Information transmission method and system for submersible

Abstract

The invention provides an information transmission method and system for a submersible, and the method comprises the following steps: a target receiving submersible broadcasts a detection signal, and enables a transmitting submersible to carry out the parametric modeling of a distorted signal. When data needs to be transmitted, the transmitting submersible generates a pre-coding signal based on the channel model, and the signal can realize space-time focusing at the position of the target receiver after being transmitted through the underwater environment. And if different data needs to be sent to a plurality of different targets at the same time, acoustic orbital angular momentum phase factors of different modes are loaded for each path of data so as to realize distinguishing. And the plurality of transmitting submersible vehicles cooperatively transmit respective pre-coded signals under the synchronization of the high-precision clock. And finally, the target receiving submersible sharpens the received focusing signal through adaptive blind equalization processing, eliminates residual interference and recovers original data. The underwater communication method has the effect of improving the underwater communication reliability of the submersible.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of submersible communication, and particularly relates to an information transmission method and system for a submersible. BACKGROUND

[0002] Underwater submersible clusters are playing an increasingly important role in the fields of ocean exploration, environmental monitoring, underwater search and rescue, and national defense security. Achieving stable and efficient information transmission between submersibles within the cluster and between the cluster and the mother ship is the basis and prerequisite for completing complex collaborative tasks. Due to the rapid attenuation of electromagnetic waves in water, sound waves are currently the only viable medium for underwater long-distance wireless communication. However, when sound waves propagate underwater, they undergo multiple reflections, refractions, and scattering from the sea surface, sea bottom, thermocline, and underwater obstacles, which causes the signal to arrive at the receiver along multiple paths with different time delays and attenuations. This phenomenon is known as the multipath effect. The multipath effect can cause severe inter-symbol interference, i.e., the "echo" of the previous signal can be superimposed with the main wave of the subsequent signal, causing severe distortion of the signal waveform, making it extremely difficult for the receiving end to accurately demodulate the original information.

[0003] To combat the multipath effect, traditional underwater acoustic communication systems usually employ extremely complex equalization algorithms and signal processing techniques at the receiving end, attempting to unravel the original data from the received chaotic signal. This method not only requires extremely high computing power for the receiving end and consumes a large amount of valuable energy, but also has a sharp decline in performance when the channel environment changes rapidly or the multipath effect is extremely severe, resulting in low communication rate and high bit error rate, and even communication interruption. Therefore, how to establish a new method that can fundamentally overcome the limitations of the multipath effect and achieve underwater high-speed, reliable, and multi-user collaborative communication is a key technical problem that needs to be solved in the current underwater information technology field. SUMMARY

[0004] The present application provides an information transmission method and system for a submersible to solve the above technical problems.

[0005] In a first aspect, the present application provides an information transmission method for a submersible, applied to a submersible cluster comprising one or more target receiving submersibles and one or more sending submersibles, the method comprising the following steps: broadcasting a preset wideband acoustic probe signal by the target receiving submersible to one or more sending submersibles, and receiving a distorted probe signal after propagation through the underwater complex environment by one or more sending submersibles; performing parameterized channel modeling on the distorted probe signal by one or more sending submersibles, respectively, decomposing the distorted probe signal into a set of multipath channel parameters including time delay, angle of arrival, amplitude, and phase, and constructing a channel model between the sending submersible and the target receiving submersible; When data needs to be transmitted by one or more sending submersibles to a target receiving submersible, a precoding signal is generated by the sending submersible based on the channel model, wherein when different data to be transmitted is simultaneously transmitted to multiple different target receiving submersibles, different preset acoustic orbital angular momentum phase factors of different modes are loaded for the data to be transmitted to different target receiving submersibles, and a precoding signal superimposed with the acoustic orbital angular momentum phase factors is generated in combination with the set of multipath channel parameters; The precoding signals generated by one or more sending submersibles are cooperatively emitted based on high-precision clock synchronization; The signal that has achieved spatial and temporal focusing at the location of the target receiving submersible through propagation in the underwater complex environment is received by the target receiving submersible, and through adaptive blind equalization processing, the received signal is sharpened according to the ideal impulse response characteristics that the signal should have to eliminate residual inter-symbol interference, and the original data is recovered through demodulation.

[0006] Optionally, the channel modeling of the distorted probe signal by one or more sending submersibles respectively includes the following steps: The time-domain windowing processing is performed on the distorted probe signal; The high-resolution direction-of-arrival estimation algorithm is used to process the distorted probe signal after the time-domain windowing processing, and a plurality of propagation path signals with different spatial angles are identified and separated; For each separated propagation path signal, the time delay and amplitude attenuation corresponding to each propagation path signal are calculated through cross-correlation operation with the wideband acoustic probe signal; The phase shift of each propagation path signal relative to the wideband acoustic probe signal is measured; The calculated time delay, angle of arrival, amplitude attenuation, and phase shift of each propagation path are combined to form a structured set of multipath channel parameters, and the channel model is constructed based on the set of multipath channel parameters.

[0007] Optionally, the precoding signal superimposed with the acoustic orbital angular momentum phase factors of different modes is generated in combination with the set of multipath channel parameters, including the following steps: The channel model is time-reversed and phase-conjugated according to the set of multipath channel parameters, and a basic precoding filter is generated for each target receiving submersible; allocating a unique and mutually orthogonal acoustic orbital angular momentum phase factor to each basis precoding filter according to the number and identity of target receiving submersibles; applying the acoustic orbital angular momentum phase factor to the corresponding basis precoding filter in the form of a complex product to generate a composite precoding filter; performing convolution operation on the data to be transmitted to different target receiving submersibles through the corresponding composite precoding filter to obtain multiple parallel encoded signal streams; linearly superimposing the multiple parallel encoded signal streams to synthesize a precoding signal.

[0008] Optionally, the method further comprises the step of dynamically correcting the channel model after generating the precoding signal: continuously monitoring the communication quality confirmation information fed back by the target receiving submersible; when the communication quality confirmation information indicates that the communication quality is lower than a preset threshold, emitting a low-power differential probe signal from the sending submersible to the target receiving submersible; receiving and processing the response of the low-power differential probe signal to calculate the perturbation change amount of the current channel relative to the previous channel model; updating the multipath channel parameter set based on the perturbation change amount to obtain a corrected channel model; adjusting the basis precoding filter and the acoustic orbital angular momentum phase factor based on the corrected channel model to generate a precoding signal adapted to the current channel state.

[0009] Optionally, the target receiving submersible further comprises the step of canceling crosstalk of multiple signals before performing the adaptive blind equalization processing: configuring the target receiving submersible with parallel receiving channels matching the number of acoustic orbital angular momentum phase factor modes; using the orthogonal demodulation operators configured by each parallel receiving channel to perform preliminary modal separation on the received space-time focused signal to obtain multiple parallel signals containing main signals and residual crosstalk signals; constructing a multiple-input multiple-output adaptive interference cancellation matrix, taking the multiple parallel signals as inputs of the adaptive interference cancellation matrix; iteratively updating the coefficients of the adaptive interference cancellation matrix to minimize the cross-correlation between the output channels of the adaptive interference cancellation matrix; sending each main signal after the adaptive interference cancellation matrix processing and crosstalk suppression to the corresponding adaptive blind equalization processing module.

[0010] Optionally, the receiving, by the target receiving submersible, the spatio-temporally focused signal propagated through the underwater complex environment and realized at the target receiving submersible location, and the demodulating the original data recovered from the spatio-temporally focused signal by adaptive blind equalization processing, includes the following steps: providing a digital adaptive transversal filter inside the target receiving submersible, and presetting a Dirac impulse function as the ideal impulse response characteristic; inputting the spatio-temporally focused signal into the adaptive transversal filter as the input signal; calculating the mean square error between the actual output signal of the adaptive transversal filter and the ideal impulse response characteristic; iteratively updating the internal tap coefficients of the adaptive transversal filter according to the mean square error and using the least mean square error algorithm or the recursive least square algorithm; filtering the subsequently received spatio-temporally focused signal using the updated adaptive transversal filter, and outputting the signal sharpened and free of residual inter-symbol interference for demodulation to obtain the original data recovered.

