar coding provided by an embodiment of this application;

[0053] Figure 3 It is a schematic diagram of SCMA non-orthogonal codeword superposition provided by an embodiment of this application;

[0054] Figure 4 It is a flowchart of chirp spread spectrum modulation provided by an embodiment of this application;

[0055] Figure 5 It is a factor graph of the allocation matrix provided by an embodiment of this application;

[0056] Figure 6 It is a flowchart of the MPA algorithm provided by an embodiment of this application;

[0057] Figure 7A graph showing how the spreading factor varies with SNR provided in an embodiment of the present application;

[0058] Figure 8 A schematic diagram of the effective information rate of the system provided in the embodiment of the present application;

[0059] Figure 9 A bit error rate diagram of a multi-user system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0060] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0061] The joint modulation and coding method proposed in this application for use in harsh channel environment scenarios innovatively proposes an adaptive chirp spread spectrum modulation technology, and cascades it with a sparse code division multiple access (Sparse Code Multiple Access, SCMA) access method based on polar codes (Polar Codes), thereby realizing a joint modulation and coding method. The method provided in this application includes:

[0062] The SCMA technology based on polarization code includes polarization channel coding and SCMA technology:

[0063] Channel coding is an important means to improve the quality of data transmission. Among many coding methods, polarization code is the only channel coding algorithm that has been proven to reach the Shannon limit capacity through rigorous mathematical methods. Its performance is better than RS code and convolutional code, and its complexity is lower. SCMA technology is a non-orthogonal multiple access technology proposed to meet the needs of 5G. Its biggest feature is that the number of accessed users can be multiple times greater than the number of resource blocks. SCMA technology serves more users while using the same resource blocks, thereby improving network throughput. At the same time, the sparsity of SCMA codewords also greatly improves the anti-interference ability of transmitted data. The joint design of polarization code and SCMA technology to build a PC-SCMA system can not only improve the communication quality of the system, but also obtain greater network throughput.

[0064] Adaptive chirp spread spectrum technology:

[0065] Chirp Spread Spectrum (CSS) is a typical low signal-to-noise ratio communication waveform. The chirp signal used has a high bandwidth and can accommodate reflected signals propagated by multiple paths. At the same time, its spectrum changes linearly in the frequency domain, and the influence of multipath signals is dispersed throughout the spectrum, so it has strong anti-multipath and anti-interference capabilities. In addition, chirp spread spectrum technology has the advantages of high signal processing efficiency and low transmission power, which is very suitable for the low-power design requirements of radio stations. The basic idea of ​​adaptive chirp modulation is to establish a feedback mechanism between the transmitter and the receiver through a reference signal known to both the transmitter and the receiver, and to exchange channel quality information in a timely manner. The receiver evaluates the quality of the wireless channel by receiving the reference signal and feeds it back to the transmitter. The transmitter automatically adjusts the chirp spread factor of the communication nodes at both ends according to the channel quality to improve the efficiency and performance of the system.

[0066] This joint modulation and coding scheme, when applied in harsh channel environments, will greatly improve the problem of poor communication performance and easy interruption under extremely low signal-to-noise ratios. At the same time, while ensuring the quality of data transmission, it will maximize the effective information rate of the system and provide theoretical and technical reserves for the development of radio station guaranteed communication technology.

[0067] This method supports the interconnection of multiple radio stations, including the transmitter and receiver;

[0068] Assume that there are J (J=6) user stations sending data at the transmitting end.

[0069] Step 11, compose the data to be sent and the reference signal into a data frame according to the frame format, and perform scrambling and interleaving processing to increase the anti-interference ability of the data;

[0070] Step 12: Polar encoder is performed on the processed data frame to introduce redundancy into the data sequence so that the receiving end can restore the source sequence with the smallest possible error probability.

[0071] Step 13, through the SCMA coding technology, the data is mapped from the bit domain to the K-dimensional complex domain codebook, and at the same time, the user codewords after SCMA coding are non-orthogonally superimposed on K orthogonal time-frequency resources; the mapping process is implemented through a pre-allocated multi-dimensional codebook, and different user data streams are distinguished according to subcarriers and codebooks.

