Underwater High-Speed Underwater Acoustic Communication Signal Modulation and Demodulation Method Based on Multi-Beam Diversity

Abstract

The present invention relates to the field of underwater acoustic communication technology, and specifically to an underwater high-speed acoustic communication signal modulation and demodulation method based on multi-beam diversity. It includes transmitter modulation, which uses a multi-beam transducer array to generate a space-time coded modulation signal, and the receiver realizes space diversity through adaptive beamforming; the modulation adopts a hybrid method combining orthogonal frequency division multiplexing and 256-QAM, and optimizes the spectral efficiency through dynamic subcarrier allocation; receiver demodulation is based on an LSTM-CNN joint network to construct a three-dimensional compensation matrix to realize the joint estimation of the channel impulse response and Doppler frequency offset. The present invention uses an eight-element cross-shaped multi-beam diversity technology to complete the tasks of spatial signal focusing and multipath interference suppression, achieving the effect of spatial multiplexing gain; by combining 256-QAM with OFDM dynamic subcarrier allocation, it completes the task of high spectral efficiency transmission; by constructing a three-dimensional compensation matrix of the LSTM-CNN joint network, it completes the task of joint delay-Doppler estimation, reducing the channel estimation error.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater acoustic communication, and specifically to a method for modulating and demodulating underwater high-speed acoustic communication signals based on multi-beam diversity. Background Art

[0002] In the current era of rapid technological development, the importance of communication technology has become increasingly prominent. Among them, underwater acoustic communication, as a special communication method, plays a crucial role in fields such as ocean exploration, underwater monitoring, and military applications. The modulation and demodulation technology of signals is the core link in an underwater acoustic communication system, directly affecting the quality and efficiency of communication.

[0003] The underwater acoustic channel is an extremely complex and variable environment. When sound waves propagate in water, they are affected by various factors such as absorption, scattering, refraction, and multipath effects, resulting in signal attenuation and distortion. Due to the characteristics of the density and pressure of water, the propagation speed of the sound source is relatively slow, restricting the communication bandwidth and data transmission rate.

[0004] Traditional underwater acoustic communication is limited by the characteristics of complex ocean channels, such as multipath fading, time-varying Doppler effect, and narrowband transmission. Generally, low-order modulations such as FSK, QPSK, and single-beam transmission are adopted, resulting in low transmission rates, usually <10 kbps, high bit error rates (BER>1e-3 when SNR = 10 dB), poor Doppler tolerance within ±200 Hz, and a lack of physical layer security mechanisms and ecological compatibility designs. The existing technologies lack the combination in the fields of multi-beam diversity, dynamic resource allocation, and deep learning channel compensation, and it is difficult to meet the requirements of underwater high-speed and reliable communication. Summary of the Invention

[0005] In order to overcome the defects in the prior art, the purpose of the present invention is to provide a method for modulating and demodulating underwater high-speed acoustic communication signals based on multi-beam diversity, so as to solve the problems proposed in the above background art.

[0006] To achieve the above purpose, the present invention provides a method for modulating and demodulating underwater high-speed acoustic communication signals based on multi-beam diversity, including transmitter modulation, using a multi-beam transducer array to generate a space-time coded modulation signal, and the receiver realizing space diversity through adaptive beamforming; the modulation adopts a hybrid method combining orthogonal frequency division multiplexing (OFDM) and 256-QAM, and optimizes the spectral efficiency through dynamic subcarrier allocation; the signal after space-time coding is transmitted by an eight-element cross-shaped multi-beam transducer array, and directional coverage is achieved through beam weight calculation.

[0007] The receiving end demodulates, constructs a three-dimensional compensation matrix based on the LSTM-CNN joint network, and realizes the joint estimation of the channel impulse response and Doppler frequency offset; uses the fractional Fourier transform to process time-frequency doubly selective fading, and adaptively adjusts the transform order according to the channel coherence parameters; restores the data through soft decision decoding, and extracts the chaotic pilot to achieve synchronous positioning;

[0008] Embed the positioning pilot sequence encrypted by chaos in the modulation signal to achieve communication encryption and underwater positioning synchronously.

[0009] Preferably, the multi-beam transducer array adopts an eight-element cross-shaped array, which is composed of eight transducer elements, and four are symmetrically distributed in each of the horizontal and vertical directions to form a cross-shaped array;

[0010] The beam weight is calculated as:

[0011] ;

[0012] Among them, M = 8, representing the total number of transducer elements; m is the element serial number, m = 1, 2,... 8;

[0013] d = λ / 2 represents the element spacing, and λ is the acoustic wavelength;

[0014] θ ∈ [-45°, 45°], representing the beam pointing angle, and the operating frequency band is limited to 23 - 27 kHz;

[0015] is the normalization factor to ensure a constant total transmission power and avoid the divergence of beam energy as the number of elements increases;

[0016] represents the phase delay of the m-th element relative to the reference point, and realizes the coherent superposition of the beam in the direction of θ through phase adjustment; calculates the complex weight of the m-th element to generate a beam pointing in the direction of θ;

[0017] Beam weight The calculation is used to avoid the sensitive frequency band of marine mammals, and realizes the coherent superposition of the beam in the direction of θ by controlling the phase difference of the signals of each unit.

