Offshore long-distance data transmission method and system based on marine laser communication

Abstract

The application relates to the technical field of communication transmission, and discloses a near-sea long-distance data transmission method and system based on marine laser communication, which comprises the following steps: emitting a laser beam carrying original data; emitting an ultrasonic wave to a water body in front of a receiving end to construct an ultrasonic wave field, causing periodic changes in refractive index, so that amplitude modulation of a preset acoustic frequency is superimposed on a speckle light field formed by multipath scattering of the laser beam; monitoring a receiving light field by using a dynamic visual sensor array, outputting a discrete event when an absolute value of a logarithmic light intensity change quantity of a pixel reaches a trigger threshold, and generating an asynchronous event stream; performing locking filtering on the asynchronous event stream based on the preset acoustic frequency, and extracting an effective signal event subset; and decoding the effective signal event subset to restore the original data. The application converts disordered scattering noise into a deterministic same-frequency envelope, breaks through frame rate limitation, filters out background light interference, and realizes high-robustness data transmission under extremely low signal-to-noise ratio and turbid water quality.

Description

Technical Field

[0001] This invention relates to the field of communication transmission technology, specifically to a method and system for near-shore long-distance data transmission based on marine laser communication. Background Technology

[0002] Underwater laser communication boasts high transmission rates and is used in marine environmental monitoring and underwater target interaction. Existing underwater laser communication systems mainly employ direct energy detection or coherent detection mechanisms, with the receiver typically relying on photomultiplier tubes, avalanche photodiodes, or conventional frame-driven area array image sensors for signal acquisition.

[0003] In nearshore and other high-turbidity water environments, the presence of a large number of suspended particles causes multipath scattering of the laser beam during transmission. After transmission through the channel, the light field reaching the receiver loses its original spatial coherence and beam collimation, degenerating into a spatially randomly distributed speckle light field with fluctuating intensity.

[0004] Faced with such complex physical channels, existing receiving mechanisms have significant limitations. When using a single-point photodetector to receive a speckle light field, photons arriving through different scattering paths cancel each other out, resulting in attenuation of the effective signal energy. If a conventional area array image sensor is used, its operating principle is based on fixed-time integral exposure. Under the interference of complex underwater background light and backscattered light, pixels are prone to charge saturation, causing the communication signal to be submerged in environmental noise. Furthermore, the time dispersion caused by multipath scattering can cause light pulse broadening, resulting in inter-symbol interference between adjacent communication symbols.

[0005] Current systems largely rely on direct modulation and demodulation of light source intensity, lacking physical layer intervention methods for random speckle channels. Under the combined influence of background noise and multipath effects, existing technologies struggle to extract effective data from the degraded received optical field, leading to a decrease in the signal-to-noise ratio, limited transmission distance, and susceptibility to link interruptions in near-shore environments. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a near-shore long-distance data transmission method and system based on marine laser communication, solving the technical problems of reduced signal-to-noise ratio, limited communication distance, and easy link interruption caused by severe multipath scattering and background light interference in turbid near-shore waters in existing underwater laser communication systems.

[0007] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of the present invention provides a near-shore long-distance data transmission method based on marine laser communication, comprising: Emit a laser beam carrying the raw data; An ultrasonic wave is emitted towards the water area in front of the receiving end to construct an ultrasonic field, causing a periodic change in the refractive index of the water medium in that area. This results in the superposition of a preset acoustic frequency amplitude modulation onto the speckle light field formed by the multipath scattering of the laser beam by the near-sea water to form a receiving light field. The received light field is monitored using a dynamic visual sensor array. When the absolute value of the logarithmic change in light intensity of a pixel reaches a preset trigger threshold, the corresponding pixel outputs a discrete event, generating an asynchronous event stream. The asynchronous event stream is subjected to locking filtering based on the preset acoustic frequency to extract a subset of valid signal events; The subset of valid signal events is decoded to restore the original data.

[0008] Furthermore, the laser beam carrying the raw data includes: The original data in binary form is mapped to a symbol set, which is defined by the spatial coordinates of the activated transmitting unit and the light pulse trigger timestamp. A two-dimensional vertical cavity surface-emitting laser array is driven to output optical pulses according to the symbol set to obtain the laser beam.

