Efficient coding linear frequency modulation transmission method for satellite Internet of Things communication

Abstract

The invention provides an efficient coding linear frequency modulation transmission method for satellite Internet of Things communication, which comprises the following steps of: sequentially carrying out coding and symbol mapping on a data bit stream generated by user information to obtain a symbol sequence; performing improved linear frequency modulation with a spreading factor of SF on the symbol sequence to obtain a linear frequency modulation signal; processing the linear frequency modulation signal through a quasi-additive Gaussian white noise channel to obtain a received signal; performing de-chirp operation and soft decision incoherent demodulation on the received signal in sequence to obtain soft decision information; obtaining the ideal channel capacity of the coding LFM system according to the soft decision information, and obtaining the approximate channel capacity of the coding LFM system by using a Taylor expansion approximation method; obtaining a corresponding reachable code rate according to the ideal channel capacity and the approximate channel capacity, and establishing a corresponding relationship between a bit signal-to-noise ratio and the channel capacity; and soft input and soft output decoding is carried out according to the soft decision information, and the original data bit stream is recovered. According to the invention, the problem of extremely low signal-to-noise ratio in satellite Internet of Things communication can be effectively solved, so that reliable user transmission is provided.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite Internet of Things communication, and particularly to an efficient coding linear frequency modulation transmission method. Background Art

[0002] At present, the rapid development of 5G communication technology is leading the Internet of Things communication system to a new height, bringing disruptive application prospects to many fields and industries. However, in special geographical environments such as ocean areas, vast deserts and remote regions, complex terrain and harsh climate conditions pose great challenges to the deployment and coverage of ground Internet of Things facilities, and ground communication facilities are vulnerable to natural disasters such as earthquakes, tsunamis and lightning, resulting in damage to communication facilities and communication interruption. In future 6G communication, the satellite Internet of Things aiming to achieve integration of air, space, land and sea will achieve global seamless coverage, providing a practical solution for Internet of Things communication in complex environments. However, the ultra-long-distance communication characteristics of the satellite Internet of Things bring serious link loss problems, resulting in very weak received signal power, thus posing higher requirements for the reliability of the communication system. Therefore, efficient coding and spread spectrum modulation technologies have become the key common technologies for reliable communication of the satellite Internet of Things. Among many spread spectrum modulation technologies, LoRa (Long Range) - an improved linear frequency modulation (LFM) technology - has become one of the mainstream technologies in the global Internet of Things field due to its characteristics such as low power consumption, wide coverage and large connection. Based on this improved LFM technology, studying its efficient transmission scheme has potential value for reliable communication of future satellite Internet of Things.

[0003] Most of the related research on this improved LFM technology uses Hamming code as the coding scheme. However, the error correction ability and coding gain of Hamming code are very limited, making it difficult to meet the reliable communication requirements of satellite Internet of Things (IoT) under extremely low signal-to-noise ratio (SNR). Therefore, researchers have started to attempt to combine high-efficiency coding (such as Turbo code) with the improved LFM technology, and screen out candidate codes suitable for satellite communication by comparing bit error performance, decoding threshold, and implementation complexity. However, the demodulation method of this improved LFM technology is hard decision demodulation or partial soft decision demodulation, so it cannot directly or better adapt to the decoding methods of high-efficiency coding. Baruffa et al. quantitatively analyzed the non-coherent detection and its error probability of the signals of the improved LFM technology in an additive white Gaussian noise (AWGN) channel in the paper "Coded LoRa performance in wireless channels", using a fixed code rate, which leads to deteriorated performance of the system under low SNR. Xu et al. provided an orthogonal binary peak difference decision method in the paper "Improving method of LoRa based on Turbo code and ODPD", converting the decision result into soft information and inputting it into the Turbo decoder. Therefore, compared with the traditional Hamming-coded LFM system, its bit error performance has been improved to a certain extent, but this method cannot generate true "soft" decision values. Fabregas et al. proposed a soft decision detection algorithm suitable for coded frequency shift keying systems in the paper "Capacity approaching codes for non-coherent orthogonal modulation", and the soft decision values generated by it have high quality, but the computational complexity of this algorithm is relatively high. In addition, the coding schemes used in the above-mentioned literatures do not consider the optimal code rate corresponding to the soft decision detection of this improved LFM. Therefore, it is necessary to deeply analyze the optimal code rate problem of the coded LFM system to optimize the system performance. In short, in the face of the problem of extremely low SNR in satellite IoT communication, the existing technologies are difficult to ensure highly reliable communication. Summary of the Invention