[0072] Step 14, select a suitable spreading factor through the reference signal-to-noise ratio (SNR) that reflects the current channel quality fed back by the receiving end, perform adaptive chirp spread spectrum modulation on the subcarrier data, obtain a transmission signal with extremely strong anti-interference ability for transmission, and synchronize the updated spreading factor to the receiving node at the same time.

[0073] The data processing process at the receiving end is the opposite of that at the sending end, and the process is as follows:

[0074] Step 21, the receiving-end radio performs non-coherent demodulation after synchronizing the received chirp-modulated signal. Compared with coherent demodulation, non-coherent demodulation has much lower computational complexity and is an ideal choice for low-power and low-cost components.

[0075] Step 22, perform channel signal-to-noise ratio (SNR) estimation; extract the received signal sequence of the time-frequency resources where the reference signal is located and calculate its power. Since the receiving end has prior information about the reference signal, based on the calculated received signal power and the known reference signal power, the signal-to-noise ratio (Signal Noise Ratio, SNR) is obtained and fed back to the sending end.

[0076] Step 23, perform SCMA multi-user detection on the demodulated data information using a low-complexity and high-performance MPA (Massage passing algorithm) to output the soft decision of each user's information.

[0077] Step 24, perform polar code decoding based on the soft decision information and perform CRC check, and finally recover the original bits sent by the sending end to achieve stable and secure communication in a harsh channel environment.

[0078] The detailed technical solution of the application is as Figure 1 shown:

[0079] The execution process of the steps is specifically described below.

[0080] Sending end:

[0081] At the sending end, the data to be sent by the J user radios and the reference signal are formed into a data frame according to the frame format, and after processes such as interleaving, they are sent to the channel encoder for polar coding.

[0082] Step 11, perform scrambling and interleaving:

[0083] Step 12, polar coding:

[0084] The channel coding technology in this solution uses polar codes. After channel polarization design of N independent channels, N polarized channels are obtained. A part of the N polarized channels tend to be perfect channels with a channel capacity of 1, and another part tend to be noise channels with a channel capacity of 0; the information bits to be sent are transmitted on K sub-channels with high reliability; fixed bits are transmitted on the remaining N - K sub-channels, and the fixed bits are known for both encoding and decoding; the mixed information bits and fixed bits are combined to obtain u N =(u1, u2,..., u N ), representing the bit sequence to be encoded sent by the radio; G N is the generating matrix, and according to c = u NG N Obtain the encoded codeword c, where the output of the j-th user is The encoding process is as Figure 2 .

[0085] Set the length of the bit sequence to be encoded as 32, and the code rate supports [0.5, 0.6, 0.7, 0.8]. Different code rates correspond to different transmission rates.

[0086] Step 13, SCMA encoding:

[0087] The SCMA encoder maps the polarized-encoded binary bits to multi-layer constellation point symbols, and at the same time, the constellation point symbols are scattered on different subcarriers; the encoding mapping process is as follows:

[0088] Adopt the SCMA(J,K) system: The SCMA encoder is defined as: where M is the constellation point set, representing the types of user transmission bit combinations; K is the number of orthogonal subcarriers. Every log2(M) information bits form an information symbol, which is mapped to a K-dimensional complex domain codeword, and the mapping relationship is determined by the SCMA codebook; after mapping, the codewords of J users are non-orthogonally superimposed on K subcarriers, and K < J;

[0089] Determine the codebook by means of constellation rotation and interleaving, introduce the design principle of the lattice constellation, and determine the compact codebook of the multi-dimensional constellation by minimizing the average symbol energy of the minimum Euclidean distance between given constellation points. The codebook is a matrix with dimensions [K, M]; the codeword is generated through the following steps:

[0090] Step 131, modulation: Map the input log2(M) bits to multi-dimensional constellation modulation symbols, and perform constellation rotation and interleaving processing to generate the modulation symbol set S K ;

[0091] Step 132, sparse expansion: Expand the modulation symbol into a sparse codeword C K , C K = F K ·S K , F K is a sparse mapping matrix, which defines the non-zero positions of each user on the resource node and represents the allocation of carrier resources for each user;

[0092] The sparse mapping matrix F K is:

[0093]

[0094] As known from the sparse mapping matrix, for each user (i.e., each column), the number of subcarriers occupied (i.e., one row in the allocation matrix) is 2, and each subcarrier resource is used by 3 users simultaneously. This sparse pattern not only limits the number of branches of each resource node but also restricts the algorithm complexity at the receiving end and reduces the interference among users.