[0011] Optionally, the cooperative transmitting, by one or more transmitting submersibles, the respective generated pre-coded signals based on high-precision clock synchronization, includes the following steps: designating a master submersible as a master node in the submersible cluster, and designating the remaining submersibles as slave nodes; periodically broadcasting, by the master submersible, a synchronization beacon signal containing the current timestamp of the master submersible; calculating, by the slave submersible, the propagation delay of the synchronization beacon signal after receiving the synchronization beacon signal, and calibrating the local clock of the slave submersible according to the propagation delay and the received timestamp; sending, by the slave submersible, a response signal back to the master submersible, so that the master submersible calculates the round-trip delay and determines the estimate of the clock bias of the slave submersible; setting a future transmission timestamp in the cooperative transmission instruction, and synchronously transmitting the respective generated pre-coded signals when all transmitting submersibles reach the transmission timestamp at their respective calibrated local clocks.

[0012] Optionally, the method further includes the steps of automatic interruption and reconstruction of the communication link: starting a retransmission timer after the transmitting submersible transmits the pre-coded signal, and waiting for the target receiving submersible to return an acknowledgement signal; retransmitting the precoded signal and accumulating a retransmission number if the acknowledgement signal is not received before the retransmission timer expires; determining that the current channel model is invalid when the accumulated retransmission number exceeds a preset upper threshold of retransmission, and aborting the current data transmission task to be sent; controlling the target receiving submersible to rebroadcast the wideband acoustic probe signal to one or more sending submersibles to start a new round of the parameterized channel modeling process.

[0013] In a second aspect, the present application further provides an information transmission system for a submersible, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the information transmission method for a submersible as described above when executing the computer program.

[0014] In a third aspect, the present application further provides a computer-readable storage medium having instructions stored thereon, wherein the instructions, when executed by a processor, cause the processor to be configured to perform the information transmission method for a submersible as described above.

[0015] The present application has the following beneficial effects: The present application changes the complex underwater environment from a communication barrier to an energy focusing lens, no longer passively compensating for signal distortion at the receiving end, but actively constructing a signal waveform that can achieve spatial and temporal focusing at the target location at the sending end, greatly improving the energy efficiency and reliability of signal transmission and significantly expanding the effective communication distance. Moreover, the present application introduces acoustic orbital angular momentum as a new dimension for multi-user differentiation, making it possible to perform parallel communication without crosstalk to multiple targets in the same frequency band by loading different modal phase factors for data of different receivers, thereby improving the communication capacity of a submersible cluster from a single-point link to a truly networked collaborative level, and achieving an order-of-magnitude increase in communication capacity. In addition, the adaptive blind equalization processing at the receiving end as the final guarantee can intelligently sharpen the focused signal and eliminate residual interference, ensuring high accuracy of data recovery. The present application builds a high-robustness, high-capacity collaborative information transmission system for a cluster of underwater submersibles. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of the information transmission method for a submersible in one embodiment of the present application.

[0017] Figure 2 A flowchart of constructing a channel model between a sending submersible and a target receiving submersible in one embodiment of the present application.

[0018] Figure 3A flowchart of a process for generating a precoded signal superimposed with an acoustic orbital angular momentum phase factor in one embodiment of the present application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be clearly described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art belong to the scope of protection of the present application.

[0020] The terms "first", "second", and the like in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally of a kind and are not limited in number, for example, the first object can be one or more. In addition, "and / or" in the specification and claims indicates at least one of the connected objects, and the character " / ", generally indicates that the front and rear associated objects are in an "or" relationship.

[0021] Figure 1 A flowchart of a process for an information transmission method for a submersible in one embodiment. It should be understood that, although Figure 1 the steps in the flowchart are displayed in sequence according to the direction of the arrows, these steps are not necessarily executed in sequence according to the direction of the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, Figure 1 at least part of the steps in the flowchart can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or sub-steps or stages of other steps. As Figure 1 shown in the flowchart, the information transmission method for a submersible disclosed in the present application is applied to a submersible cluster including one or more target receiving submersibles and one or more sending submersibles, and specifically includes the following steps: S101. Broadcasting a preset wideband acoustic probe signal by the target receiving submersible to one or more sending submersibles, and receiving a distorted probe signal after propagation through an underwater complex environment by the one or more sending submersibles.

[0022] The channel sounding process is initiated by the target receiver submersible as the information sink, which broadcasts a pre-designed, wide-spectrum acoustic sounding signal to one or more predetermined sender submersibles as the information source. This wideband signal, such as a chirp signal, is chosen for its good autocorrelation properties and ability to resist multipath interference. When the signal propagates underwater, it will experience multipath effects caused by reflections from the water surface and bottom, as well as refractions from different medium layers in the water, accompanied by signal strength attenuation and phase distortion, resulting in a severely distorted version of the original signal received by the sender submersible. Each sender submersible independently captures this distorted signal containing rich channel information. The essence of this step is to have the target receiver actively provide a reference, forcing the underwater environment to imprint its transmission characteristics on the reference signal, so that the sender can passively and safely obtain an acoustic channel response snapshot from its location to the target location.

[0023] S102. Parameterized channel modeling of the distorted sounding signal by one or more sender submersibles to decompose the distorted sounding signal into a set of multipath channel parameters including time delay, angle of arrival, amplitude, and phase, and to construct a channel model between the sender submersible and the target receiver submersible.

[0024] After obtaining the distorted sounding signal, each sender submersible needs to deeply analyze the original data and convert it into a mathematical model for calculation and utilization. This parameterized channel modeling process aims to decompose the continuous and complex distorted waveform into a set of discrete and clearly physically meaningful parameters. In specific implementation, first, a high-resolution direction-of-arrival estimation algorithm is used to process the received distorted signal. This kind of algorithm can accurately identify signal components arriving from different spatial angles, thereby separating multiple propagation paths mixed together in the spatial dimension. For each separated path signal, cross-correlation operation with the original wideband acoustic sounding signal can accurately calculate the propagation time delay and signal amplitude attenuation of the path relative to the direct path. Finally, all path parameters such as time, angle of arrival, amplitude, and phase are integrated to form a structured multipath channel parameter set.

[0025] S103. When data needs to be sent by one or more sender submersibles to the target receiver submersible, a pre-coded signal is generated by the sender submersible based on the channel model, wherein when different to-be-sent data are simultaneously sent to multiple different target receiver submersibles, different preset modal acoustic orbital angular momentum phase factors are loaded for the to-be-sent data sent to different target receiver submersibles, and the pre-coded signal superimposed with the acoustic orbital angular momentum phase factor is generated in combination with the multipath channel parameter set.

[0026] Wherein, after the establishment of the precise channel model, when the sending submersible needs to send actual data to the target receiving submersible, the pre-coding generation process of the signal is started, and the channel model is used to pre-process the signal to be sent, so that after the signal propagates through the actual underwater channel, various distortions can be exactly offset, and finally perfectly converge at the target position. This process is based on the time reversal principle, that is, the conjugate of the channel model is reversed in time. When different data needs to be sent to multiple different target receiving submersibles at the same time, in order to avoid information confusion, the acoustic orbital angular momentum technology is introduced. A unique and mutually orthogonal acoustic orbital angular momentum mode is loaded for the data stream sent to different targets, which is equivalent to giving each signal a unique spiral phase tag. Finally, all processed signal streams are linearly superimposed to form a composite pre-coded signal.