[0004] Aiming at the technical problem that it is difficult to ensure reliable communication when the signal-to-noise ratio is extremely low in the long-distance communication of satellite Internet of Things (IoT), the present invention proposes an efficient coding linear frequency modulation transmission method for satellite IoT communication, which enhances the reliable transmission ability of the system in an extremely low SNR environment by combining high-efficiency coding with the improved linear frequency modulation (LFM) technology, and introduces an approximate metric method to reduce the computational complexity of the system.

[0005] To achieve the above object, the technical solution of the present invention is realized as follows:

[0006] An efficient coded linear frequency modulation transmission scheme for satellite Internet of Things communication, comprising the following steps:

[0007] S1: Sequentially encode and symbol map the data bit stream generated by user information to obtain a symbol sequence {d k};

[0008] S2: Perform improved LFM modulation with a spreading factor of SF on the symbol sequence {d k} to obtain a linear frequency modulation signal s k ;

[0009] S3: Obtain a received signal r k through a quasi-additive Gaussian white noise channel for the linear frequency modulation signal s k ;

[0010] S4: Sequentially perform de-chirping operation and soft-decision non-coherent demodulation on the received signal r k to obtain soft-decision information Λ(c k,j );

[0011] S5: Obtain the ideal channel capacity C opt of the coded LFM system according to the soft-decision information and obtain the approximate channel capacity C app of the coded LFM system by using the first-order approximation method of Taylor expansion;

[0012] S6: Respectively obtain the corresponding achievable code rates R opt and R app according to the ideal channel capacity C opt and the approximate channel capacity C app , and establish the relationship between the bit signal-to-noise ratio and the channel capacity C opt or C app ;

[0013] S7: Perform soft-input soft-output decoding according to the soft-decision information to recover the original data bit stream.

[0014] The encoding method described in step S1 is:

[0015] The encoding method uses Turbo coding. The data bit stream generated by user information is represented as {b i}, and the binary coding sequence obtained by Turbo coding is {c j}; Turbo coding is used to improve the error correction ability at extremely low signal-to-noise ratios;

[0016] The method of the symbol mapping is:

[0017] Map every SF bits of the binary coding sequence {c j} after Turbo coding to a symbol sequence to obtain a decimal symbol sequence c k,j represents the j-th bit in the k-th decimal symbol sequence, k represents the symbol period index, and SF represents the spreading factor.

[0018] The method for performing improved LFM modulation with a spreading factor of SF on the symbol sequence described in step S2 is as follows: Using the modulation signal acquisition formula to perform linear frequency modulation with a spreading factor of SF on the symbol sequence {d k} to obtain the linear frequency modulation signal sequence s k} of s k = {s k (l)}, s k represents the modulation signal transmitted within the k-th symbol period of the symbol sequence {d k}, and the modulation signal acquisition formula is:

[0019]

[0020] where s k (l) represents the modulation signal at the l-th position within the k-th symbol period, k = 0, 1,..., N - 1 represents the symbol period index, N is the transmission length of the symbol sequence {d k}, l = 0, 1,..., 2 SF - 1 represents the chirp index, M = 2 SF is the number of chirps of the symbol sequence {d k}, θ k is the phase offset, is the imaginary unit.

[0021] The method for obtaining the received signal r k described in step S3 is as follows:

[0022] r k = s k + n k ;

[0023] In the formula, r k = {r k (0),..., r k (l),..., r k (M - 1)} represents the received signal sequence within the k-th symbol period, r k (l) represents the l-th received signal within the k-th symbol period, n k = {n k (0),..., n k (l),..., n k (M - 1)} represents the additive white Gaussian noise signal sequence within the k-th symbol period, n k(l) represents the l-th additive white Gaussian noise signal within the k-th symbol period;

[0024] The method of the de-chirping operation described in step S4 is as follows:

[0025] For the l-th received signal r k (l) = s k (l) + n k (l) within the k-th symbol period of the received signal rk, multiply it by a down-chirping signal to obtain the de-chirped signal z k (l) as:

[0026]

[0027] The method of the soft decision non-coherent demodulation is as follows:

[0028] (1), Perform an M-point discrete Fourier transform on the de-chirped signal z k (l) to obtain the frequency-domain signal Z k (q);

[0029] (2), Calculate the log-likelihood probability of the received signal according to the frequency-domain signal Z k (q);

[0030] (3), Calculate the demodulation log-likelihood ratio of the i-th bit c k} of the transmitted symbol sequence {d k,i} as the soft decision information Λ(c k,i ).

[0031] The frequency-domain signal Z k (q) is:

[0032]

[0033] where q is the frequency index used to identify different frequency points in the frequency domain;

[0034] The log-likelihood probability is:

[0035]

[0036] where I0(·) is the modified Bessel function of the first kind of order zero, used to obtain the definite metric when q = d k ; is the modulus value of the frequency-domain signal, is the noise variance;

[0037] The soft decision information is:

[0038]

[0039] Among them, Λ(c k,i ) is the soft decision value encoding the LFM signal, and i = 0, 1,..., SF - 1 represents the binary bit stream {c k} of the symbol sequence {d k,0 , c k,1 , …, c k,SF-1} that makes up the transmitted data, where the index of bit c k,i is i, and respectively represent the index set where the i-th position is 0 and the index set where the i-th position is 1 in the binary bit representation corresponding to the decimal numbers 0 to M - 1 mapped according to the natural mapping.

[0040] The method for obtaining the approximate channel capacity C app of the encoded LFM system using the first-order approximation method of Taylor expansion in step S5 is as follows:

[0041] (1) Based on the ideal metric and the determined metric in the soft decision information, obtain the ideal channel capacity C opt of the encoded LFM system;

[0042] (2) Perform a first-order Taylor expansion of the first-kind zero-order modified Bessel function at x = 0 and estimate the determined metric and the ideal metric according to the first-order Taylor expansion;

[0043] (3) According to the estimation result, obtain the approximate channel capacity C app .

[0044] The ideal channel capacity C opt is as follows:

[0045]

[0046] The first-order Taylor expansion of the Bessel function at x = 0 is as follows:

[0047]

[0048] The method for estimating the determined metric and the ideal metric is as follows:

[0049]

[0050] where is the ideal metric;

[0051] The calculation method of the approximate channel capacity C app is as follows:

[0052]

[0053] Among them, is the conditional expectation operator, indicating that the expectation of the received signal r k is taken under the condition of q = d k .

[0054] The method of obtaining the corresponding achievable code rates R opt and R app respectively according to the ideal channel capacity C opt and the approximate channel capacity C app , and establishing the relationship between the bit signal-to-noise ratio and the channel capacity C opt or C app is as follows:

[0055] The method of obtaining the corresponding achievable code rates R opt and R app respectively according to the ideal channel capacity C opt and the approximate channel capacity C app is as follows:

[0056] or

[0057] The method of establishing the relationship between the bit signal-to-noise ratio and the channel capacity C opt or C app is as follows:

[0058] Given the spreading factor SF and the symbol signal-to-noise ratio E s / N0, calculate the relationship between the bit signal-to-noise ratio E b / N0 and the channel capacity:

[0059] or

[0060] Among them, E s is the symbol energy, E b is the bit energy, and N0 represents the unilateral power density.

[0061] The method of performing soft-input soft-output decoding according to the soft decision information in step S7 to recover the original data bit stream is as follows: Input the soft decision information Λ(c k,i ) into the soft decision decoder, use the BCJR algorithm to decode the Turbo code, and convert the decoded codeword into the original data bit stream through symbol demapping.

[0062] Advantages of the present invention:

[0063] 1. Enhanced error correction ability and coding gain: The introduction of efficient coding improves the error correction ability of the system, enabling the system to adapt to a lower signal-to-noise ratio environment. FromFigure 4 and Figure 5 It can be seen that the bit error performance of the Turbo-LFM system is 8 - 10 dB better than that of the Hamming-LFM system, that is, the highly efficient coded Turbo code can adapt to a lower signal-to-noise ratio.