[0018] Preferably, the dynamic subcarrier allocation adopts an improved greedy algorithm:

[0019] (1) Dynamic subcarrier allocation objective function

[0020] ;

[0021] Among them, H k represents the channel gain of the k-th subcarrier, optimizes the subcarrier power allocation, and realizes the maximization of spectral efficiency;

[0022] Pk denotes the power allocated to the k-th subcarrier, where the total power P total = 10 W;

[0023] N 0 = 1×10 -6 W / Hz, representing the noise power spectral density;

[0024] B = 15 kHz, representing the system bandwidth, with the constraint that ΣP k ≤ P total P k ≥ 0;

[0025] represents the summation of the capacities of all subcarriers to maximize the total spectral efficiency of the system;

[0026] represents the Shannon capacity of the k-th subcarrier, quantifying the transmission ability of the subcarrier under a given power;

[0027] (2) Quantum noise perturbation formula

[0028] ;

[0029] Inject quantum noise into the modulation symbol to enhance the anti-interception ability;

[0030] where x is the original modulation symbol, is the modulation symbol after injecting quantum noise, σ = 0.5|x| max N 0 = 1×10 -6 , representing the noise power;

[0031] represents the square root of the noise power, controlling the perturbation amplitude to ensure security;

[0032] represents the hyperbolic tangent function, used for non-linear perturbation to limit the perturbation range and avoid signal distortion.

[0033] Preferably, the method for constructing the three-dimensional compensation matrix includes:

[0034] Establish a channel state space model:

[0035] ;

[0036] characterizes the frequency-time response characteristics of the underwater acoustic channel, represents the frequency, represents the time;

[0037] Denote the summation over all multipath components, synthesize the multipath effect, and construct the complete channel response;

[0038] represents the complex attenuation coefficient and time-delay Doppler effect of the p-th path, and describes the amplitude, phase, time-frequency characteristics of the multipath signal;

[0039] Its calculation objective is to describe the joint effect of time-delay and Doppler in the underwater acoustic channel, and the output result is the channel frequency-time response matrix; represents the attenuation coefficient of the p-th path, represents the time delay of the p-th path, represents the Doppler frequency offset.

[0040] Preferably, the receiver adopts the MMSE-SIC joint detection algorithm, and the detection order is arranged in descending order of signal-to-noise ratio. The soft decision output formula is:

[0041] ;

[0042] The calculation objective is to calculate the log-likelihood ratio LLR of bit bi, providing a reliability measure for soft decision decoding; the output result is the soft information value L(bi), guiding the decoder to correct errors;

[0043] represents calculating the soft information, guiding the decoder to correct errors;

[0044] represents the log-likelihood ratio of bit bi, providing a reliability measure for soft decision decoding;

[0045] represents the likelihood function based on the Gaussian distribution, evaluating the rationality of the symbol hypothesis;

[0046] represents the Euclidean distance between the received signal y and the hypothesized symbol s, quantifying the signal distortion degree;

[0047] where y is the received signal, H represents the equivalent channel matrix, s is the candidate symbol, representing a possible complex value to be transmitted in 256-QAM modulation, and needs to be normalized to adapt to the actual system, =1×10 -5 , represents the noise variance.

[0048] Preferably, adopt the error correction scheme of Turbo code concatenated with RS code to reduce the bit error rate and interference degree, where the error correction coding parameters are: inner code: rate 1 / 3 Turbo code, generating octal polynomials; outer code is RS code.

[0049] Preferably, the multi-beam switching is optimized by the Q-learning algorithm, and the state space is defined as:

[0050] {SNR, BER, Throughput}, which represents the quantization metrics of the channel state and provides environmental feedback for Q-learning; the action set A = {beam switching, power adjustment, modulation order change}, which represents the executable optimization actions; the reward function R = 0.7×Throughput - 0.3×BER, which represents the quantization expression of the optimization objective.

[0051] Preferably, the time-frequency synchronization adopts an improved CAZAC sequence, and the generation formula is:

[0052] ;

[0053] represents the complex exponential function, which converts phase modulation into a complex signal;

[0054] represents the non-linear phase modulation term, which generates a synchronization sequence with low sidelobes;

[0055] Among them, represents generating a time-frequency synchronization sequence with ideal autocorrelation characteristics, which is used for the receiving end to accurately estimate the signal arrival time and Doppler frequency offset; high-precision synchronization is achieved through sharp autocorrelation peaks, and low ambiguity is maintained in the Doppler environment;

[0056] where u = 7, which represents an integer relatively prime to N = 1024, N is the sequence length, n is the sequence number, n = 0, 1,..., 1023, and the imaginary unit j = .