[0027] S104. Based on high-precision clock synchronization, one or more sending submersibles cooperatively emit the pre-coded signals generated by each sending submersible.

[0028] Wherein, if there is a deviation in the emission time of each sending submersible, the phase information carried by the pre-coded signal will be disordered during propagation, resulting in a significant reduction in focusing effect or even complete failure. High-precision clock synchronization is usually achieved using a master-slave synchronization scheme. In the submersible cluster, a submersible is pre-designated as the master node, and the remaining submersibles are slave nodes. The master node periodically broadcasts a synchronization beacon containing its current precise timestamp. Upon receiving the beacon, the slave nodes estimate the propagation delay of the signal based on the propagation speed and distance of the signal, and then calibrate their local clocks based on the received master node timestamp and propagation delay. The calibration relationship can be expressed as: tlocal = tmaster + tpropagation. To improve accuracy, the slave nodes also send response signals back to the master node, allowing the master node to calculate the round-trip delay and more accurately estimate and compensate for the clock deviation of each slave node. When the cooperative emission instruction is issued, the instruction will contain a future unified emission timestamp. All sending submersibles emit their respective pre-coded signals simultaneously at the specified timestamp based on their calibrated local clocks.

[0029] S105. The target receiving submersible receives the signal that has been spatially and temporally focused at the target receiving submersible's location through the complex underwater environment, and through adaptive blind equalization processing, sharpens the received spatially and temporally focused signal based on the ideal impulse response characteristics that the spatially and temporally focused signal should have to eliminate residual inter-symbol interference, and demodulates to recover the original data.

[0030] The signal is transmitted cooperatively and propagates underwater, and finally forms a strong signal pulse focused in space and time at the location of the target receiving submersible. Although the pre-coding process has greatly suppressed channel distortion, due to slight dynamic changes in the environment or inherent errors in channel modeling, there may still be residual inter-symbol interference in the received signal, i.e. there is tailing and overlap between the front and rear signal pulses, affecting the accuracy of data demodulation. To solve this problem, the target receiving submersible will start an adaptive blind equalization processing flow. The core of blind equalization is that the receiving end corrects the signal based on the ideal statistical characteristics or structural characteristics of the signal without knowing the specific content of the original transmitted data. In this method, the ideal signal characteristic is set as a perfect Dirac impulse function, because it represents a single data symbol without any inter-symbol interference. The received focused signal is sent to an adaptive transversal filter with adjustable internal tap coefficients. The filter calculates the mean square error between its actual output and the ideal impulse response. Then an iterative algorithm such as the least mean square error is used to continuously adjust the tap coefficients of the filter according to the error. Through repeated iterations, the filter can automatically learn and compensate for the residual distortion in the channel that has not been completely eliminated by pre-coding, sharpening the received signal pulse and making its shape approximate to the ideal impulse function.

[0031] In one embodiment, the channel model between the transmitting submersible and the target receiving submersible is constructed by one or more transmitting submersibles respectively performing parametric channel modeling on the distorted probe signal, decomposing the distorted probe signal into a set of multipath channel parameters including time delay, angle of arrival, amplitude and phase. S201. Time domain windowing processing is performed on the distorted probe signal; S202. A high-resolution direction of arrival estimation algorithm is used to process the distorted probe signal after time domain windowing processing, and a plurality of propagation path signals with different spatial angles are identified and separated; S203. For each separated propagation path signal, the time delay and amplitude attenuation corresponding to each propagation path signal are calculated by cross-correlation operation with the wideband acoustic probe signal; S204. The phase shift of each propagation path signal relative to the wideband acoustic probe signal is measured; S205. The calculated time delay, angle of arrival, amplitude attenuation and phase shift of each propagation path are combined to form a structured multipath channel parameter set, and a channel model is constructed based on the multipath channel parameter set.

[0032] In this embodiment, due to the multipath effect of the underwater channel, the response of the probed signal lasts for a period of time, but its main energy and information are concentrated within a finite time range. To focus the analysis on this effective signal while avoiding spectral leakage introduced by directly truncating the signal, windowing is necessary. This process multiplies the received distorted signal sequence point-by-point with a window function of finite length and smooth transitions at both ends, such as a Hanning window or a Hamming window. This operation smoothly attenuates the amplitude of the beginning and end portions of the signal to zero, ensuring a smooth transition of the signal segment and thus greatly suppressing spectral leakage. After completing the time-domain windowing, the next core task is to identify and separate the signal components from different physical paths from the windowed signal. Since these paths originate from different spatial directions, their directions of arrival (DOA) can be determined using a sensor array deployed on the transmitting submersible. To accurately distinguish paths with extremely similar angles in a complex multipath environment, a high-resolution ODA estimation algorithm, such as a multi-signal classification algorithm, is required. The principle of this algorithm is to first calculate the covariance matrix of the data received by the sensor array, and then perform eigenvalue decomposition on the matrix to divide the signal space into a signal subspace dominated by the signal and a noise subspace dominated by the noise.

[0033] Since the signal propagation direction vector is necessarily orthogonal to the noise subspace, all possible angles of arrival can be searched by constructing a spatial spectrum function. The expression for this spatial spectrum function is: .in It corresponds to the direction angle. The array guide vector, It is a matrix composed of eigenvectors of the noise subspace, and H represents the conjugate transpose operation. When in a certain direction... When a signal path exists, the corresponding steering vector will be nearly orthogonal to the noise subspace, causing the denominator to approach zero, thus forming a sharp peak in the spatial spectrum. By searching for these peaks, the angle of arrival of each multipath signal can be accurately identified, ultimately separating the multipath signals that were originally mixed together in the time domain into clear separations in the spatial domain, laying the foundation for analyzing the characteristics of each path one by one.

[0034] After the spatial angle of each propagation path is accurately identified, the specific parameters of these paths in time and amplitude, i.e. time delay and amplitude attenuation, need to be further determined. This task is accomplished by cross-correlation operation. Cross-correlation is a mathematical tool to measure the similarity of two signals at different time delays. In specific operation, the original broadcasted, undistorted wideband acoustic probe signal is taken as the reference signal, and cross-correlation calculation is performed with each spatially separated propagation path signal. When the time axis of the reference signal is shifted to a certain specific value relative to the path signal, if the waveforms of the two are highly matched at this position, a peak value of the cross-correlation function will appear. The time shift corresponding to this peak value accurately reflects the propagation time delay of the path signal relative to the direct path or the reference time zero point . At the same time, the amplitude of this peak value, after normalization processing, directly corresponds to the amplitude attenuation of the path signal in the propagation process . In addition to time delay and amplitude, the phase change experienced by the signal in the propagation process is also a key component of the channel characteristics, especially for time reversal technology which requires coherent processing. Therefore, the phase shift introduced by each propagation path signal must be accurately measured. This information can also be extracted from the cross-correlation operation results of the previous step.

[0035] When calculating the cross-correlation function, the obtained peak value is actually a complex number, which contains not only amplitude information but also phase information. The phase angle of this complex peak value represents the phase shift of the path signal relative to the original wideband acoustic probe signal . Specifically, if the cross-correlation peak value calculated at time delay is a complex number , then the phase shift of the path is the argument of , i.e. . This phase shift comprehensively reflects the phase velocity change of the signal in the propagation medium, as well as the phase jump occurring in the interface reflection and other physical effects. After completing this measurement step, the four basic physical parameters of each propagation path, i.e. arrival angle, time delay, amplitude attenuation and phase shift, are all accurately obtained. Finally, these discrete parameters are systematically combined to construct a structured channel model that can fully describe the complex underwater acoustic environment between the sending submersible and the target receiving submersible.