[0064] 2. Improve communication reliability: By combining highly efficient coding with improved linear frequency modulation (LFM) technology, the reliable transmission ability of the system in a low signal-to-noise ratio environment is enhanced, ensuring stable communication in a complex satellite Internet of Things environment.

[0065] 3. Reduce complexity: By introducing an approximate metric method, the present invention simplifies the complex calculation of traditional Bessel functions and reduces the computational complexity of the system. This improvement makes the scheme easier to implement in real time, especially in satellite Internet of Things devices with limited resources, improving the practicality and deployment feasibility of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0067] Figure 1 It is a transmission system model under satellite Internet of Things based on Turbo-LFM technology used in the present invention.

[0068] Figure 2 It is a relationship curve between signal-to-noise ratio and achievable code rate based on ideal metric in the present invention.

[0069] Figure 3 It is a relationship curve between signal-to-noise ratio and achievable code rate based on approximate metric in the present invention.

[0070] Figure 4 It is the performance of Hamming-LFM and Turbo-LFM transmission systems based on ideal metric and approximate metric when SF = 7 in the present invention.

[0071] Figure 5 It is the performance of Hamming-LFM and Turbo-LFM transmission systems based on ideal metric and approximate metric when SF = 12 in the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0072] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0073] An efficient coding linear frequency modulation transmission method for satellite Internet of Things communication, as Figure 1 shown, includes the steps:

[0074] S1: Sequentially encode and symbol map the data bit stream generated by user information to obtain a symbol sequence {d k}.

[0075] The encoding method used is (768,384), (1152,384), and (1920,384) Turbo encoding. The data bit stream generated by user information is represented as {b i}, and the binary encoding sequence obtained by Turbo encoding is {c j}; Turbo encoding is used to improve the error correction ability under extremely low signal-to-noise ratios.

[0076] The method of symbol mapping is as follows:

[0077] Map every SF bits of the binary encoding sequence {c j} after Turbo encoding to a symbol sequence to obtain a decimal symbol sequence {d k}:

[0078]

[0079] where c k,j represents the jth bit in the kth symbol sequence, k represents the symbol period index, and SF is the spreading factor.

[0080] S2: Perform improved LFM modulation with a spreading factor of SF on the symbol sequence {d k} to obtain a linear frequency modulation signal s k .

[0081] The method of performing improved LFM modulation with a spreading factor of SF on the symbol sequence is as follows:

[0082] Use the modulation signal acquisition formula to perform improved LFM modulation with a spreading factor of SF on the symbol sequence {d k} to obtain a linear frequency modulation signal sequence s k = {s k = {s k(l)} represents the modulation signal transmitted within the k-th symbol period of the symbol sequence {d k} and the formula for obtaining the modulation signal is:

[0083]

[0084] where k = 0, 1,..., L - 1 is the symbol period index, L is the transmission length of the symbol sequence {d k}, l = 0, 1,…, 2 SF -1 represents the chirp index, s k (l) represents the l-th modulation signal within the k-th symbol period, M = 2 SF is the number of chirps of the symbol sequence {d k}, θ k is the phase offset, is the imaginary unit.

[0085] S3: The chirp signal s k passes through a quasi-additive Gaussian white noise channel to obtain the received signal r k .

[0086] The method for obtaining the received signal is:

[0087] r k = s k + n k ;

[0088] where r k = {r k (0),..., r k (l),..., r k (M - 1)} represents the received signal sequence within the k-th symbol period, r k (l) represents the l-th received signal within the k-th symbol period, n k = {n k (0),..., n k (l),..., n k (M - 1)} represents the additive Gaussian white noise signal sequence within the k-th symbol period, n k (l) represents the l-th additive Gaussian white noise signal within the k-th symbol period.

[0089] S4: The received signal r k is successively subjected to de-chirping operation and soft decision non-coherent demodulation to obtain the soft decision information Λ(c k,j ).

[0090] The method for the de-chirping operation is:

[0091] The l-th received signal r in the k-th symbol period of the received signal rk k (l) = s k (l) + n k (l) is multiplied by a down-chirp signal to obtain the de-chirped signal z k (l):

[0092]

[0093] wherein, is still Gaussian noise, and its statistical characteristics are consistent with n k (l).