[0095] To further elaborate on this application, the following specifically elaborates on step 13 with specific examples.

[0096] An SCMA(6,4) system is adopted, where the number of users is 6 and the number of subcarriers is 4. A codebook is designed by a method based on constellation rotation and interleaving. The design principle of the lattice constellation is introduced, and a multi-dimensional constellation compact codebook design is carried out by minimizing the average symbol energy of the minimum Euclidean distance between given constellation points. The codebook is a matrix with dimensions [K, M], where K = 4, representing 4 orthogonal subcarriers; M = 4, log2(M) = 2, indicating that every 2 information bits form 1 information symbol. The information symbols represented by columns 1 to 4 are: 00, 01, 10, 11. The information symbols are mapped to the 4-dimensional complex domain, and the codebooks (CodeBook, CB) of the 6 users are as follows:

[0097]

[0098] Suppose the information bits transmitted by user 1 at a certain moment are 11, those of user 2 are 10, those of user 3 are 10, those of user 4 are 00, those of user 5 are 01, and those of user 6 are 11. The information bits of the 6 users are respectively encoded according to the corresponding codebooks to generate their respective complex-domain codewords (CodeWord, CW):

[0099]

[0100]

[0101] Then, the codewords of the 6 users are multiplexed on four orthogonal subcarriers. Among them, the subcarrier 1 transmits the superposition of the codewords of user 1, user 3, and user 5, the subcarrier 2 transmits the superposition of the codewords of user 1, user 4, and user 6, the subcarrier 3 transmits the superposition of the codewords of user 2, user 3, and user 6, and the subcarrier 4 transmits the superposition of the codewords of user 2, user 4, and user 5. The specific superposition process is as Figure 3 shown.

[0102] The allocation situation of SCMA for each user's carrier resources is represented by the allocation matrix F:

[0103]

[0104] As known from the allocation matrix, each user (column) occupies 2 subcarriers (rows), and each subcarrier resource can be used by 3 users simultaneously. This sparse pattern not only limits the number of branches of each resource node but also restricts the algorithm complexity at the receiving end and reduces the interference among users.

[0105] Currently, most radio stations adopt the TDMA multiple access mode. In a TDMA system, time is divided into periodic frames, and each frame is further divided into several non-overlapping time slots. Each radio station can only use the specified time slot within each frame to send the data to be transmitted, resulting in low transmission efficiency. The introduction of the SCMA multiple access mode enables multiple radio station data to be transmitted simultaneously without interference within the same time slot and the same frequency band resource, greatly improving the utilization rate of limited spectrum resources and thus increasing the data transmission rate.

[0106] Step 14, Adaptive chirp spread spectrum modulation:

[0107] The complex sequence after coding adopts adaptive chirp spread spectrum modulation. Before modulation, the current channel quality is evaluated according to the SNR fed back by the receiving end, and the spreading factor suitable for the current channel conditions is dynamically selected; the definition of extremely low SNR varies in different standards and studies. For example, in some standards of the Telecommunication Standardization Sector of the International Telecommunication Union (ITU-T), the SNR below 12 dB is defined as low SNR, and in some specific cases, such as in a strong noise environment, the SNR may be as low as -6 dB or even lower. In this application, the extremely low SNR range is below -5 dB.