[0036] The structured channel model is mathematically represented as the channel impulse response. In implementation, the time delay, angle of arrival, amplitude attenuation and phase shift calculated for each propagation path k in the previous steps are integrated. Each path can be modeled as a Dirac impulse function with a specific delay, scaling and phase shift. By linearly superimposing the responses of all L identified paths, the final set of multipath channel parameters and the corresponding channel model H are formed. The model can be expressed as:

[0037] where L is the total number of identified multipaths, represents the Dirac impulse function. This mathematical expression reproduces the effect that would be produced at the receiving end when an ideal, infinitely narrow acoustic impulse is transmitted through the underwater channel.

[0038] In one embodiment, the pre-coded signal is generated by loading different preset acoustic OAM phase factors to the data to be transmitted to different target receiving submersibles, and combining the multipath channel parameter set to generate a pre-coded signal superimposed with acoustic OAM phase factors, including the following steps: S301. Time reversal and phase conjugation processing is performed on the channel model according to the multipath channel parameter set, to generate a basic pre-coding filter for each target receiving submersible; S302. A unique and mutually orthogonal acoustic OAM phase factor is assigned to each basic pre-coding filter according to the number and identification of the target receiving submersibles; S303. The acoustic OAM phase factor is applied to the corresponding basic pre-coding filter in the form of a complex product to generate a composite pre-coding filter; S304. The data to be transmitted to different target receiving submersibles are respectively convolved by the corresponding composite pre-coding filter to obtain a plurality of parallel encoded signal streams; S305. The plurality of parallel encoded signal streams are linearly superimposed to form a pre-coded signal.

[0039] In this embodiment, the primary task of the pre-coded signal generation process is to tailor a basic signal shaping tool, i.e. a pre-coding filter, for each target receiving submersible. The core of this process is the application of the time reversal principle. This principle states that if the effects of signal propagation in the channel are recorded, and then the recorded effects are reversed in time and phase conjugated, and then applied to the original signal, the complex multipath effects of the channel itself will become a kind of boost, so that the signal will automatically focus at the original receiving point when it propagates through the same channel. In implementation, the channel model previously constructed, i.e. the channel impulse response , a mathematical transformation is performed. Time reversal corresponds to replacing the time variable t with -t, and phase conjugation is taking the conjugate of all complex values in the model. Thus, the basis pre-coding filter generated for the i-th target receiver can be expressed as where the asterisk represents the conjugate operation.

[0040] In order to transmit independent data to multiple different target receivers in the same time, same frequency band without interference, an additional dimension of differentiation needs to be introduced on top of time reversal. Acoustic orbital angular momentum technology provides an ideal solution for this purpose. Acoustic orbital angular momentum imparts a helical wavefront structure to the sound wave, and its mode number determines the degree and direction of the twist of the helix. Acoustic orbital angular momentum beams of different mode numbers are physically orthogonal to each other, which means they can propagate in space overlapping each other without confusion. According to the number of target receivers that need to be communicated simultaneously, a unique and mutually orthogonal acoustic orbital angular momentum phase factor is assigned to each basis pre-coding filter. For example, the first target can be assigned a mode number , the second target a mode number , and so on, where all are different integers. This phase factor mathematically manifests as a complex exponential function related to the azimuthal angle . Here j is the imaginary unit.

[0041] After generating the basis pre-coding filter for each target and the unique acoustic orbital angular momentum phase factor, the two need to be combined to form a composite pre-coding filter. This composite filter contains both the time reversal information that enables spatial focusing and the acoustic orbital angular momentum mode information that differentiates between different receivers. The way this combination is achieved is by performing a complex multiplication operation. The acoustic orbital angular momentum phase factor assigned to a particular target receiver is applied to the corresponding basis pre-coding filter in the form of a complex multiplication. This operation grafts a helical phase structure onto each component of the time reversal filter. The composite pre-coding filter generated by this can be expressed as the product of the basis filter and the phase factor: . Through this step, a highly specialized and customized composite filter is generated for each target. This filter not only knows how to direct the signal to overcome multipath effects to reach the target location, but also pre-encodes the data stream it carries with a specific helical wavefront morphology.

[0042] Next, the original data to be transmitted is modulated into a baseband signal. Then, this data signal representing specific information is convolved with a composite precoding filter tailored for its target receiver vehicle. Convolution is a mathematical operation whose physical meaning is that each instant value of one signal is replaced by the weighted sum response of another signal. This process completely imprints the complex time delay, phase rotation and helical phase structure contained in the composite filter onto the original data signal. In this way, a corresponding, deeply encoded signal stream is generated for each piece of data to be transmitted to different target receiver vehicles . This process is parallelized, i.e. all data streams to be transmitted are simultaneously and independently convolved through their respective composite filters. After generating independent, deeply encoded signal streams for all target receiver vehicles, the last step is to integrate these parallel signal streams into a single, final precoding signal. This integration process is very straightforward, i.e. linear superposition is used. All parallel processed encoded signal streams are point-by-point added in the time domain. If there are N target receiver vehicles, the final synthesized precoding signal is the sum of all encoded signal streams: .

[0043] The reason why this direct addition is feasible is that the acoustic orbital angular momentum phase factors loaded for each signal stream are mutually orthogonal. It is this orthogonality that ensures that even if all signals are mixed together and emitted, they can be effectively distinguished during propagation and demodulation at the respective target receiver, without causing serious mutual interference.

[0044] In one embodiment, the method further comprises the step of dynamically modifying the channel model after generating the precoding signal: continuously monitoring the communication quality confirmation information fed back by the target receiver vehicle; when the communication quality confirmation information indicates that the communication quality is below a preset threshold, transmitting a low-power differential probe signal by the transmitting vehicle to the target receiver vehicle; receiving and processing the response of the low-power differential probe signal, and calculating the perturbation change of the current channel relative to the previous channel model; updating the set of multipath channel parameters based on the perturbation change to obtain a modified channel model; adjusting the base precoding filter and the acoustic orbital angular momentum phase factor based on the modified channel model to generate a precoding signal adapted to the current channel state.

[0045] In this embodiment, after successfully demodulating the data, the target receiving submersible calculates a series of key performance indicators, such as signal-to-noise ratio, bit error rate, or mean square error of equalizer output, based on the quality of the received signal. These indicators are quantified and packaged into short communication quality confirmation information, which is fed back to the transmitting submersible through a low-rate uplink. The transmitting submersible continuously receives and analyzes these feedback information, and compares the performance indicators in them with a pre-set threshold representing the lower limit of acceptable communication quality in real time. This process constitutes a complete feedback loop, enabling the transmitting end to monitor the signal quality at the remote receiving point in real time and non-intrusively. Once it is detected that the communication quality confirmation information fed back by the target receiving submersible indicates that the current link performance has fallen below the pre-set threshold, the transmitting submersible will immediately initiate the channel correction process. Unlike the initial omnidirectional channel sounding, the target at this time is not to rebuild the model from scratch, but to quickly capture the small changes in the channel. To this end, the transmitting submersible transmits a specially designed low-power differential sounding signal to the target receiving submersible. The reason for using low power is to conduct the sounding without interrupting the main data communication, while also saving energy and reducing the probability of being detected by non-cooperative parties. The principle of designing the differential sounding signal is to highlight the differences between the new and old channels. One way to achieve this is to convolve the original wideband sounding signal with the conjugate time-reversed form of the previously established channel model to generate this differential signal. When this signal passes through the actual, already changed channel, its response at the receiving end will no longer be an ideal impulse, and its distortion pattern directly reflects the degree of deviation of the current channel from the old model.

[0046] When the target receiving submersible receives the response of the low-power differential sounding signal, it returns the response signal containing channel change information to the transmitting submersible. The core task of the transmitting submersible after receiving this response is to accurately calculate the perturbation change of the current channel relative to the last channel model from it. The principle of processing the response is to compare it with an ideal response. The ideal response refers to the waveform that the differential sounding signal should present after passing through the channel if the channel has not changed at all, which is usually a sharp pulse close to the Dirac impulse function. The difference between the actual received response and this ideal response constitutes the perturbation change of the channel. This perturbation change can be modeled mathematically as a separate perturbation channel model superimposed on the original channel model. Therefore, the true channel model at the current time can be approximately represented as the linear addition of the old channel model and the calculated perturbation change.