[0009] Furthermore, the step of emitting ultrasonic waves towards the water area in front of the receiving end to construct an ultrasonic field includes: The ultrasonic waves are emitted using a ring-shaped ultrasonic transducer placed at the front end of the receiver to construct an ultrasonic standing wave. The refractive index of the water medium is fluctuated by the radiation pressure of the ultrasonic standing wave, thereby constructing a dynamic refractive index field with the preset acoustic frequency, and the speckle light field is superimposed with the periodic amplitude modulation.

[0010] Furthermore, the step of monitoring the received light field using a dynamic visual sensor array and generating an asynchronous event stream includes: The pixels in the dynamic vision sensor array independently monitor the logarithm of the incident light intensity. When the absolute value of the logarithmic change in light intensity between the current moment and the previous trigger moment reaches the preset trigger threshold, the following triggering model is satisfied:

[0011] In the formula, For the current moment Falling on coordinates Instantaneous light intensity sensed on pixels, The previous trigger time, A preset trigger threshold is set; when the above conditions are met, the corresponding pixel outputs the discrete event containing spatial coordinates, timestamp, and polarity and updates the trigger time; all discrete events generated within the receiving period constitute the asynchronous event stream.

[0012] Further, the step of performing lock-in filtering on the asynchronous event stream based on the preset acoustic frequency to extract a subset of valid signal events includes: The density of the asynchronous event stream on the time axis is statistically analyzed to obtain the event trigger rate function; Perform time-frequency analysis on the event trigger rate function to extract the spectral peaks in the frequency domain; The effective signal event subset is extracted by using bandpass filtering to remove noise by utilizing the significant spectral peak features at the preset acoustic frequency.

[0013] Further, decoding the subset of valid signal events to reconstruct the original data includes: Three-dimensional spatiotemporal clustering is performed on the subset of effective signal events to construct the spatiotemporal feature vector of the receiver; Calculate the Euclidean distance between the spatiotemporal feature vector of the receiving end and the feature vectors of each ideal symbol template in the preset mapping dictionary:

[0014] In the formula, The spatiotemporal feature vector of the receiving end and the first Euclidean distance between dictionary templates The total number of elements contained in the feature vector. , , Representing the The first dictionary template Standard three-dimensional parameters of each element , , Represents the first in the spatiotemporal feature vector of the receiving end The three-dimensional parameters of each element, The time dimension is used as the dimensional weighting coefficient; the template index with the minimum Euclidean distance is output as the decision symbol, and the original data is restored according to the inverse mapping rule.

[0015] A second aspect of the present invention provides a near-shore long-distance data transmission system based on marine laser communication, comprising: Laser emitting equipment used to emit laser beams carrying raw data into bodies of water; The acoustic and optical conditioning device is placed at the receiving end and is used to emit ultrasonic waves into the water area in front of the receiving end to create an ultrasonic field. An event-driven receiving device includes a dynamic visual sensor array for monitoring the received light field and outputting discrete events when the absolute value of the pixel logarithmic light intensity change reaches a preset trigger threshold, generating an asynchronous event stream; The processor and the memory storing the computer program are respectively connected in communication with the laser emitting device, the acousto-optic conditioning device and the event-driven receiving device; When the processor executes the computer program, it controls the laser emitting device and the acousto-optic conditioning device to work together, performs locking filtering based on a preset acoustic frequency on the asynchronous event stream generated by the event-driven receiving device to extract a subset of valid signal events, and decodes the subset of valid signal events to restore the original data.

[0016] Furthermore, the laser emitting device includes a two-dimensional vertical cavity surface-emitting laser array; the processor is specifically used to: map the raw data in binary form into a symbol set defined by spatial coordinates and timestamps, and drive the two-dimensional vertical cavity surface-emitting laser array to output light pulses according to the symbol set to obtain the laser beam.

[0017] Furthermore, the acoustic-optical conditioning device includes a ring-shaped ultrasonic transducer; the processor is specifically used to: control the ring-shaped ultrasonic transducer to emit ultrasonic waves to construct ultrasonic standing waves, thereby constructing a dynamic refractive index field with a preset acoustic frequency in the water body.

[0018] Furthermore, the memory is pre-configured with a mapping dictionary; the processor is specifically used to: acquire the asynchronous event stream generated by the event-driven receiving device; extract a subset of valid signal events by performing lock-in filtering on the asynchronous event stream based on the preset acoustic frequency; perform spatiotemporal clustering on the subset of valid signal events to obtain the spatiotemporal feature vector of the receiving end, and reconstruct the original data based on the minimum Euclidean distance matching decision symbol with the feature vector in the mapping dictionary.