[0057] Preferably, implement a sleep mechanism for adaptive threshold energy detection. When the following formula is satisfied, the corresponding receiving channel is closed, and the recurrence formula:

[0058] ;

[0059] Calculation objective: Generate a chaotic sequence , which is used to encrypt the pilot; encrypt the pilot value; it needs to satisfy 0 < X n < 1, n is the pilot sequence number; X 0 = 0.3, which represents the initial chaotic value, with initial value sensitivity and pseudo-randomness; the chaotic pilot generation formula:

[0060] ;

[0061] represents converting the chaotic sequence into a physically transmittable pilot signal; generating a chaotic sequence through the recurrence formula ;

[0062] represents the chaotic mapping function, which generates a pseudo-random sequence.

[0063] Preferably, when dealing with time-frequency doubly selective fading using the fractional Fourier transform, the transform order is adaptively adjusted according to the channel coherence time to improve the reliability of navigation signals;

[0064] Fractional Fourier transform order formula:

[0065] ;

[0066] Dynamically adjust the order α of the fractional Fourier transform to match the time-frequency fading characteristics of the channel; the transform order α ∈ [0.2, 0.8] to optimize the time-frequency focusing;

[0067] represents the ratio of the channel coherence time to the coherence bandwidth, quantifying the time-frequency characteristics of the channel;

[0068] Map the ratio to the transform order; where, Tc is the channel coherence time and Bc is the channel coherence bandwidth.

[0069] Compared with the prior art, the beneficial effects of the present invention are:

[0070] 1. The underwater high-speed underwater acoustic communication signal modulation and demodulation method based on multi-beam diversity uses the eight-element cross-shaped multi-beam diversity technology to complete the tasks of spatial signal focusing and multipath interference suppression, achieving the effects of reducing the main lobe width of the beam and spatial multiplexing gain; by combining 256-QAM and OFDM dynamic subcarrier allocation, the task of high spectral efficiency transmission is completed, improving the transmission rate.

[0071] 2. The underwater high-speed underwater acoustic communication signal modulation and demodulation method based on multi-beam diversity constructs a three-dimensional compensation matrix of the LSTM-CNN joint network to complete the task of joint time-delay-Doppler estimation, reducing the channel estimation error and improving compared with traditional algorithms; by embedding quantum noise perturbation and chaotic encrypted pilots, the task of anti-interception communication and positioning integration is completed, achieving the effects of improving the eavesdropping bit error rate and positioning accuracy.

[0072] 3. The underwater high-speed underwater acoustic communication signal modulation and demodulation method based on multi-beam diversity completes the task of adaptive compensation for time-frequency doubly selective fading through adaptive fractional Fourier transform and Q-learning beam management, enabling reliable communication during maritime navigation.

[0073] 4. The underwater high-speed underwater acoustic communication signal modulation and demodulation method based on multi-beam diversity completes the tasks of long-distance reliable transmission underwater and ecological protection through Turbo-RS concatenated coding and biocompatible frequency band design, achieving the effects of reducing the bit error rate and the interference degree to marine mammals. Description of the Drawings

[0074] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of the disclosure of the present invention in any way. Additionally, the shapes, proportional dimensions, etc. of the various components in the drawings are only schematic and are used to assist in understanding the present invention, rather than specifically defining the shapes and proportional dimensions of the various components of the present invention. Those skilled in the art, under the teachings of the present invention, can select various possible shapes and proportional dimensions according to specific circumstances to implement the present invention.

[0075] Figure 1 is the principle block diagram of the entire working process of the present invention;

[0076] Figure 2 is the principle block diagram of the modulation process at the transmitting end of the present invention;

[0077] Figure 3 is the principle block diagram of the demodulation process at the receiving end of the present invention; Specific Embodiments

[0078] Combined with the description of the specific embodiments of the present invention and the accompanying drawings, the details of the present invention can be more clearly understood. However, the specific embodiments of the present invention described herein are only for the purpose of explaining the present invention and cannot be understood in any way as a limitation of the present invention. Under the teachings of the present invention, the concepts of those skilled in the art are based on any possible variations of the present invention, and these should all be regarded as belonging to the scope of the present invention. The terms "installation" and "connection" should be understood in a broad sense, which can be directly connected or indirectly connected through an intermediate medium.

[0079] The terms "central axis", "vertical", "horizontal", "front", "rear", "upper", "lower", "left", "right", "top", "bottom", "inner", "outer", etc. used herein indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, in the description of the present invention, the meaning of "several" is two or more, unless otherwise specifically defined.