[0047] After the macroscopic perturbation is calculated, its influence needs to be decomposed and applied to the specific set of multi-path channel parameters that constitute the channel model. The channel model is described by a series of parameters with clear physical meaning (time delay, angle of arrival, amplitude, phase), so updating the model means updating these parameters. This process is not a complete high-resolution parameter estimation, but a fine-tuning of the original parameters based on the perturbation. Various tracking algorithms can be used, such as Kalman filter, which treats each multi-path parameter as a state vector and the calculated perturbation as a new observation. The algorithm will correct the estimate of each parameter based on the new observation. For example, the amplitude parameter of the kth path can be updated as where the amplitude correction is obtained by analyzing the energy change in the perturbation model corresponding to the path delay and angle. The update of time delay, angle of arrival and phase also follows similar logic.

[0048] After successfully updating the multi-path channel parameter set based on the perturbation and obtaining the corrected channel model, the entire adaptive adjustment process enters the final and most critical stage: applying the new channel model to optimize signal transmission. The basic precoding filter used to generate the precoded signal and the related acoustic orbital angular momentum factor are calculated based on the old channel model and are no longer optimal. Therefore, they must be re-adjusted based on the corrected channel model. In implementation, first, an updated basic precoding filter is recalculated using the new channel model through time reversal and phase conjugation processing. Then, the new basic precoding filter is combined with the original acoustic orbital angular momentum phase factor assigned to distinguish different targets to generate a new composite precoding filter. Finally, the subsequent data to be sent will be processed through this newly generated composite precoding filter that fully adapts to the current channel state to generate a new precoded signal.

[0049] In one embodiment, the target receiving submersible further includes a step of canceling the crosstalk of the multi-path signal before performing adaptive blind equalization processing: The target receiving submersible is configured with parallel receiving channels matching the number of acoustic orbital angular momentum phase factor modes; Using the orthogonal demodulation operators configured in each parallel receiving channel, the received spatio-temporally focused signal is preliminarily separated into multiple parallel signals containing the main signal and residual crosstalk signals; An adaptive interference cancellation matrix is constructed, and the multi-path parallel signals are used as the input of the adaptive interference cancellation matrix; The coefficients of the adaptive interference cancellation matrix are iteratively updated to minimize the cross-correlation between the output channels of the adaptive interference cancellation matrix. The main signals after adaptive interference cancellation matrix processing and crosstalk suppression are sent to the respective adaptive blind equalization processing module.

[0050] In this embodiment, if the transmitter broadcasts three modalities (e.g. modalities number -1, +1, +2) simultaneously to three different targets, the receiver needs to establish three parallel digital signal processing channels. Physically, this usually relies on a ring array composed of multiple hydrophone units, which can completely capture the spatial information of the arriving sound field, especially the phase variation in the azimuth. In the software layer, the raw data from the array is simultaneously sent into the three independent virtual channels. Each channel is pre-configured to be responsible for decoding a specific acoustic orbital angular momentum modality. After the establishment of parallel receiving channels, each channel uses the specific orthogonal demodulation operator assigned to it to perform preliminary modal separation on the received mixed signals that have been spatially focused. The core principle of the orthogonal demodulation operator is to use the mathematical orthogonality between acoustic orbital angular momentum modalities. The transmitter loads the phase factor where is the modality number of the i-th signal, is the azimuth. Then in the corresponding i-th channel at the receiver, a demodulation operator is configured which is the conjugate of the operator. Multiply the received mixed signal with this demodulation operator over the entire azimuth range and integrate, theoretically only the signal components with the same modality as the operator can produce a non-zero coherent accumulation result, while the signal components of other modalities will be cancelled due to orthogonality. The mathematical expression of this process is: , where is the received signal field, is the i-th channel signal after the initial separation. However, due to channel distortion, array geometry error, and other non-ideal factors, this orthogonality will be partially destroyed, resulting in a small amount of residual crosstalk signals from other modes being mixed into the separation results in addition to the expected main signal. In order to remove the residual crosstalk signals that still exist after the initial modal separation, a more powerful adaptive processing mechanism needs to be introduced. The core of this mechanism is to construct an adaptive interference cancellation matrix for multiple inputs and multiple outputs. This matrix is a digital signal processing module in concept, which has an equal number of input ports and output ports as the number of parallel receiving channels. The multiple channels of parallel signals containing the main signal and residual crosstalk obtained in the previous step are fed into the input end of this matrix as a signal vector. The matrix is composed of a series of adjustable coefficient digital filters inside, which can perform complex linear combinations on the input signals. The design goal is to actively cancel the interference components from other channels contained in each output signal by appropriately weighting, delaying, and subtracting each channel signal. For example, it will learn to extract crosstalk features from the second and third channel signals, and then subtract these features from the first channel signal.

[0051] After constructing the adaptive interference cancellation matrix, it must be given a clear optimization goal and a set of learning rules so that it can automatically adjust the internal coefficients to achieve the best interference cancellation effect. The optimization goal is set to minimize the cross-correlation between the output channels of the matrix. Cross-correlation is an indicator that measures the statistical correlation between two signals. If there is still a high correlation between two output signals, it means that they still share some common components, i.e., residual crosstalk. Therefore, the system will continuously calculate the cross-correlation values between any two different output signals, and the sum of the squares of all these cross-correlation values will form a cost function J. The goal of the algorithm is to minimize this cost function J by adjusting the matrix coefficients, which is expressed as . Where represents the expected value, and are the i-th and k-th output signals of the matrix, and the asterisk represents the complex conjugate. Using iterative algorithms such as gradient descent, the coefficients of the matrix will be continuously fine-tuned in the direction that reduces the cost function until the signals of each output channel become statistically independent of each other.

[0052] When the adaptive interference cancellation matrix is updated by iteration, the internal coefficients converge to the optimal state, and the crosstalk between the signals obtained from the output end of the adaptive interference cancellation matrix is maximally suppressed. At this time, each output signal can be regarded as an independent and clean communication link. However, although the interference between signals is eliminated, each signal itself can still be subject to intersymbol interference caused by its own propagation path, that is, the tailing and broadening of the signal pulse. Therefore, the last step of this processing flow is to send the main signals after the interference cancellation matrix processing, whose crosstalk has been effectively suppressed, to their respective self-adaptive blind equalization processing modules respectively and independently. The subsequent blind equalization processing will focus on solving the intersymbol interference problem of a single signal without worrying about the interference from other modal signals. Finally, the problem is decomposed and isolated by first solving the crosstalk between multiple signals and then solving the distortion within a single signal, ensuring the order and efficiency of the entire signal processing link.

[0053] In one embodiment, the signal that is spatially and temporally focused at the target receiving submersible location by propagating through the underwater complex environment is received by the target receiving submersible, and the received spatially and temporally focused signal is sharpened by adaptive blind equalization processing according to the ideal impulse response characteristics that the spatially and temporally focused signal should have to eliminate residual intersymbol interference, and the original data is recovered by demodulation, including the following steps: A digital adaptive transversal filter is provided inside the target receiving submersible, and a Dirac impulse function is preset as the ideal impulse response characteristic; The spatially and temporally focused signal is used as the input signal of the adaptive transversal filter; The mean square error between the actual output signal of the adaptive transversal filter and the ideal impulse response characteristic is calculated; According to the mean square error, the internal tap coefficients of the adaptive transversal filter are iteratively updated using the least mean square error algorithm or the recursive least squares algorithm; The updated adaptive transversal filter is used to filter the subsequently received spatially and temporally focused signal, and the signal that is sharpened and has eliminated residual intersymbol interference is output for demodulation to obtain the recovered original data.