[0108] When the SNR is low, a larger spreading factor, i.e., SF > 8, is selected to improve the anti-interference ability and give priority to ensuring the transmission quality; when the SNR is high, a smaller spreading factor, i.e., SF < 8, is selected to maximize the effective information transmission rate on the premise of ensuring the transmission quality. The SNR is used as the spreading factor switching threshold, and the corresponding relationship is:

[0109] Reference Signal Signal-to-Noise Ratio Spreading Factor SNR < -23 SF = 15 -23 ≤ SNR < -21.5 SF = 14 -21.5 ≤ SNR < -20 SF = 13 -20 ≤ SNR < -18.5 SF = 12 -18.5 ≤ SNR < -17 SF = 11 -17 ≤ SNR < -15.5 SF = 10 -15.5 ≤ SNR < -14 SF = 9 -14 ≤ SNR < -12.5 SF = 8 -12.5 ≤ SNR < -11 SF = 7 -11 ≤ SNR < -9.5 SF = 6 -9.5 ≤ SNR < -8 SF = 5 -8 ≤ SNR < -6.5 SF = 4 -6.5 ≤ SNR < -5 SF = 3 SNR ≥ -5 SF = 2

[0110] The process of adaptive chirp spread spectrum modulation is as Figure 4 . Before modulation, the complex sequence after SCMA coding is first digitally processed into a binary bit sequence f(n) through the Matlab function f(n)=dec2bin(abs(complex sequence), λ), where n ∈ [0, M - 1[, and M = 2 λ , and λ is the spreading factor SF, that is, the number of binary bits transmitted by each modulation symbol; after the processing is completed, chirp spread spectrum modulation is performed: in the discrete time domain, the CSS symbol composed of M sampling values is expressed as S(n)=f(n)c u (n), where f(n) is the information symbol, expressed as cu (n) is an upward chirp symbol, denoted as

[0111] Receiving end:

[0112] Step 21, Receiver non-coherent detection:

[0113] When the chirp-modulated spread-spectrum signal is transmitted through the RF antenna, due to the unknown channel state information, the receiving-end radio uses a non-coherent detection mechanism to demodulate the chirp spread-spectrum signal. Considering that the information-bearing element of the CSS symbol is the value of the spreading factor FS, therefore, non-coherent detection involves the estimation of k. With non-coherent detection, the transmitted spreading factor FS is estimated as:

[0114]

[0115] where y(n) is the signal received by the radio, is the local conjugate unmodulated chirp signal of the receiving radio; for the FS estimation using non-coherent detection, first estimate the DFT of the product signal , and the estimation result is R(k); then, estimate the absolute value of R(k) and find the frequency index with the maximum peak.

[0116] Step 22, SNR estimation:

[0117] Extract the received signal sequence of the time-frequency resource where the reference signal is located and count the power. Since the receiving end has the prior information of the reference signal, based on the calculated received signal power and the known reference signal power, the signal-to-noise ratio (SignalNoiseRatio, SNR) is obtained and fed back to the transmitting end.

[0118] Step 23, SCMA multi-user detection:

[0119] The SCMA system decodes the received data through a multi-user detection algorithm to obtain the data before SCMA coding. This scheme uses the MPA algorithm with lower complexity for multi-user detection; from the SCMA coding, it can be known that the mapping of SCMA users to carrier resources can be determined by the resource allocation matrix F, (F) kj = 1 indicates that user j occupies subcarrier k, conversely, (F) kj = 0 indicates that user j does not occupy subcarrier k; regard the users as variable nodes, represented by circles, and the carriers as function nodes, represented by squares, and connect the variable nodes and function nodes with (F) kj = 1 with lines to obtain the factor graph of the allocation matrix F as Figure 5 :

[0120] The MPA algorithm updates the prior information by transmitting extrinsic information between variable nodes and function nodes. Assume that initially, the probabilities of all symbols are equal. First, the conditional probabilities and prior probabilities are initialized, and then the variable nodes and function nodes are continuously updated through iteration. Finally, the log-likelihood ratio (LLR) of the transmission probability of each codeword of the variable node is calculated and input to the channel decoder.