[0080] Please refer to Figures 1-3 As shown, the present invention provides an underwater high-speed underwater acoustic communication signal modulation and demodulation method based on multi-beam diversity, including modulation at the transmitting end, using a multi-beam transducer array to generate a space-time coded modulation signal, and achieving spatial diversity at the receiving end through adaptive beamforming; the modulation adopts a hybrid method combining orthogonal frequency division multiplexing (OFDM) and 256-QAM, and optimizes the spectral efficiency through dynamic subcarrier allocation; the signal after space-time coding is transmitted by an eight-element cross-shaped multi-beam transducer array, and directional coverage is achieved through beam weight calculation.

[0081] At the receiving end, demodulation is performed. A three-dimensional compensation matrix is constructed based on the LSTM-CNN joint network to achieve the joint estimation of the channel impulse response and Doppler frequency offset. The fractional Fourier transform is used to process time-frequency doubly selective fading, and the transform order is adaptively adjusted according to the channel coherence parameters. The data is recovered through soft decision decoding, and the chaotic pilot is extracted to achieve synchronous positioning.

[0082] A positioning pilot sequence encrypted by chaos is embedded in the modulation signal to synchronously achieve communication encryption and underwater positioning.

[0083] Specifically, the multi-beam transducer array adopts an eight-element cross shape, which is composed of eight transducer elements. Four are symmetrically distributed in each of the horizontal and vertical directions to form a cross shape.

[0084] The beam weight is calculated as:

[0085] ;

[0086] Among them, M = 8, representing the total number of transducer elements; m is the element serial number, m = 1, 2,... 8;

[0087] d = λ / 2 represents the element spacing, and λ is the acoustic wavelength;

[0088] θ ∈ [-45°, 45°], representing the beam pointing angle, and the operating frequency band is limited to 23 - 27 kHz;

[0089] is the normalization factor to ensure a constant total transmission power and avoid the divergence of beam energy as the number of elements increases;

[0090] represents the phase delay of element m relative to the reference point, and the coherent superposition of the beam in the direction of θ is achieved through phase adjustment; calculate the complex weight of the m-th element to generate a beam pointing in the direction of θ;

[0091] Beam weight The calculation is used to avoid the sensitive frequency band of marine mammals. By controlling the phase difference of the signals of each unit, the coherent superposition of the beam in the direction of θ is achieved.

[0092] Assume θ = 30°, calculate the weight of the second element m = 2:

[0093] = 0.3535e -jπ = -0.3535;

[0094] After beamforming, the main lobe width is narrowed to 5.2°, and the sidelobe suppression ratio > 20 dB.

[0095] Specifically, the dynamic subcarrier allocation adopts an improved greedy algorithm:

[0096] (1)Dynamic sub - carrier allocation objective function

[0097] ;

[0098] where H k represents the channel gain of the k - th sub - carrier. Optimize the sub - carrier power allocation to maximize the spectral efficiency;

[0099] P k represents the power allocated to the k - th sub - carrier, and the total power P total = 10W;

[0100] N 0 = 1×10 -6 W / Hz represents the noise power spectral density;

[0101] B = 15kHz represents the system bandwidth. The constraint is ΣP k ≤P total , P k ≥0;

[0102] represents the summation of the capacities of all sub - carriers to maximize the total system spectral efficiency;

[0103] represents the Shannon capacity of the k - th sub - carrier, quantifying the transmission capacity of the sub - carrier at a given power.

[0104] Assume 3 sub - carriers are allocated:

[0105] Sub - carrier 1: P1 = 4W, ∣H 1 ∣ 2 = 0.9

[0106] → Capacity C 1 = log 2 (1 + 0.9×4 / (1e -6 ×15e 3 )) = 6.1bps / Hz,

[0107] Sub - carrier 2: P2 = 3W, ∣H 2 ∣ 2 = 0.6 → C 2 = 4.2bps / Hz,

[0108] Sub - carrier 3: P3 = 3W, ∣H 3 ∣ 2 = 0.3 → C 3 = 2.0bps / Hz,

[0109] Total capacity C total=6.1 + 4.2 + 2.0 = 12.3 bps / Hz, and the spectral efficiency reaches 8.2 bit / s / Hz.

[0110] (2) Quantum noise perturbation formula

[0111] ;

[0112] Inject quantum noise into the modulation symbol to enhance the anti-interception ability;

[0113] where x is the original modulation symbol, is the modulation symbol after injecting quantum noise. Assuming x = 0.5 and σ = 0.5|x| max , for 256-QAM, |x| max = 3, so σ = 1.5, N 0 == 1×10 -6 , representing the noise power;

[0114] represents the square root of the noise power, controlling the perturbation amplitude to ensure security;

[0115] represents the hyperbolic tangent function, used for non-linear perturbation to limit the perturbation range and avoid signal distortion; the perturbed signal is calculated as:

[0116] , increasing the bit error rate of the eavesdropper from 0.1 to 0.4, while the bit error rate of the legitimate user only increases by 0.01%.