[0054] In this embodiment, to effectively eliminate the residual inter-symbol interference still existing in the spatially and temporally focused signal after precoding and channel propagation, an adaptive transversal filter is constructed inside the digital signal processing unit of the target receiving submersible. This filter is essentially a finite impulse response structure, consisting of a series of delay units and adjustable weighting coefficients (i.e. tap coefficients). More importantly, to guide the filter to optimize itself, it must be set with an explicit and ideal performance target. In this technical solution, the ideal target is pre-set as the Dirac impulse function. The Dirac impulse function is a mathematically perfect impulse, characterized by having an infinite amplitude at the time zero point and zero amplitude at all other time points, with a total energy of one. It is chosen as the ideal impulse response characteristic because it represents a signal form without any time spread and without any front and back tail, and is the ultimate embodiment of eliminating all inter-symbol interference.

[0055] After the adaptive transversal filter and its ideal target are set, the actual signal processing process is officially started. The spatially and temporally focused signal from the receiving array after preliminary processing is sent to the input end of the adaptive transversal filter as a sequence of continuous digital sampling points. Although the energy of this input signal has been highly concentrated, its pulse form will still exhibit a certain degree of spread and oscillation due to the subtle dynamic changes of the channel or modeling errors, which are the residual inter-symbol interference that needs to be eliminated. When the signal sequence flows through the transversal filter, it moves on the internal delay chain, and each delay unit outputs a signal version that is delayed by one sampling period relative to the previous unit. At each time point, the filter multiplies the outputs of all delay units by their respective tap coefficient values at the current time point, and then adds all the products to generate a final output signal. The output can be represented as where is the input signal delayed by k steps, is the kth tap coefficient, and N is the total order of the filter.

[0056] After the preliminary output signal is generated, the core link of the adaptive mechanism, performance evaluation, begins. The goal of this link is to accurately quantify the gap between the current filter output and the pre-set ideal target. Specifically, the mean square error between the actual output signal of the adaptive transversal filter and the Dirac impulse function as the ideal impulse response characteristic is calculated. First, at each sampling time n, the instantaneous error between the two is calculated, which is defined as the difference between the ideal signal and the actual output signal: Since instantaneous errors can be positive or negative, direct summation will cancel each other out and fail to reflect the overall error size. Therefore, it is necessary to square the instantaneous error to get the instantaneous squared error This ensures that all error contributions are positive and larger errors are given higher weights. Finally, by averaging the instantaneous squared error over a period of time, the mean squared error is obtained. This value is a non-negative scalar that intuitively reflects the degree to which the filter output signal deviates from the ideal impulse function in shape. A smaller mean squared error means that the output signal has sharper pulses and less intersymbol interference, while a larger value is the opposite.

[0057] After calculating the mean squared error as a performance evaluation indicator, the next step is to use this error information to intelligently and iteratively update the internal tap coefficients of the adaptive transversal filter, so that its performance evolves towards the direction of optimization. This task is usually accomplished by the least mean squared error algorithm, which is known for its simple calculation and stable performance. The principle of the least mean squared error algorithm is to search along the negative gradient direction of the mean squared error performance surface to find the point with the smallest error as quickly as possible. In each iteration period, the algorithm calculates the adjustment amount of the tap coefficient vector according to the currently calculated error signal and the input signal vector of the filter. The mathematical expression of coefficient update is In this formula, is a normal number called step factor, which controls the amplitude of each iteration adjustment and the convergence speed. This update process is continuous: the filter processes the input signal to get the output, calculates the error, then updates the coefficients using the error, and then processes the next input signal with the new coefficients, and so on. Through this iterative process, the tap coefficients of the filter will gradually converge, so that the filter itself is more and more close to an inverse filter that can perfectly eliminate channel residual distortion.

[0058] After a sufficient number of iterations, the tap coefficients of the adaptive transversal filter will gradually converge and stabilize around a set of optimal values, at which point the performance of the filter is at its best. After this, the filter, which has been "trained", can be used to perform real-time filtering on the subsequent continuously received spatio-temporally focused signals. When the subsequent signal stream passes through this optimized filter, the waveform of its output signal will be greatly sharpened, with the main lobe energy of the signal pulse more concentrated, and the sidelobes and tails caused by intersymbol interference significantly suppressed, so that the entire signal form is maximally approximated to the ideal Dirac impulse function. This clean signal stream, which has been sharpened and residual intersymbol interference eliminated, is then sent to the final demodulator. The demodulator can make decisions on the amplitude or phase of each signal pulse with little interference, thereby accurately recovering the original transmitted binary data. Finally, the last obstacle affecting data recovery can be completely removed, greatly reducing the bit error rate of communication, and ensuring that even in complex and variable underwater environments, information transmission can meet the requirements of high reliability and high fidelity.

[0059] In one embodiment, the transmitting one or more submersibles cooperatively transmitting the respective generated precoded signals based on the high-precision clock synchronization comprises the following steps: designating a master submersible in the cluster of submersibles and the remaining submersibles as slave submersibles; periodically broadcasting, by the master submersible, a synchronization beacon signal containing a current timestamp of the master submersible; calculating, by the slave submersible, a propagation delay of the synchronization beacon signal upon receiving the synchronization beacon signal, and calibrating the local clock of the slave submersible according to the propagation delay and the received timestamp; sending, by the slave submersible, a response signal back to the master submersible, so that the master submersible calculates a round-trip delay and determines an estimate of the clock bias of the slave submersible; setting a future transmission timestamp in the cooperative transmission instruction, and synchronously transmitting the respective generated precoded signals when all transmitting submersibles reach the transmission timestamp on their respective calibrated local clocks.

[0060] In this embodiment, the selection of the master node can be based on a pre-configured arrangement or determined through a dynamic election protocol, which can take into account factors such as the long-term stability of each vehicle's clock, the remaining power, or whether it can receive an external high-precision time signal. The vehicle designated as the master node will serve as the sole time authority for the entire cluster, with its local clock becoming the reference for all slave nodes to calibrate against. The role of the slave nodes is to passively listen for time information from the master node and actively adjust their local clocks to align with the master node. This master-slave structure avoids conflicts and oscillations caused by multiple time standards in the network, creating an ordered, stable, and traceable synchronization framework. After establishing the master-slave structure, the time synchronization process is initiated by the master node. The master vehicle will periodically broadcast a specially marked synchronization beacon signal in all directions. The core of this broadcasting process is that the master node captures the current time reading of its high-precision local clock at the precise moment when the acoustic transmission system is about to emit the beacon signal. This reading is called the time stamp. Subsequently, this time stamp containing specific time, minute, second, and higher-precision time information is encoded into digital information and embedded as the core data payload in the acoustic waveform of the synchronization beacon signal. This process is repeated periodically, for example, once every second, to ensure that slave nodes can continuously track the small drift of the master node clock and enable new vehicles to quickly integrate into the synchronization network.

[0061] When a slave vehicle receives the synchronization beacon signal from the master node in its hydrophone, it immediately starts the calibration procedure for the local clock. First, the slave node records the local time stamp of receiving the beacon. Then, it decodes the data payload of the beacon signal to extract the original time stamp sent by the master node. At this point, the slave node obtains two key time points, but it cannot simply set its clock to the original time stamp because the beacon signal takes a certain amount of time to propagate from the master node to the slave node, i.e., the propagation delay. To perform accurate calibration, the slave node must estimate this propagation delay. At the initial synchronization, this delay can be preliminarily estimated by dividing the approximate distance between the vehicles by the speed of sound in water. Subsequently, the slave node can calculate the correct time that the master node's clock should indicate when the signal arrives, i.e., the master node's sending time stamp plus the propagation delay. The correct time of the slave node's local clock should be should be where is the master node's sending time stamp decoded from the beacon, is the estimated signal propagation delay. Finally, the slave node compares this calculated correct time with its recorded reception time, obtains the deviation between the two, and adjusts the counter of its local clock according to the deviation to complete a calibration.