[0019] This invention provides a method and system for near-shore long-distance data transmission based on marine laser communication. It has the following beneficial effects: 1. To address the problem of severe optical signal distortion and communication interruption caused by near-shore multipath scattering, this invention transmits ultrasonic waves into the water ahead of the receiver to construct a dynamic refractive index field, forcibly superimposing an amplitude modulation of a preset acoustic frequency onto the unpredictable random speckle optical field. This physical-level active channel modification mechanism transforms disordered ocean scattering noise into a deterministic, co-frequency envelope that can be precisely tracked by the system, thereby effectively overcoming the optical transmission bottleneck in turbid water.

[0020] 2. To address the technical limitation of traditional photodetectors being easily overwhelmed by strong background light during integral exposure, this invention utilizes a dynamic visual sensor array for pixel-level asynchronous monitoring, outputting discrete events only when the logarithmic change in light intensity reaches a threshold. This asynchronous sensing architecture avoids the limitations of a fixed frame rate and, with its microsecond-level capture capability and wide dynamic range for high-frequency transient light signals, effectively filters out the slow drift interference of underwater ambient light, achieving robust signal sensing under extremely low contrast conditions.

[0021] 3. To address the challenge of signal analysis under extremely low signal-to-noise ratios, this invention performs locked filtering based on a preset acoustic frequency on the generated asynchronous event stream, selectively extracting a subset of valid signal events to reconstruct the data. This processing logic utilizes the frequency labels assigned by the prior acoustic-optical coupling to accurately remove broadband background noise generated by random channel disturbances, re-establishing the photon spatiotemporal correlation disrupted by scattering in the extremely harsh underwater multipath channel, ensuring reliable reconstruction of long-distance data transmission. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the physical hardware topology of the system of the present invention; Figure 3 This is a schematic diagram of the multidimensional spatiotemporal joint modulation principle of the transmitter of the present invention; Figure 4 This is a schematic diagram of the acoustic-optical coupling physical mechanism and sensing of the receiving end of the present invention; Figure 5 This is a flowchart of the locking filter algorithm based on acoustic frequency according to the present invention. Figure 6 This is a flowchart of the three-dimensional spatiotemporal density clustering and dictionary decoding and restoration process of the present invention. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see the appendix Figure 1 - Appendix Figure 6 This invention provides a method and system for near-shore long-distance data transmission based on marine laser communication. The method for near-shore long-distance data transmission based on marine laser communication may include the following steps: Emit a laser beam carrying the raw data; An ultrasonic wave is emitted towards the water area in front of the receiving end to construct an ultrasonic field, causing a periodic change in the refractive index of the water medium in that area. This results in the superposition of a preset acoustic frequency amplitude modulation onto the speckle light field formed by the multipath scattering of the laser beam by the near-sea water to form a receiving light field. The received light field is monitored using a dynamic visual sensor array. When the absolute value of the logarithmic change in light intensity of a pixel reaches a preset trigger threshold, the corresponding pixel outputs a discrete event, generating an asynchronous event stream. The asynchronous event stream is subjected to locking filtering based on the preset acoustic frequency to extract a subset of valid signal events; The subset of valid signal events is decoded to restore the original data.

[0025] This invention provides a near-shore long-distance data transmission system based on marine laser communication and a corresponding implementation method. The system is deployed in near-shore underwater environments with high turbidity and strong suspended particulate matter, and is used to solve the problem of sharp deterioration in signal-to-noise ratio and communication interruption caused by severe multipath scattering in traditional coherent optical communication or direct energy detection mechanisms.

[0026] The entire data transmission system is physically distributed across the transmitting and receiving ends of the communication link. The transmitting end is equipped with a laser emitting device, while the receiving end integrates an acousto-optic conditioning device, an event-driven receiving device, a processor, and a memory. To ensure nanosecond to microsecond-level timestamp accuracy and acousto-optic synchronization, the processor is preferably a field-programmable gate array (FPGA) or a digital signal processor (DSP). The memory not only non-volatilely stores the mapping dictionary described below, but also, when executed by the processor, automatically controls the laser emitting device, the acousto-optic conditioning device, and the event-driven receiving device to work collaboratively according to a preset timing sequence, thereby implementing the various steps of the data transmission method described in this invention.