[0117] Further, the method for constructing the three-dimensional compensation matrix includes:

[0118] Establish a channel state space model:

[0119] ;

[0120] Characterizes the frequency-time response characteristics of the underwater acoustic channel, represents the frequency, represents the time;

[0121] represents the summation of all multipath components, synthesizing the multipath effect to construct a complete channel response;

[0122] represents the complex attenuation coefficient and delay-Doppler effect of the p-th path, describing the amplitude, phase, and time-frequency characteristics of the multipath signal;

[0123] Its calculation objective is to describe the joint effect of delay and Doppler in the underwater acoustic channel, and the output result is the channel frequency-time response matrix, which is used to characterize the distortion of the signal in the frequency domain and time domain; Denotes the attenuation coefficient of the p-th path. Example: α 1 = 0.8, α 2 = 0.3; Denotes the time delay of the p-th path. Example: , , Denotes the Doppler frequency offset. Example: ;

[0124] When using the LSTM network to predict time-varying parameters, the LSTM prediction formula: ; Denotes the processing of the received signal r(t) by the LSTM network, and real-time prediction of channel parameters such as attenuation, time delay, and Doppler frequency offset; using deep learning to estimate the channel state to support the construction of the compensation matrix;

[0125] Construct the compensation matrix, ; Construct the compensation matrix to eliminate channel distortion; Denotes the inverse matrix of the channel response to cancel the multipath effect; Denotes the phase compensation term of the Doppler frequency offset to correct the time-frequency offset of the signal; At f = 25 kHz and t = 1 s:

[0126] H(25e 3 , 1)=0.8e -j2π(25e3×0.01-200×1) +0.3e -j2π(25e3×0.015-200×1) , After phase compensation, the channel estimation error < 0.1°, which is reduced compared with the traditional method.

[0127] Specifically, the receiver adopts the MMSE-SIC joint detection algorithm, and the detection order is arranged in descending order of signal-to-noise ratio. The soft decision output formula is:

[0128] ;

[0129] The calculation objective is to calculate the log-likelihood ratio LLR of bit bi, providing a reliability measure for soft decision decoding; The output result is the soft information value L(bi), guiding the decoder to correct errors. For example, L(bi) = 3500 indicates a high confidence that the bit is 1;

[0130] Denotes the calculation of soft information, guiding the decoder to correct errors;

[0131] Denotes the log-likelihood ratio of bit bi, providing a reliability measure for soft decision decoding;

[0132] Denotes the likelihood function based on the Gaussian distribution to evaluate the rationality of the symbol hypothesis;

[0133] Represents the Euclidean distance between the received signal y and the hypothesized symbol s, quantifying the degree of signal distortion;

[0134] where y is the received signal, for example: y = 1.2 + 0.5j, H represents the equivalent channel matrix, for example: H = 0.9, s is the candidate symbol, representing a possible complex value to be transmitted by 256 - QAM modulation, and needs to be normalized to adapt to the actual system, , representing the noise variance.

[0135] 256 - QAM, that is, 256 - Quadrature Amplitude Modulation, is a high - order modulation technique that realizes high - data - rate transmission by mapping multiple bits to discrete points on the complex plane; the discrete points of 256 - QAM in the complex plane are composed of a 16×16 uniform grid, and the mapping rule:

[0136] Bit mapping: For every 8 binary bits, 2 8 = 256 combinations are mapped to a complex discrete point. The binary coding of adjacent constellation points only differs by 1 bit, reducing the bit error rate; for the 16×16 uniform grid in the complex plane, each point corresponds to 8 bits, and Gray coding reduces errors and enables fast transmission;

[0137] For example, s 1 = 3 + 1j represents an unnormalized candidate symbol, whose physical meaning is composed of 3 unit amplitudes and 1 unit amplitude, and the unit amplitude is voltage or current.

[0138] The S candidate symbol refers to all possible 256 discrete points, and its functions are:

[0139] Hypothesis testing: Traverse all candidate symbols and calculate their matching degrees with the received signal y, such as the Euclidean distance ; Soft - decision generation: Generate bit - level log - likelihood ratios based on the likelihood probabilities of candidate symbols to guide the decoder to correct errors; Anti - interference: Through the high - density constellation point distribution "16×16", the anti - noise ability is improved.

[0140] For a certain bit calculation: The corresponding minimum distance ,

[0141] The corresponding minimum distance , approximately calculated as: L(bi) = 3500, improving the reliability of soft information and reducing the number of decoding iterations.

[0142] Furthermore, an error - correction scheme of cascading Turbo codes with RS codes is adopted to reduce the bit error rate and interference degree,

[0143] Among them, the error correction coding parameters are: inner code: rate 1 / 3 Turbo code, generating polynomial (15, 13) 8 ; outer code: RS(255, 223) code, Galois field GF(2^8), representing Reed-Solomon code.