[0094] The method of soft decision non-coherent demodulation is as follows:

[0095] (1), Perform M-point discrete Fourier transform on the de-chirped signal z k (l) to obtain the frequency-domain signal Z k (q):

[0096]

[0097] where q is the frequency index, used to identify different frequency points in the frequency domain, is the DFT result of the noise term , and δ(·) is the Kronecker delta function.

[0098] (2), According to the frequency-domain signal Z k (q), calculate the log-likelihood probability of the received signal:

[0099]

[0100] where r k = {r k (l)} is the received signal sequence in the k-th symbol period, I0(·) is the modified Bessel function of the first kind of zero order (Bessel function), which converts the relationship between the amplitude of the frequency-domain signal and the noise into a metric value to measure the probability characteristics of the received signal under specific conditions, is the modulus value of the frequency-domain signal, reflecting the amplitude information of the frequency-domain signal. In the formula, the first term is called the normalization constant independent of the frequency index q, and the second term is an important metric related to the frequency index q, is the noise variance.

[0101] The log-likelihood probability is used in the communication system to measure the possibility of the received signal appearing under a specific hypothesis, and is an important calculation link in soft decision non-coherent demodulation, providing a basis for subsequent calculations such as demodulation log-likelihood ratio.

[0102] (3), According to the log-likelihood probability of the received signal, calculate the transmitted data symbol dk the \(i\)-th bit \(c\) k,i of the demodulated log-likelihood ratio (LLR), i.e., the soft decision value:

[0103]

[0104] where \(\Lambda(c\) k,i ) is the soft decision value of the encoded LFM signal. When its value is greater than 0, the probability that the bit is "1" is relatively large; when it is less than 0, the probability that the bit is "0" is relatively large. \(i = 0, 1, \cdots, SF - 1\) represents the symbol sequence \(d\) k of the transmitted data, and the binary bit stream \(\{c\) k,0 , c\) k,1 , \cdots, c\) k,SF-1 \}, and the index of the bit \(c\) k,i . and respectively represent the index set where the \(i\)-th position is 0 and the index set where the \(i\)-th position is 1 in the binary bit representation corresponding to the decimal numbers 0 to \(M - 1\) according to the natural mapping.

[0105] When the improved LFM technology uses the Hamming code as the coding scheme, it performs hard decision non-coherent demodulation, mainly based on the amplitude or phase of the signal, ignoring the soft information of the signal (such as signal-to-noise ratio, likelihood ratio, etc.), resulting in a large amount of information loss. The orthogonal binary peak difference decision method generates soft information only based on the relative difference in signal amplitude, and does not fully quantify the confidence of the signal, resulting in the decoder being unable to effectively use the soft information for iterative optimization. The soft decision non-coherent demodulation adopted by the present invention outputs the soft information of each symbol or bit, reflecting its correct probability (such as the log-likelihood ratio LLR). These soft information can be directly input into the decoder of an efficient code (such as Turbo code) to improve the decoding performance and reduce the bit error rate.

[0106] S5: Obtain the ideal channel capacity \(C\) k,i of the encoded LFM system according to the soft decision information \(\Lambda(c\) opt and the approximate channel capacity \(\widetilde{C}\) app of the encoded LFM system obtained by using the first-order approximation method of Taylor expansion.

[0107] The method for obtaining the approximate channel capacity of the encoded LFM system by using the Taylor expansion approximation method is as follows:

[0108] (1) According to the ideal metric and the definite metric in the soft decision information , taking the transmitted symbol sequence \(d\) k = 0 as an example, obtain the ideal channel capacity of the encoded LFM system:

[0109]

[0110] In the formula, C opt is the channel capacity derived from the ideal metric and the determined metric. For an ergodic channel, the time average can be replaced by the statistical average, thus obtaining the final result of the above formula.