[0062] To overcome the errors caused by inaccurate one-way propagation delay estimation and achieve higher-precision clock synchronization, a refinement step involving bidirectional time exchange is required. After the slave node completes its initial calibration, it immediately sends a response signal back to the master node. This process involves four key timestamps: the master node at time... Send the initial beacon; the slave node at its local time Received beacon; slave node at its local time Send a response signal; the master node at its local time A response signal has been received. When sending a response signal, the slave node will include its recorded... and These two timestamps are included in the signal data. Upon receiving the response, the master node possesses all four timestamps. At this point, the master node can accurately calculate the one-way propagation delay of the signal in the water, because the total round-trip time... Subtract the time the signal stays inside the node. This is exactly equal to the pure propagation time of the signal traveling back and forth twice. Therefore, the one-way propagation delay... The exact value is: The master node can use this precisely calculated propagation delay to deduce the precise deviation of the slave node's clock relative to its own clock, and send this deviation value to the slave node in the next broadcast, or use it directly in subsequent coordination commands.

[0063] After all transmitting submersibles have achieved high-precision clock synchronization with the master node through the aforementioned steps, the foundation for executing precise coordinated actions is established. To ensure that all submersibles transmit signals at exactly the same instant, the design of the coordinated transmission command must take into account the time required for the command itself to propagate through the network. Therefore, the command does not contain an "immediately transmit" order, but instead sets a specific, future absolute transmission timestamp. This future transmission timestamp is set sufficiently late to ensure that even the submersible furthest from the command originating point and with the slowest reception has ample time to receive and parse the entire command. Upon receiving the command, each submersible acting as a transmitting node extracts this future transmission timestamp and sets a countdown or alarm clock based on its local clock. Then, the submersible enters a standby state, quietly waiting for its calibrated local clock to reach the precise moment specified in the command. When the local clocks of all transmitting submersibles simultaneously reach the preset transmission timestamp, they synchronously and instantaneously activate their respective transmitting transducers, transmitting the pre-generated precoded signal.

[0064] In one implementation, the method further includes an automatic interruption and reconstruction step of the communication link: After sending the precoded signal transmitted by the submersible, start the retransmission timer and wait for the target to receive the confirmation signal transmitted back by the submersible; If no acknowledgment signal is received before the retransmission timer expires, the precoded signal is retransmitted and the number of retransmissions is accumulated. When the cumulative number of retransmissions exceeds the preset retransmission upper limit threshold, the current channel model is determined to be invalid, and the current data transmission task to be sent is terminated. The control target receives the submersible and rebroadcasts broadband acoustic detection signals to one or more transmitting submersibles to initiate a new round of parameterized channel modeling.

[0065] In this embodiment, after the transmitting submersible completes a coordinated transmission of a precoded signal, it does not immediately send the next set of data, but instead enters a waiting state with a defined time limit. Simultaneously, an internal retransmission timer is activated and begins counting down. The preset duration of this timer is not arbitrarily set, but is based on a precise estimate of the round-trip time of the current communication link. This duration is typically set to the estimated signal round-trip propagation delay, plus the processing time required for the target receiving submersible to decode, verify, and generate an acknowledgment signal, with an additional small safety margin to handle minor network jitter. During this period, the transmitting submersible continuously monitors the uplink from the target receiving submersible, waiting for a positive acknowledgment signal. This acknowledgment signal is usually a short data packet explicitly indicating that the previously transmitted data has been successfully received and verified.

[0066] If the transmitting submersible fails to receive a valid acknowledgment signal from the receiving submersible by the time the retransmission timer reaches zero, this timeout event triggers a pre-defined retransmission protocol. A timeout indicates a problem in the communication link. Possible causes include the transmission of the precoded signal being lost during propagation due to depth fading or strong noise interference, or the receiving submersible successfully receiving the data but losing its return acknowledgment signal on the way back. Regardless of the cause, the transmitting submersible retrieves the exact same precoded signal from its internal buffer and retransmits it. Simultaneously, an internal counter associated with this data transmission task increments, recording the number of retransmissions. After retransmission, the retransmission timer is reset to its initial duration and restarted, initiating a new round of acknowledgment waiting. This step proactively addresses momentary communication failures, overcoming occasional signal loss through retries, thereby improving the success rate of data transmission.

[0067] While retransmission mechanisms are effective in handling transient communication failures, they cannot continue indefinitely. To avoid the submersible getting stuck in an endless retransmission loop due to persistent link interruptions, thus wasting valuable energy and communication resources, the system sets a preset retransmission upper limit threshold. After each retransmission, the system compares the cumulative number of retransmissions with this upper limit threshold. If, after multiple attempts, such as five consecutive retransmissions, no acknowledgment signal is received, causing the cumulative number of retransmissions to exceed the preset upper limit, the system will no longer attribute this communication failure to accidental, transient interference. At this point, a higher-level judgment will be made: the persistent failure is most likely caused by a more fundamental problem, namely, that the current underwater channel model is severely mismatched and no longer accurately reflects the actual channel conditions. Based on this judgment, the system will immediately suspend the currently attempted data transmission task, mark the data packet as a transmission failure, and stop all related retransmission activities.

[0068] Once the current channel model is determined to be invalid and data transmission is suspended, the ultimate goal is to automatically restore the communication link, rather than abandoning it completely. This involves an automatic interruption and reconstruction process for the communication link. The first step in this process is for one or more transmitting submersibles to send a specific control command to the target receiving submersible. This command requests the target receiving submersible to restart the channel sounding procedure. Upon receiving this control command, the target receiving submersible switches its operating mode, transforming from a passive data receiver back into an active sound source. It will broadcast the preset broadband acoustic sounding signal with good autocorrelation characteristics to the surrounding waters, just as it did when the communication link was initially established. Simultaneously, the transmitting submersibles that received the reconstruction command also switch modes, ceasing data transmission and activating their respective receiving arrays to enter a state of channel parameter monitoring and acquisition.

[0069] The present invention also discloses an information transmission system for a submersible, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the information transmission method for a submersible as described in any of the above embodiments.

[0070] The processor can be a central processing unit (CPU). Of course, depending on the actual use, it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc., and this application does not limit it.

[0071] The memory can be an internal storage unit of a computer device, such as a hard disk or RAM, or an external storage device, such as a plug-in hard disk, smart memory card (SMC), secure digital card (SD), or flash memory card (FC) provided on the computer device. Furthermore, the memory can be a combination of internal storage units and external storage devices of a computer device. The memory is used to store computer programs and other programs and data required by the computer device. The memory can also be used to temporarily store data that has been output or will be output. This application does not limit this.

[0072] The present invention also discloses a computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the information transmission method for a submersible described in any of the above embodiments.

[0073] The computer program can be stored in a machine-readable medium. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or certain middleware. The machine-readable medium includes any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the machine-readable medium includes, but is not limited to, the above-mentioned components.

[0074] The information transmission method for a submersible described in the above embodiments is stored in the computer-readable storage medium and loaded and executed on a processor to facilitate the storage and application of the above method.

[0075] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of protection of this application is limited to these examples; within the framework of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of one or more embodiments of this application as described above, which are not provided in detail for the sake of brevity.

[0076] One or more embodiments in this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of this application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments in this application should be included within the protection scope of this application.