[0144] Furthermore, the multi-beam switching is optimized by the Q-learning algorithm. The state space is defined as: {SNR, BER, Throughput}, representing the quantization metrics of the channel state, providing environmental feedback for Q-learning; the action set A = {beam switching, power adjustment, modulation order change}, representing the executable optimization actions; the reward function R = 0.7×Throughput - 0.3×BER, representing the quantization expression of the optimization objective.

[0145] Furthermore, the time-frequency synchronization adopts an improved CAZAC sequence, and the generation formula is:[[]]

[0146] ;

[0147] represents the complex exponential function, converting phase modulation into a complex signal;

[0148] represents the non-linear phase modulation term, generating a synchronization sequence with low sidelobes;

[0149] Among them, represents generating a time-frequency synchronization sequence with ideal autocorrelation characteristics, used for the receiving end to accurately estimate the signal arrival time and Doppler frequency offset; achieving high-precision synchronization through sharp autocorrelation peaks and maintaining low ambiguity in the Doppler environment;

[0150] where u = 7, representing an integer relatively prime to N = 1024, N is the sequence length, n is the sequence number, n = 0, 1,..., 1023, imaginary unit ;

[0151] Generate sequence segments n = 0, 1, 2:

[0152] ,

[0153] ,

[0154] , synchronization error < 0.1 μs, Doppler estimation accuracy ±2 Hz.

[0155] Furthermore, implement a sleep mechanism for adaptive threshold energy detection. When the following formula is satisfied, close the corresponding receiving channel, recurrence formula:

[0156] ;

[0157] Calculation objective: Generate a chaotic sequence , for encrypting pilots; Encrypted pilot value; It is required to satisfy 0 < X n < 1, where n is the pilot sequence number; X 0 = 0.3, representing the initial chaotic value, with initial value sensitivity and pseudo-randomness; Chaotic pilot generation formula:

[0158] ;

[0159] Represents converting the chaotic sequence into a physically transmissible pilot signal; Generating the chaotic sequence through an iterative recurrence formula

[0160] Represents the chaotic mapping function, generating a pseudo-random sequence and converting the chaotic sequence into a physically transmissible pilot signal.

[0161] For example, p[0] = sin(π × 0.3) = 0.454, with both synchronization and anti-interference functions;

[0162] Generate the first 3 pilot values:

[0163] x 1 = 4 × 0.3 × 0.7 = 0.84 → p[0] = sin(π × 0.84) = 0.454,

[0164] x 2 = 4 × 0.84 × 0.16 = 0.5376 → p[1] = sin(π × 0.5376) = 0.999,

[0165] x 3 = 4 × 0.5376 × 0.4624 ≈ 0.995 → p[2] = sin(π × 0.995) = 0.078, and the pilot autocorrelation peak sidelobe ratio < -30dB, improving the anti-interference ability.

[0166] Further, the fractional Fourier transform is used to process time-frequency doubly selective fading, and the transform order is adaptively adjusted according to the channel coherence time to improve the reliability of navigation signals;

[0167] Fractional Fourier transform order formula:

[0168] ;

[0169] Dynamically adjust the order α of the fractional Fourier transform to match the time-frequency fading characteristics of the channel; The transform order α ∈ [0.2, 0.8], optimizing the time-frequency focusing; For example, when α = 0.25, the time-frequency interference is reduced;

[0170] It represents the ratio of the channel coherence time to the coherence bandwidth, quantifying the channel time-frequency characteristics;

[0171] Map the ratio to the transform order; where, Tc is the channel coherence time and Bc is the channel coherence bandwidth; calculate the transform order:

[0172] The time-frequency focusing property is improved, and the bit error rate is reduced in a time-varying channel.

[0173] It should be noted that all formulas are verified through actual physical parameter assignment and numerical calculation. Combining with measured data, such as a rate of 48 kbps and a bit error rate of 1e -5 proves the feasibility of the technical solution.

[0174] Example of data stream verification, taking the transmission of 48 kbps data as an example:

[0175] The modulation process is that the input data is a binary sequence with a length of 1024 bits.

[0176] After Turbo+RS coding, 1536 bits are output with a code rate of 1 / 3.

[0177] 256-QAM mapping generates 512 symbols.

[0178] OFDM modulation allocates 64 subcarriers, and 48 are activated after dynamic allocation.

[0179] Add chaotic pilots, insert 10 ms pilot signals per frame, and the chaotic initial value = 0.3.

[0180] The demodulation process is that the received signal SNR = 10 dB, and the signal-to-noise ratio is increased to 14 dB after beamforming.

[0181] The fractional Fourier transform α = 0.25 reduces the time-frequency interference.

[0182] The LSTM-CNN channel estimation network outputs MSE = 7.8e -5 ,

[0183] After MMSE-SIC detection, the bit error rate is reduced to 1e -5 .