[0027] The aforementioned devices are physically connected via a high-speed differential bus and share the same global high-precision hardware clock reference source. To eliminate system-level physical delays caused by the difference between the underwater speed of sound (approximately 1500 m / s) and the speed of light, a phase delay compensation module is configured inside the processor. After the system powers on, a calibration process is executed: the processor controls the acoustic-optical conditioning device to emit a single-pulse ultrasound, records the emission trigger timestamp, and then uses an event-driven receiving device to detect the light intensity disturbance generated by ultrasonic cavitation and record the reception timestamp. The difference between the two is the sound wave propagation delay. During normal operation, the processor advances the time axis of the ultrasonic drive signal. Compensation is performed to ensure that the optical modulation envelope arriving at the sensor target surface is aligned with the frequency reference of the event timestamp, so that it is uniformly scheduled by the processor to execute the transmission task.

[0028] In the complex physical channels of nearshore waters, the presence of abundant plankton and inorganic suspended matter results in extremely turbid water. When the coherent laser beam carrying the original data emitted by the laser transmitting device penetrates this water channel, severe multipath scattering occurs. At this point, the optical field reaching the receiver loses its original spatial coherence and beam collimation, degenerating into a randomly distributed speckle field.

[0029] To accurately define the impact of this physical process on the subsequent receiving mechanism, the following optical field evolution model is established. The complex amplitude of the initial incident light field output by the laser emitting device is set as... ,in It is a three-dimensional spatial coordinate vector. The time variable is used. The nearshore water channel is equivalently described as a time-varying multipath scattering impulse response function with extremely high spatial variability. .

[0030] After long-distance transmission, the complex amplitude of the speckle optical field reaches the water front region of the receiving end. Expressed as a convolution of the incident light field and the water body's impulse response:

[0031] In the formula, This represents the time delay parameter caused by different scattering paths due to multipath propagation.

[0032] Correspondingly, the spatial distribution of light intensity generated by this speckle light field on the cross-section of the receiver. It is expressed as the square of the complex amplitude mode:

[0033] The above formula shows that the light intensity reaching the receiver... It exhibits extremely uneven speckle characteristics that fluctuate wildly and randomly over time. When conventional photodetector arrays receive this type of light intensity, the energy received by each pixel unit will cancel each other out in phase or the signal-to-noise ratio will be overwhelmed by the speckle noise.

[0034] In order to reduce the extremely degraded speckle intensity This invention innovatively deploys an acousto-optic conditioning device in front of the optical field of view of the event-driven receiving device, re-extracting high-dimensional features carrying the original data. This device directly acts on the local water medium in front of the receiver, rather than on the light beam itself.

[0035] Specifically, the acoustic-optical conditioning device continuously injects ultrasonic waves into the localized water body to create an ultrasonic field. This physical action transforms the originally static or slowly flowing localized water medium into an acoustically driven dynamic optical modulator. By forcibly altering the spatial distribution of water density through ultrasonic radiation pressure, the water's refractive index undergoes high-frequency periodic changes.

[0036] When the extremely scattered speckle light field passes through this localized water body dynamically modified by the ultrasonic field, it will change the original random light intensity. Based on this, a global amplitude modulation envelope with a defined physical frequency (i.e., a preset acoustic frequency) is forcibly superimposed. This process transforms unpredictable near-shore scattering noise into a deterministic modulation signal that the receiving device can actively lock onto and track, thereby generating the received light field ultimately projected onto the target surface of the event-driven receiving device. The event-driven receiving device, in conjunction with a processor, utilizes a dedicated asynchronous event triggering mechanism and filtering algorithm to complete the data restoration and reconstruction from this received light field.

[0037] In this embodiment, to ensure robust data transmission in the extremely harsh ocean multipath scattering channels, the present invention employs a specific multidimensional spatiotemporal joint design for the signal loading mechanism at the transmitter. This design abandons the conventional methods of relying on a single light source intensity or phase modulation in traditional marine optical communication, and extends the data carrying dimension to a combined domain of spatial coordinates and time microsecond-level slices.

[0038] In the signal processing layer of the transmitting device, the processor first receives the raw data stream in binary form from the external input. This continuous binary data stream is divided into fixed-length data blocks by serial-to-parallel conversion logic. Subsequently, each independent binary data block is strictly mapped to a specific symbol according to a preset mapping rule.

[0039] All of the above independent symbols together constitute a complete symbol set. .in, The total number of symbols available in the symbol set is represented by the bit width of a single data block.

[0040] In the technical solution of this invention, the symbol set Any specific symbol in It is not a single level amplitude value, but rather a set of spatial coordinate vectors of the activated transmitting units. and the corresponding optical pulse trigger timestamp sequence Joint definition. This definition method can be represented by a specific structure tuple as follows: .