Claims

1. An information transmission method for a submersible, characterized in that, Applied to a submersible cluster comprising one or more target receiving submersibles and one or more transmitting submersibles, the method includes the following steps: The target receiving submersible broadcasts a preset broadband acoustic detection signal to one or more transmitting submersibles, and the one or more transmitting submersibles receive the detection signal that has become distorted after propagating through the complex underwater environment. The distorted detection signal is parametrically modeled by one or more transmitting submersibles. The distorted detection signal is decomposed into a multipath channel parameter set containing time delay, angle of arrival, amplitude and phase, and a channel model between the transmitting submersible and the target receiving submersible is constructed. When data needs to be transmitted from one or more transmitting submersibles to a target receiving submersible, a precoded signal is generated by the transmitting submersible based on the channel model. When different data to be transmitted is transmitted to multiple different target receiving submersibles at the same time, the data to be transmitted to different target receiving submersibles are loaded with preset acoustic orbital angular momentum phase factors of different modes, and a precoded signal superimposed with the acoustic orbital angular momentum phase factors is generated by combining the multipath channel parameter set. Based on high-precision clock synchronization, the precoded signals generated by each of one or more transmitting submersibles are transmitted in a coordinated manner. The target receiving submersible receives a signal that propagates through the complex underwater environment and is spatiotemporally focused at the target receiving submersible's location. Through adaptive blind equalization processing, the received spatiotemporally focused signal is sharpened according to the ideal impulse response characteristics that the spatiotemporally focused signal should have in order to eliminate residual inter-symbol interference, and the original data is demodulated and recovered.

2. The information transmission method for a submersible according to claim 1, characterized in that, The step of parametrically modeling the distorted detection signal using one or more transmitting submersibles, decomposing the distorted detection signal into a multipath channel parameter set including time delay, angle of arrival, amplitude, and phase, and constructing a channel model between the transmitting submersible and the target receiving submersible includes the following steps: The distorted detection signal is subjected to time-domain windowing processing; A high-resolution direction-of-arrival estimation algorithm is used to process the distorted detection signal after the time-domain windowing process, and multiple propagation path signals with different spatial angles are identified and separated. For each separated propagation path signal, the time delay and amplitude attenuation corresponding to each propagation path signal are calculated by performing cross-correlation operation with the broadband acoustic detection signal; Measure the phase shift of the signal along each propagation path relative to the broadband acoustic detection signal; The calculated delay, angle of arrival, amplitude attenuation, and phase offset of each propagation path are combined to form a structured multipath channel parameter set, and the channel model is constructed based on the multipath channel parameter set.

3. The information transmission method for a submersible according to claim 1, characterized in that, The process of loading preset acoustic orbital angular momentum phase factors of different modes onto the data to be transmitted to different target receiving submersibles, and generating a precoded signal superimposed with the acoustic orbital angular momentum phase factors in combination with the multipath channel parameter set includes the following steps: Based on the multipath channel parameter set, the channel model is time-reversed and phase-conjugated to generate a basic precoding filter for each target receiving submersible. Each basic precoding filter is assigned a unique and mutually orthogonal acoustic orbital angular momentum phase factor based on the number and identification of the target receiving submersible. The acoustic orbital angular momentum phase factor is applied to the corresponding basic precoding filter in the form of a complex product to generate a composite precoding filter. The data to be transmitted to different target receiving submersibles are convolved through the corresponding composite precoding filters to obtain multiple parallel encoded signal streams; The multiple parallel encoded signal streams are linearly superimposed to synthesize a precoded signal.

4. The information transmission method for a submersible according to claim 3, characterized in that, The method further includes a step of dynamically correcting the channel model after generating the precoded signal: Continuously monitor the communication quality confirmation information fed back by the submersible from the target; When the communication quality confirmation information indicates that the communication quality is lower than a preset threshold, the transmitting submersible transmits a low-power differential detection signal to the target receiving submersible. The system receives and processes the response of the low-power differential probe signal, and calculates the perturbation change of the current channel relative to the previously described channel model. The multipath channel parameter set is updated based on the perturbation change to obtain the corrected channel model. The base precoding filter and the acoustic orbital angular momentum phase factor are adjusted based on the modified channel model to generate the precoded signal adapted to the current channel state.

5. The information transmission method for a submersible according to claim 3, characterized in that, Before performing the adaptive blind equalization process, the target receiving submersible also includes a step of eliminating crosstalk between multiple signals: Configure the target receiving submersible with a parallel receiving channel that matches the number of acoustic orbital angular momentum phase factor modes; Using the orthogonal demodulation operators configured in each parallel receiving channel, the received spatiotemporally focused signal is subjected to preliminary mode separation to obtain a multi-channel parallel signal containing the main signal and residual crosstalk signal; Construct a multi-input multi-output adaptive interference cancellation matrix, using the multi-channel parallel signals as inputs to the adaptive interference cancellation matrix; The coefficients of the adaptive interference cancellation matrix are iteratively updated with the goal of minimizing the cross-correlation between the output channels of the adaptive interference cancellation matrix. Each main signal, after being processed by the adaptive interference cancellation matrix and with crosstalk suppressed, is sent to its respective adaptive blind equalization processing module.

6. The information transmission method for a submersible according to claim 1, characterized in that, The process of receiving a signal that propagates through the complex underwater environment and is spatiotemporally focused at the target receiving submersible's location, and then sharpening the received spatiotemporally focused signal by adaptive blind equalization based on the ideal impulse response characteristics of the spatiotemporally focused signal to eliminate residual inter-symbol interference, and demodulating to recover the original data includes the following steps: A digital adaptive transverse filter is installed inside the target receiving submersible, and a Dirac impulse function is preset as the ideal impulse response characteristic. The spatiotemporally focused signal is used as the input signal of the adaptive transverse filter; Calculate the mean square error between the actual output signal of the adaptive transverse filter and the ideal impulse response characteristics; Based on the mean square error, and using the least mean square error algorithm or the recursive least squares algorithm, the internal tap coefficients of the adaptive transverse filter are iteratively updated; The updated adaptive transverse filter is used to filter the subsequently received spatiotemporally focused signal, and the output signal, which has been sharpened and freed from residual inter-symbol interference, is used for demodulation to obtain the recovered original data.

7. The information transmission method for a submersible according to claim 1, characterized in that, The process of transmitting the precoded signals generated by one or more transmitting submersibles in coordination based on high-precision clock synchronization includes the following steps: In the submersible cluster, designate one master node submersible and treat the remaining submersibles as slave nodes. The master node submersible periodically broadcasts a synchronization beacon signal containing the current timestamp of the master node submersible; After receiving the synchronization beacon signal, the slave node submersible calculates the propagation delay of the synchronization beacon signal and calibrates the local clock of the slave node submersible based on the propagation delay and the received timestamp. The slave node submersible sends a response signal back to the master node submersible, enabling the master node submersible to calculate the round-trip time and determine an estimate of the clock skew of the slave node submersible. A future launch timestamp is set in the coordinated launch command, and when all the transmitting submersibles reach their respective launch timestamps on their calibrated local clocks, they synchronously launch the precoded signals they have generated.

8. The information transmission method for a submersible according to claim 7, characterized in that, The method also includes automatic interruption and reconstruction steps for the communication link: After the transmitting submersible transmits the precoded signal, a retransmission timer is started and the target waits for the submersible to send back an acknowledgment signal. If the acknowledgment signal is not received before the retransmission timer expires, the precoded signal is retransmitted and the number of retransmissions is accumulated. When the cumulative number of retransmissions exceeds the preset retransmission upper limit threshold, it is determined that the current channel model has failed, and the current data transmission task to be sent is terminated. The target receiving submersible is controlled to rebroadcast the broadband acoustic detection signal to one or more transmitting submersibles to initiate a new round of the parameterized channel modeling process.

9. An information transmission system for a submersible, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the information transmission method for a submersible as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When executed by a processor, the instruction causes the processor to be configured to perform the information transmission method for a submersible according to any one of claims 1 to 8.

Images