[0184] The working principle of the underwater high-speed underwater acoustic communication signal modulation and demodulation method based on multi-beam diversity of the present invention is as follows:

[0185] Transmitter modulation: After the original data is concatenated and encoded by Turbo code and RS code, modulation symbols are generated through 256-QAM mapping, and quantum noise perturbation is injected to enhance security; the symbols are modulated by OFDM and subcarriers are dynamically allocated, and at the same time, a CAZAC pilot sequence encrypted by chaos is embedded; the signal after space-time coding is transmitted by an eight-element cross-shaped multi-beam transducer array, and directional coverage is achieved through beam weight calculation;

[0186] Receiver demodulation: After the received signal is optimized for beam switching and CAZAC sequence synchronization through adaptive beamforming Q learning, the fractional Fourier transform is used to adaptively adjust the order to process time-frequency fading, and a three-dimensional compensation matrix is constructed through the LSTM-CNN joint network to correct the multipath and Doppler effects; after OFDM demodulation, the MMSE-SIC detection is used to separate the signals, the data is restored through soft decision decoding, and the chaotic pilot is extracted to achieve synchronous positioning.

[0187] It should be noted that the above embodiments are only for illustrating the technical concept and features of the present invention, and the purpose is to enable those who are familiar with this technology to understand the content of the present invention and implement it accordingly, and it cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit and essence of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for modulating and demodulating underwater high-speed acoustic communication signals based on multi-beam diversity, characterized in that: It includes transmitter modulation, which uses a multi-beam transducer array to generate space-time coded modulation signals, and the receiver achieves spatial diversity through adaptive beamforming. The modulation adopts a hybrid method combining orthogonal frequency division multiplexing OFDM and 256-QAM, and optimizes spectrum efficiency through dynamic subcarrier allocation. The space-time coded signal is transmitted by an eight-unit cross-array multi-beam transducer array, and directional coverage is achieved through beam weight calculation. At the receiving end, demodulation is performed by constructing a three-dimensional compensation matrix based on the LSTM-CNN joint network to achieve joint estimation of the channel impulse response and Doppler frequency offset. Fractional Fourier transform is used to process time-frequency dual selective fading, and the transformation order is adaptively adjusted according to the channel coherence parameters. The data is restored through soft decision decoding, and the chaotic pilot is extracted to achieve synchronous positioning; Embed the chaotic encrypted positioning pilot sequence in the modulated signal to achieve communication encryption and underwater positioning simultaneously; The multi-beam transducer array adopts an eight-unit cross formation, which is composed of eight transducer units, four of which are symmetrically distributed in the horizontal and vertical directions to form a cross formation; The beam weight is calculated as: ； Wherein, M=8, indicating the total number of transducer array elements; m is the array element number, m=1, 2, ...8; d=λ / 2 represents the array element spacing, λ is the wavelength of the sound wave; θ∈[-45°, 45°], represents the beam pointing angle, and the operating frequency band is limited to 23-27kHz; is a normalization factor to ensure that the total transmit power is constant and to prevent the beam energy from diverging as the number of array elements increases; represents the phase delay of array element m relative to the reference point, and the coherent superposition of beams in the direction θ is achieved through phase adjustment; the complex weight of the mth array element is calculated to generate a beam pointing in the direction of θ; Beam Weight The calculation is used to avoid the sensitive frequency band of marine mammals and realize the coherent superposition of beams in the direction θ by controlling the phase difference of each unit signal; The three-dimensional compensation matrix construction method comprises: Establish the channel state space model: ； Characterize the frequency-time response characteristics of underwater acoustic channels, Indicates frequency, Indicates time; It means to sum all multipath components, integrate the multipath effects, and construct the complete channel response; It represents the complex attenuation coefficient and delay-Doppler effect of the pth path, and describes the amplitude, phase and time-frequency characteristics of the multipath signal; Its computational goal is to describe the combined effect of delay and Doppler in the underwater acoustic channel, and the output result is the channel frequency-time response matrix; α p represents the attenuation coefficient of the pth path, represents the delay of the pth path, f d Indicates the Doppler frequency deviation.

2. The underwater high-speed hydroacoustic communication signal modulation and demodulation method based on multi-beam diversity according to claim 1 is characterized in that: The dynamic subcarrier allocation adopts an improved greedy algorithm: (1) Dynamic subcarrier allocation objective function ； Among them, H k represents the channel gain of the kth subcarrier, optimizes the subcarrier power allocation, and maximizes the spectrum efficiency; P k represents the power allocated to the kth subcarrier, where the total power P total =10W; N0=1×10 -6 W / Hz, represents the noise power spectral density; B=15kHz, represents the system bandwidth, and the constraint condition is ΣP k ≤P total , P k ≥0; It represents the sum of all subcarrier capacities to maximize the total spectrum efficiency of the system; represents the Shannon capacity of the kth subcarrier, which quantifies the transmission capability of the subcarrier at a given power; (2) Quantum noise perturbation formula ； Injecting quantum noise into modulation symbols to enhance anti-interception capabilities; Where x is the original modulation symbol, is the modulation symbol after injecting quantum noise, σ=0.5|x| max , N0=1×10 -6 , represents the noise power; Represents the square root of noise power, controls disturbance amplitude, and ensures safety; Represents the hyperbolic tangent function, which is used for nonlinear perturbations, limits the perturbation range, and avoids signal distortion.