[0111] (2) Perform a first-order Taylor expansion of the Bessel function at x = 0, that is:

[0112]

[0113] According to the result of the first-order Taylor expansion, estimate the ideal metric and the determined metric to obtain:

[0114]

[0115] Among them, is the modulus value of the frequency-domain signal. The conditions for the approximate equality in the above two formulas are that the signal-to-noise ratio is moderate, neither extremely low (such as 0 - 5 dB, noise-dominated) nor extremely high (such as above 20 dB, signal-dominated). Obviously, there is Then and

[0116] (3) According to the estimation results of the ideal metric and the determined metric , obtain the approximate channel capacity C app :

[0117]

[0118] Among them, C app is the approximate channel capacity, which is used to measure the approximate maximum information transmission rate that can be achieved by this communication system under the current conditions, eliminating the influence of the noise variance opt in the channel capacity C . is the conditional expectation operator, indicating that the expectation of the received signal r k is taken under the condition of q = 0.

[0119] S6: According to the ideal channel capacity C opt and the approximate channel capacity C app , respectively obtain the corresponding achievable code rates R opt and R app , and establish the relationship between the bit signal-to-noise ratio and the channel capacity C opt or C app .

[0120] Consider a coded LFM system with a spreading factor of SF, and the corresponding achievable code rates R opt and R app are obtained respectively according to the ideal channel capacity C opt and the approximate channel capacity C app by the following calculation method:

[0121]

[0122] The method for establishing the relationship between the bit signal-to-noise ratio and the channel capacity C opt or C app is as follows:

[0123] Given the spreading factor SF and the symbol signal-to-noise ratio E s / N0, calculate the relationship between the bit signal-to-noise ratio E b / N0 and the channel capacity C opt or C app :

[0124] or

[0125] where E s is the symbol energy, E b is the bit energy, and N0 represents the one-sided power spectral density.

[0126] For the complex analysis of the channel capacity C opt under the ideal metric and the channel capacity C app under the approximate metric, since the logarithmic operation can be quickly implemented by the binary search algorithm, the complexity brought by the logarithmic operation is not considered. Taking the real multiplication operation as an example, considering the m-th order Taylor expansion of the Bessel function I0(x) at x = 0 (theoretically m takes infinity, and in actual simulation, m can take a very large value), then the computational complexity of the ideal metric is O(m 2 NM); while the computational complexity of the approximate metric is O(NM). Obviously, the complexity of the ideal metric is much higher than that of the approximate metric.

[0127] S7: The method for soft input soft output decoding according to the soft decision information to recover the original data bit stream is as follows:

[0128] The soft decision information Λ(c k,i ) generated by soft decision non-coherent demodulation represents the probability information that each bit is "0" or "1". Input the soft decision information Λ(c k,i ) into the soft decision decoder, where the Turbo code decoding adopts BCJR

[0129] The (Bahl-Cocke-Jelinek-Raviv) algorithm performs iterative information exchange among component decoders, gradually approaching maximum a posteriori probability (MAP) decoding through extrinsic information update, and finally converting the decoded codeword into the original data bitstream through symbol demapping.

[0130] To further illustrate the beneficial effects of the present invention, comparative descriptions are provided through simulation experiments in this embodiment, which are specifically as follows:

[0131] Simulation 1:

[0132] Simulation conditions: In a coded non-coherent LFM system, assume that the transmission bandwidth B = 125 kHz, and the spreading factors SF are 7, 8, 9, 10, 11, 12 (corresponding to the number of orthogonal chirps M = 2 SF ).

[0133] Results and analysis: Figure 2 and Figure 3 respectively give the achievable code rate curves of the coded LFM systems based on ideal metric and approximate metric under six spreading factors in the present invention.

[0134] From Figure 2 and Figure 3 of the simulation results, it can be found that whether it is the ideal metric or the approximate metric, under different spreading factors SF, the curves of the achievable code rate R and the signal-to-noise ratio E b / N0 are in a "C" shape. Specifically, for the ideal metric and the approximate metric, their optimal achievable code rates are 1 / 2 and 1 / 5 respectively. That is, the signal-to-noise ratios E b / N0 corresponding to these two optimal achievable code rates are the lowest.

[0135] Simulation 2:

[0136] Simulation conditions: Assume that the spreading factors SF are 7 and 12, and the coding schemes adopt (7,4) Hamming code and (768,384), (1152,384), and (1920,384) Turbo codes. For the Hamming-LFM transmission system, the demodulation method adopted is hard decision non-coherent demodulation; for the Turbo-LFM transmission system, the demodulation method adopted is soft decision non-coherent demodulation.