[0121] Step 24, polar decoding:

[0122] The SCMA decoding operation outputs the initial log-likelihood ratio values of J receiving stations for channel decoding. In terms of the selection of the channel decoding algorithm, the Successive Cancellation List (SCL) decoding algorithm is selected. It is an iterative decoding algorithm used to correct linear block codes. In the SCL decoding algorithm, the received codeword undergoes multiple rounds of information update and decision-making processes. Each round utilizes the decision results of the previous round to improve the decoding performance. By continuously iterating and gradually canceling the already decoded information, the SCL algorithm gradually reduces the error probability and finally obtains the correct decoding result. The SCL algorithm process includes two steps: information update and path cancellation. In the information update stage, the algorithm updates the path metric value by calculating the soft information and hard information of the current node, and then selects a more reliable path. In the path cancellation stage, the algorithm selects a path to cancel according to the hard information of the current node and updates the metric values of its child nodes. Through this iterative method, the SCL algorithm can continuously improve the decoding performance and achieve a decoding result close to the optimal under the premise of ensuring a certain computational complexity.

[0123] This application first innovatively proposes the adaptive chirp spread spectrum technology. By real-time evaluating the state of the communication channel and selecting the most suitable spreading factor according to the channel quality, the performance and capacity of the system are improved. Secondly, the PC-SCMA system based on polar codes and SCMA technology is integrated with the adaptive chirp spread spectrum modulation technology to implement a radio adaptive joint coding modulation scheme applied to harsh channel environments. It can recover data bits from highly attenuated and interfered signals and adaptively increase the transmission rate when the channel quality improves, while achieving interconnection of a large number of radios with less time-frequency resources.

[0124] The present invention is simulated in Matlab and compared with a non-adaptive joint coding modulation scheme with a fixed spreading factor to prove the improvement of the system performance by adaptivity. In the non-adaptive joint coding modulation scheme, the spreading factor of the chirp spread spectrum modulation is fixed and does not dynamically adjust with the change of the channel quality. Other link modules are exactly the same as those of the present invention. Denote this adaptive joint coding modulation scheme as Scheme 1 and the non-adaptive joint coding modulation scheme as Scheme 2.

[0125] Let the original sequence sent by each radio station be a random bit sequence with \(L = 1024\), and the reference signal be a ZC sequence with \(L = 128\). Dynamically adjust the signal-to-noise ratio of the wireless channel from -25 dB to 0 dB to represent a strong noise environment. The data frame composed of the reference signal and the sequence to be sent is subjected to Polar channel coding according to \((64,128)\) and \((6,4)\) SCMA coding, and then the encoded complex sequence is binary digitized and chirp spread spectrum modulation is performed. In Scheme 1, the receiving end estimates the channel signal-to-noise ratio and feeds it back to the sending end, and dynamically adjusts the spreading factor according to the signal-to-noise ratio. In Scheme 2, the spreading factor \(SF = 10\) is fixed. ref A comparison chart of the switching situation of the spreading factor with the change of the channel SNR under the two schemes is given. Figure 7 A comparison chart of the switching situation of the spreading factor with the change of the channel SNR under the two schemes is given. Figure 8 The change of the effective information rate of the system with the spreading factor in the two schemes is given. Figure 9 The BER of 6 radio station users under different signal-to-noise ratios in the two schemes is given.

[0126] Take the SNR corresponding to \(BER = 10\) -2 as the operating point. From Figure 9 it can be obtained that the operating point of Scheme 1 is \(SNR=-17dB\), the corresponding spreading factor is 11, and the effective transmission rate is 2.9 kbps; the operating point of Scheme 2 is \(SNR=-16.5dB\), the corresponding spreading factor is 10, and the corresponding effective transmission rate is 5 kbps. When \(SNR=-16\), for both schemes with \(SF = 10\), correct decoding can be achieved and the effective transmission rates are the same. When \(SNR > -16\), the spreading factor of Scheme 1 decreases as the channel condition improves, and the corresponding effective transmission rate increases significantly, while the effective transmission rate of Scheme 2 remains unchanged at 5 kbps. When the signal-to-noise ratio is greater than -5, the effective transmission rate of Scheme 1 reaches 230 kbps, which is 46 times that of Scheme 2.

[0127] It can be seen from the simulation that the adaptive joint coding and modulation scheme can achieve guaranteed communication at a low rate at extremely low signal-to-noise ratios, and can achieve a high communication rate when the signal-to-noise ratio improves, meeting the high-speed transmission of data services such as videos.

[0128] The above-described embodiments of the present application do not constitute a limitation on the protection scope of the present application.