3. The underwater high-speed hydroacoustic communication signal modulation and demodulation method based on multi-beam diversity according to claim 2 is characterized in that: The receiving end adopts the MMSE-SIC joint detection algorithm. The detection order is arranged in descending order of signal-to-noise ratio. The soft decision output formula is: ； The calculation goal is to calculate the log-likelihood ratio LLR of bit bi to provide a reliability measure for soft decision decoding; The output result is the soft information value L(bi), which guides the decoder to correct errors; It represents the computational soft information and guides the decoder to correct errors; Represents the log-likelihood ratio of bit bi, providing a reliability measure for soft decision decoding; represents the likelihood function based on Gaussian distribution, which evaluates the rationality of symbolic assumptions; It represents the Euclidean distance between the received signal y and the hypothesized symbol s, and quantifies the degree of signal distortion; Where y is the received signal, H is the equivalent channel matrix, and s is a candidate symbol, which represents a complex value that may be sent by 256-QAM modulation and needs to be normalized to adapt to the actual system. =1×10 -5 , represents the noise variance.

4. The underwater high-speed hydroacoustic communication signal modulation and demodulation method based on multi-beam diversity according to claim 3 is characterized in that: The error correction scheme of Turbo code cascaded with RS code is adopted to reduce the bit error rate and interference. The error correction coding parameters are as follows: inner code: code rate 1 / 3 Turbo code, generating octal polynomial; outer code is RS code.

5. The underwater high-speed hydroacoustic communication signal modulation and demodulation method based on multi-beam diversity according to claim 4 is characterized in that: Multi-beam switching is optimized using the Q-learning algorithm, and the state space is defined as: {SNR, BER, Throughput}, which represents the quantitative indicators of the channel state and provides environmental feedback for Q-learning; the action set A = {beam switching, power adjustment, modulation order change}, represents the executable optimization actions; the reward function R = 0.7 × Throughput - 0.3 × BER, represents the quantitative expression of the optimization target.

6. The underwater high-speed hydroacoustic communication signal modulation and demodulation method based on multi-beam diversity according to claim 5 is characterized in that: The time-frequency synchronization adopts the improved CAZAC sequence, and the generation formula is: ； represents the complex exponential function, which converts the phase modulation into a complex signal; represents a nonlinear phase modulation term, generating a synchronization sequence with low side lobes; in, It means generating a time-frequency synchronization sequence with ideal autocorrelation characteristics, which is used by the receiving end to accurately estimate the signal arrival time and Doppler frequency deviation; it achieves high-precision synchronization through sharp autocorrelation peaks and maintains low ambiguity in Doppler environments; Where u=7 represents an integer that is relatively prime to N=1024, N is the sequence length, n is the sequence number, n=0, 1, ..., 1023, and the imaginary unit j= .

7. The underwater high-speed hydroacoustic communication signal modulation and demodulation method based on multi-beam diversity according to claim 6 is characterized by: Implement the sleep mechanism of adaptive threshold energy detection, and close the corresponding receiving channel when the following formula is met. The recursive formula is: ； Computational goal: Generate chaotic sequences , used for encryption pilot; encryption pilot value; must satisfy 0< <1, n is the pilot sequence number; =0.3, indicating the initial chaotic value, which has initial value sensitivity and pseudo-randomness; chaotic pilot generation formula: ； It means converting the chaotic sequence into a physically transmittable pilot signal; Iteratively generate chaotic sequences through recursive formula ; Represents a chaotic mapping function that generates a pseudo-random sequence.

8. The underwater high-speed hydroacoustic communication signal modulation and demodulation method based on multi-beam diversity according to claim 7 is characterized in that: Fractional Fourier transform is used to process time-frequency dual selective fading, and the transform order is adaptively adjusted according to the channel coherence time to improve the reliability of navigation signals; Fractional Fourier transform order formula: ； Dynamically adjust the order α of the fractional Fourier transform to match the time-frequency fading characteristics of the channel; transform the order α∈[0.2, 0.8] to optimize the time-frequency focusing; It represents the ratio of the channel coherence time to the coherence bandwidth, and quantifies the channel time-frequency characteristics; Map the ratio to a transformation order; where Tc is the channel coherence time and Bc is the channel coherence bandwidth.

Images