[0137] Results and analysis: Figure 4 and Figure 5 respectively give the performances of the Hamming-LFM and Turbo-LFM transmission systems based on ideal metric and approximate metric under different spreading factors and code rates in the present invention.

[0138] From Figure 4 and Figure 5From the simulation results, it can be seen that the bit error performance of the Turbo-LFM system under different spreading factors and optimal code rates is improved by 8-10 dB compared with that of the Hamming-LFM system, that is, the Turbo code can adapt to a lower signal-to-noise ratio. Therefore, the Turbo-LFM scheme is more suitable for long-distance communications such as satellite Internet of Things. In addition, the low-code-rate Turbo-LFM based on the ideal metric has a performance improvement of about 1 dB compared with the low-code-rate Turbo-LFM based on the approximate metric. However, compared with the ideal metric, the approximate metric simplifies the Bessel function through Taylor expansion, thus reducing the computational complexity.

[0139] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An efficient coded linear frequency modulation transmission scheme for satellite Internet of Things communication, characterized in that, It includes the following steps: S1: Sequentially perform encoding and symbol mapping on the data bit stream generated by the user information to obtain a symbol sequence {d k}; S2: Perform improved LFM modulation with a spreading factor of SF on the symbol sequence {d k}, to obtain a chirp signal s k ; S3: For the chirp signal s k obtain the received signal r through a quasi-additive white Gaussian noise channel k ; S4: Perform de-chirping operation and soft-decision non-coherent demodulation on the received signal r k in sequence to obtain the soft-decision information Λ(c k,j ); S5: Obtain the ideal channel capacity C of the coded LFM system according to the soft decision information opt and obtain the approximate channel capacity C of the coded LFM system by using the first-order approximation method of Taylor expansion app ; S6: Obtain the corresponding achievable code rates R opt and R app respectively according to the ideal channel capacity C opt and the approximate channel capacity C app , and establish the relationship between the bit signal-to-noise ratio and the channel capacity C opt or C app ; S7: Perform soft-input soft-output decoding based on the soft decision information to recover the original data bit stream.

2. The efficient coding linear frequency modulation transmission scheme for satellite Internet of Things communication according to claim 1, wherein The coding method described in step S1 is: The described encoding method uses Turbo coding, and the data bit stream generated by user information is represented as {b i}, and the binary coding sequence obtained by Turbo coding is {c j}; Turbo coding is used to improve the error correction ability at extremely low signal-to-noise ratios; The method of symbol mapping is: Map each SF bits of the binary coded sequence {c j} after Turbo coding into a symbol sequence to obtain a decimal symbol sequence c k,j represents the j-th bit in the k-th decimal symbol sequence, k represents the symbol period index, and SF represents the spreading factor.

3. The efficient coding linear frequency modulation transmission scheme for satellite Internet of Things communication according to claim 2, wherein The method of performing improved LFM modulation with a spreading factor of SF on the symbol sequence described in step S2 is: Use the modulation signal acquisition formula to perform chirp modulation with a spreading factor of SF on the symbol sequence {d k} to obtain the chirp modulation signal sequence s k} of {d k ={s k (l)}, where s k represents the modulation signal transmitted within the k-th symbol period of the symbol sequence {d k}, and the modulation signal acquisition formula is: where s k (l) represents the modulation signal of the l-th in the k-th symbol period, k = 0, 1,..., N - 1 represents the symbol period index, N is the transmission length of the symbol sequence {d k}, l = 0, 1,..., 2 SF - 1 represents the chirp index, M = 2 SF is the number of chirps of the symbol sequence {d k}, θ k is the phase offset, is the imaginary unit.