Claims

1. A radio station adaptive joint coding modulation method applied to a harsh channel environment, characterized in that: The method supports the interconnection of multiple radio stations, including a transmitting end and a receiving end; Assume that there are J user stations sending data at the transmitting end; The data processing process at the sending end is as follows: Step 11, compose the data to be sent and the reference signal into a data frame according to the frame format, and perform scrambling and interleaving processing to increase the anti-interference ability of the data; Step 12, performing polarization coding on the processed data frame to introduce redundancy into the data sequence; Step 13, using SCMA coding technology, the data is mapped from the bit domain to a K-dimensional complex domain codebook, and the SCMA-coded user codewords are non-orthogonally superimposed on K orthogonal time-frequency resources; the mapping process is implemented through a pre-assigned multi-dimensional codebook, and different user data streams are distinguished according to subcarriers and codebooks; Step 14, selecting a suitable spreading factor through the reference signal noise ratio reflecting the current channel quality fed back by the receiving end, performing adaptive chirp spreading modulation on the subcarrier data, obtaining a transmission signal for transmission, and synchronizing the updated spreading factor to the receiving node at the same time; The data processing process at the receiving end is as follows: Step 21, the receiving end station performs non-coherent demodulation on the received chirp modulation signal after synchronization; Step 22, perform channel signal-to-noise ratio estimation: extract the received signal sequence of the time-frequency resource where the reference signal is located and count its power. Since the receiving end has prior information about the reference signal, the signal-to-noise ratio is obtained based on the calculated received signal power and the known reference signal power and fed back to the transmitting end; Step 23, using MPA to perform SCMA multi-user detection on the demodulated data information, and outputting soft decisions of each user information; Step 24: perform polar code decoding according to the soft decision information, and perform CRC check to finally restore the original bits sent by the transmitter.

2. The method according to claim 1, characterized in that Step 12, performing polarization coding on the processed data frame to introduce redundancy into the data sequence, comprising: After the channel polarization design, N independent channels are obtained to obtain N polarized channels. Some of the N polarized channels have a channel capacity of 1, and the other part tends to be a noise channel, that is, the channel capacity is 0. The information bits to be sent are transmitted on K sub-channels with high reliability. Fixed bits are transmitted on the remaining NK sub-channels, and the fixed bits are known to both the encoding and decoding. The information bits and fixed bits are mixed to obtain u N =(u1,u2,...,u N ), represents the bit sequence to be encoded sent by the radio station; G N To generate the matrix, according to c = u N G N Get the encoded codeword c, where the jth user output is The length of the bit sequence to be encoded is set to 32, and the bit rate supports [0.5, 0.6, 0.7, 0.8]. Different bit rates correspond to different transmission rates.

3. The method according to claim 1, characterized in that Step 13, the SCMA encoding process is as follows: The SCMA encoder maps the polarization-coded binary bits to multiple layers of constellation point symbols, and the constellation point symbols are dispersed to different subcarriers. The coding and mapping process is as follows: Using the SCMA(J,K) system: the SCMA encoder is defined as: Where M is the constellation point set, which indicates the number of types of bit combinations sent by the user; K is the number of orthogonal subcarriers, and each log2(M) information bit constitutes an information symbol, which is mapped to a K-dimensional complex domain codeword. The mapping relationship is determined by the SCMA codebook; after mapping, J user codewords are non-orthogonally superimposed on K subcarriers, and K<J; The codebook is determined by a constellation rotation and interleaving method, and the design principle of the lattice constellation is introduced. The compact codebook of the multidimensional constellation is determined by minimizing the average symbol energy of the minimum Euclidean distance between given constellation points. The codebook is a matrix with a dimension of [K, M]. The codeword is generated by the following steps: Step 131, modulation: Map the input log2(M) bits to multidimensional constellation modulation symbols, perform constellation rotation and interleaving, and generate a modulation symbol set S K ; Step 132, sparse expansion: expand the modulation symbol into a sparse codeword C K ,C K =F K ·S K ,F K It is a sparse mapping matrix, which defines the non-zero position of each user on the resource node and represents the allocation of carrier resources to each user; Sparse mapping matrix F K for: It is known from the sparse mapping matrix that the subcarriers occupied by each user, that is, one column, that is, one row in the allocation matrix is ​​2, and each subcarrier resource is used by three users at the same time.