[0041] In order to physically determine the aforementioned spatial coordinate vectors For instantiated output, the laser emitting device of this invention employs a two-dimensional vertical cavity surface-emitting laser array. This array consists of densely arranged independent microcavity laser emitting units, forming a regular two-dimensional planar grid.

[0042] For symbols Its spatial coordinate vector is expressed as:

[0043] In the formula, Precisely indicating the number of times the specific symbol was sent on the array panel. The two-dimensional physical coordinates of a microcavity laser unit to be lit and activated, parameters The total number of spatial luminescent units invoked by this symbol.

[0044] Meanwhile, the timestamp sequence corresponding to this symbol is defined as In the formula, The coordinates are defined as The specific time points at which the light-emitting units output extremely narrow light pulses are determined. By strictly corresponding the spatial location with the time points, the initial spatiotemporal relative isolation of photons at the channel entrance is artificially constructed.

[0045] After completing the above mapping, the processor directly drives the two-dimensional vertical-cavity surface-emitting laser array. The hardware driver circuit follows the symbol... The definition is used to generate and inject a time- and space-varying two-dimensional driving current matrix into the array. Its physical control model satisfies the following formula:

[0046] In this control model, The rated threshold bias current amplitude for a single laser unit; This represents the two-dimensional Dirac function, used to precisely lock and filter the target space coordinates to be activated from the entire continuous array plane; This represents the preset single-cycle optical pulse time envelope waveform function.

[0047] Two-dimensional vertical cavity surface-emitting laser array in driving current matrix Under strict pumping, the emitting units at different spatial coordinates sequentially or in parallel excite discrete light pulses at their corresponding timestamps. These high-speed light pulse groups, rich in spatiotemporal specific patterns, are coupled by a front-end collimating lens group and converge in free space, ultimately obtaining a laser beam carrying the original data and possessing extremely strong anti-scattering properties. This laser beam is then injected into the complex near-shore seawater medium and evolves into the speckle light field described in subsequent steps.

[0048] In this embodiment, to overcome the physical limitations of traditional photoelectric receiving mechanisms in multipath scattering environments, the present invention deploys an acousto-optic conditioning device composed of a ring ultrasonic transducer in the leading edge region of the optical path at the receiving end. The active intervention of this device forcibly injects deterministic physical characteristics into the disordered ocean speckle channel, fundamentally changing the signal extraction logic.

[0049] During system operation, the acoustic-optical conditioning device, precisely controlled by an internal processor, continuously emits ultrasonic waves into a localized water body in front of the event-driven receiving device's field of view. In its specific spatial deployment, the annular ultrasonic transducer is preferably coaxially mounted outside the lens of the event-driven receiving device, with the sound field focusing point located within a water body range of 0.1 to 2.0 meters in front of the sensor. The ultrasonic waves radiated through this specially shaped geometry interfere and converge within a specific spatial region, thereby constructing a stable and intense ultrasonic standing wave field underwater. This standing wave field generates spatially fixed antinodes and nodes in the water medium.

[0050] It should be noted that although this embodiment uses an ultrasonic standing wave field as the optimal implementation, in other alternative embodiments, the ultrasonic field can also be constructed by emitting unidirectional ultrasonic traveling waves or high-intensity focused ultrasound: When using a traveling wave scheme, a unidirectional ultrasonic beam is emitted through a parallel ultrasonic transducer array, forming a periodic refractive index grating moving along the propagation direction in the water body. The modulation effect is equivalent to the standing wave scheme, only requiring the center frequency of the bandpass filter to be locked to the traveling wave ultrasonic frequency. When using a focused ultrasound scheme, a spherical focusing transducer concentrates the sound field into a smaller water area, achieving a higher modulation depth, suitable for high-turbidity extreme scenarios. All methods that can cause a periodic change in the refractive index of the local water body do not depart from the scope of protection of this invention.

[0051] Let the preset acoustic frequency of the driving ring ultrasonic transducer be . The corresponding acoustic angular frequency is The dynamic sound pressure distribution model established by this ultrasonic standing wave field in a local water body. Satisfy the following mathematical relationship:

[0052] In this sound pressure physical model, This indicates the peak amplitude of the standing wave sound pressure. The spatial wave vector of the preset ultrasonic wave, Represents a three-dimensional spatial coordinate vector in the water body. It is a time variable.