4. The efficient coding linear frequency modulation transmission scheme for satellite Internet of Things communication according to claim 3, wherein The method for obtaining the received signal r described in step S3 k is as follows: r k = s k + n k ; where r k = {r k (0),..., r k (l),..., r k (M - 1)} represents the received signal sequence within the k-th symbol period, r k (l) represents the l-th received signal within the k-th symbol period, n k = {n k (0),..., n k (l),..., n k (M - 1)} represents the additive white Gaussian noise signal sequence within the k-th symbol period, n k (l) represents the l-th additive white Gaussian noise signal within the k-th symbol period; The method of the de-chirping operation described in step S4 is: For the received signal r k The l-th received signal r within the k-th symbol period of k (l) = s k (l) + n k (l) is multiplied by a down-chirp signal to obtain the de-chirped signal z k (l) is:

5. The efficient coding linear frequency modulation transmission scheme for satellite Internet of Things communication according to claim 4, wherein The method of soft decision non-coherent demodulation is: (1), Perform an M-point discrete Fourier transform on the de-chirped signal z k (l) to obtain the frequency-domain signal Z k (q); (2) Calculate the log-likelihood probability of the received signal according to the frequency-domain signal Z k (q). (3) Calculate the demodulation log-likelihood ratio of the \(i\)-th bit \(c\) of the transmitted symbol sequence \(\{d\}\) k as soft decision information \(\Lambda(c\) k,i ) according to the log-likelihood probability of the received signal. k,i ​ 6. The efficient coding linear frequency modulation transmission scheme for satellite Internet of Things communication according to claim 5, wherein The frequency-domain signal Z k (q) is as follows: where q is the frequency index, used to identify different frequency points in the frequency domain; The log-likelihood probability is: where, I0(·) is the first-kind modified Bessel function of order zero, which is used to obtain the determination metric when q = d k at this time is the magnitude of the frequency-domain signal, is the noise variance; The soft decision information is: Among them, Λ(c k,i ) is the soft decision value encoding the LFM signal, and i = 0, 1, ..., SF - 1 represents the binary bit stream {c k} of the symbol sequence {d k,0 , c k,1 , …, c k,SF-1} that makes up the transmitted data. The index of bit c k,i , and respectively represent the index sets where the i-th position is 0 and the index sets where the i-th position is 1 in the binary bit representations corresponding to the decimal numbers 0 to M1 according to the natural mapping.

7. The efficient coding chirp transmission scheme for satellite Internet of Things communication according to claim 5 or 6, characterized in that, The method for obtaining the approximate channel capacity C of the coded LFM system by using the first-order approximation method of Taylor expansion described in step S5 app is as follows: (1) Based on the determined metrics in the ideal metric and soft decision information, obtain the ideal channel capacity C of the coded LFM system opt ; (2) First kind of zero-order modified Bessel function Perform a first-order Taylor expansion at x = 0 and estimate the determined metric and the ideal metric based on the first-order Taylor expansion; (3) According to the estimation result, obtain the approximate channel capacity C app .

8. The efficient coding linear frequency modulation transmission scheme for satellite Internet of Things communication according to claim 7, wherein The ideal channel capacity C opt is as follows: The Bessel function described above The first-order Taylor expansion at x = 0 is as follows: The method of estimating the determination metric and the ideal metric is: wherein is the ideal metric; The approximate channel capacity C app is calculated as follows: wherein, is a conditional expectation operator, representing taking the expectation of the received signal r k under the condition that q = d k by taking the expectation.

9. The efficient coding linear frequency modulation transmission scheme for satellite Internet of Things communication according to claim 8, wherein According to the ideal channel capacity C described in step S6 opt and the approximate channel capacity C app respectively obtain the corresponding achievable code rates R opt and R app , and the method for establishing the relationship between the bit signal-to-noise ratio and the channel capacity C opt or C app is as follows: The described method for obtaining the corresponding achievable code rates R opt and approximate channel capacity C app respectively is as follows: opt and R app is: or The method for establishing the relationship between the bit signal-to-noise ratio and the channel capacity C opt or C app is as follows: Given the spreading factor SF and the symbol signal-to-noise ratio E s / N0, the bit signal-to-noise ratio E b / N0 and the relationship between the channel capacity: or where E s is the symbol energy, and E b is the bit energy, and N0 represents the one-sided power spectral density.

10. The efficient coding linear frequency modulation transmission scheme for satellite Internet of Things communication according to any one of claims 5, 6, 8 or 9, characterized in that The method of soft - input soft - output decoding according to the soft - decision information in step S7 to recover the original data bit - stream is as follows: input the soft - decision information Λ(c k,i ) into a soft - decision decoder, use the BCJR algorithm to decode the Turbo code, and convert the decoded codeword into the original data bit - stream through symbol demapping.

Images