4. The method according to claim 1, characterized in that: Step 14, adaptive chirp spread spectrum modulation: comprising: Before modulation, the current channel quality is evaluated based on the SNR fed back by the receiving end, and the spreading factor suitable for the current channel conditions is dynamically selected; the extremely low signal-to-noise ratio range is below -5dB; When the SNR is low, a larger spreading factor, that is, SF>8, is selected to improve the anti-interference capability; when the SNR is high, a smaller spreading factor, that is, SF<8, is selected; using SNR as the spreading factor switching threshold, the corresponding relationship is: Before modulation, the complex sequence encoded by SCMA is digitized into a binary bit sequence f(n) by Matlab f(n)=dec2bin(abs(complex sequence), λ) function, where n∈[0,M-1], and M=2 λ , λ is the spreading factor SF, that is, the number of binary bits transmitted by each modulation symbol; after the processing is completed, chirp spread spectrum modulation is performed: in the discrete time domain, the CSS symbol composed of M sampling values ​​is represented by S(n) = f(n)c u (n), where f(n) is the information symbol, expressed as c u (n) is the upward chirp symbol, expressed as 5. The method according to claim 1, characterized in that Step 21, the receiving end station performs non-coherent demodulation on the received chirp modulation signal after synchronization, including: Using non-coherent detection, the spreading factor FS of the transmission is estimated as: Where y(n) is the signal received by the radio station, and s(n) is the local conjugate unmodulated linear frequency modulation signal of the receiving radio station; for FS estimation using non-coherent detection, the DFT of the product signal r(n)=y(n)s(n) is first estimated, and the estimated result is R(k); then, the absolute value of R(k) is estimated to find the frequency index with the largest peak.

6. The method according to claim 1, characterized in that Step 22, estimating the channel signal-to-noise ratio, includes: The received signal sequence of the time-frequency resource where the reference signal is located is extracted and the power is counted. Since the receiving end has prior information about the reference signal, the signal-to-noise ratio is obtained based on the calculated received signal power and the known reference signal power and fed back to the sending end.

7. The method according to claim 1, characterized in that Step 23, using MPA to perform SCMA multi-user detection on the demodulated data information, and outputting soft decisions of each user information, including: MPA algorithm is used for multi-user detection. From SCMA coding, we can know that the mapping of SCMA users to carrier resources can be determined by the resource allocation matrix F, (F) kj =1 means user j occupies subcarrier k, otherwise, (F) kj = 0 means that user j does not occupy subcarrier k; consider the user as a variable node, represented by a circle, and the carrier as a function node, represented by a square. kj = 1, the variable nodes and function nodes are connected by lines to obtain the factor graph of the allocation matrix F; The MPA algorithm updates the prior information by passing external information between variable nodes and function nodes; it assumes that the probability of all symbols appearing is equal at the beginning; first, the conditional probability and prior probability are initialized, and then the variable nodes and function nodes are continuously updated through iteration; finally, the log-likelihood ratio is calculated by the transmission probability of each codeword of the variable node and input into the channel decoder.

8. The method according to claim 1, characterized in that Step 24, decoding the polar code according to the soft decision information, and performing CRC check, and finally recovering the original bits sent by the transmitter; including: The SCMA decoding operation outputs the initial log-likelihood ratio values ​​of J receiving stations for channel decoding; in the selection of the channel decoding algorithm, a serial list cancellation decoding algorithm is selected; In the SCL decoding algorithm, the received codewords go through multiple rounds of information updates and decision-making processes, and each round uses the decision results of the previous round to improve the decoding performance. Through continuous iteration and gradual cancellation of the information that has been decoded, the SCL algorithm gradually reduces the error probability and finally obtains the correct decoding result; the SCL algorithm process includes two steps: information update and path cancellation: in the information update stage, the algorithm updates the path metric value by calculating the soft information and hard information of the current node, and then selects a path with higher credibility; in the path cancellation stage, the algorithm selects a path to cancel based on the hard information of the current node, and updates the metric value of its child nodes at the same time.

Images