[0053] In specific engineering implementation environments, to ensure that the signal can effectively penetrate near-shore waters and generate a significant acousto-optic coupling effect, the laser emitting device preferably employs a two-dimensional vertical-cavity surface-emitting laser array with wavelengths in the 450nm to 550nm attenuation window in the blue-green band; the operating frequency (i.e., the preset acoustic frequency) of the ring ultrasonic transducer in the acousto-optic conditioning device... The preferred setting is within the 50kHz to 2MHz range, and its transmission power must ensure that a peak amplitude is formed at the focal point of the water body. An ultrasonic standing wave of not less than 0.1 MPa; the event-driven receiving device preferably employs a dynamic visual sensor with microsecond-level time resolution (e.g., 1 μs), and the preset trigger threshold... The value is dynamically set based on the ambient light intensity of the scene, and is usually between 0.1 and 0.3 natural logarithmic units.

[0054] Furthermore, in terms of overall system timing configuration, to avoid aliasing between modulation frequency and symbol rate, it is necessary to ensure that the transmission period of a single symbol contains at least 20 ultrasonic modulation cycles. This ensures that the subsequent frequency-locked filtering algorithm has sufficient sampling points to complete spectral peak identification. Under typical parameters, when the ultrasonic frequency emitted by the acousto-optic conditioning device is set to 1MHz, the symbol transmission rate corresponding to the laser emitting device is preferably configured to be between 10kBaud and 50kBaud.

[0055] Under the direct radiation pressure of the aforementioned strong sound pressure gradient distribution, the fluid particles within the water medium undergo periodic microscopic aggregation and dispersion. Based on the cross-mechanism of fluid mechanics and acousto-optics, the dramatic fluctuations in sound pressure trigger both spatial and temporal fluctuations in water density, thereby forcibly causing a synchronous change in the optical refractive index of the water medium in that region.

[0056] This local dynamic refractive index field induced by ultrasonic standing waves The physical evolution law is expressed as:

[0057] In the formula, The background resting refractive index of near-shore water when it is not disturbed by a sound field. The pressure-optic coefficient characterizes the specific water medium. This refractive index field formula shows that the ultrasonic standing wave field actually temporarily transforms the originally homogeneous or slowly varying frontal water body into a three-dimensional acousto-optic phase grating with high-frequency oscillation characteristics.

[0058] When a speckle light field carrying the original data but already subject to severe multipath scattering interference penetrates the dynamic refractive index field, significant nonlinear acousto-optic coupling occurs between the two. This is due to the long ultrasonic wave length of the dynamic refractive index field. Acousto-optic interaction length With laser wavelength The combination satisfies the Klein-Cook parameters The physical boundary conditions are such that it conforms to the Raman-Nass diffraction mechanism. The cumulative optical phase delay of the high-frequency scintillation speckle field during its travel is periodically modulated.

[0059] After the aforementioned physical process of acoustic-optical interweaving, the originally chaotic speckle light intensity, upon reaching the sensing target surface of the event-driven receiving device, is forcibly superimposed with a global amplitude modulation at the same frequency as the preset acoustic frequency. The final light intensity distribution of the received light field projected onto the sensor array... Expressed as:

[0060] In this modulation formula, This represents the background speckle intensity formed by multipath scattering from near-shore seawater before acoustic field modulation; This represents the acousto-optic amplitude modulation depth coefficient induced by ultrasonic radiation pressure, the magnitude of which depends on the peak amplitude of the applied ultrasonic sound pressure. In typical applications in near-shore waters, the modulation depth is controlled within the range of 5% to 30% by adjusting the ultrasonic power. The amplitude variation within this range can stably trigger the logarithmic threshold circuit of the dynamic vision sensor, ensuring effective detection of the audio tag. This refers to the fixed system phase offset that accompanies the acousto-optic coupling penetration process. Through this acousto-optic conditioning step, the present invention superimposes a deterministic audio envelope resistant to multipath interference onto a random speckle background, laying a solid physical channel foundation for subsequent noise stripping and event locking extraction at the pixel level.

[0061] In this embodiment, to address the received light field containing weak modulation envelopes and strong background scattering noise formed after acousto-optic conditioning, the present invention abandons the frame-level integration exposure mechanism of conventional photoelectric sensors in the event-driven receiving device, and innovatively adopts a dynamic visual sensor array for pixel-level asynchronous monitoring. This array has a microsecond-level instantaneous response capability to extremely high frequency light intensity transients.

[0062] In the hardware architecture of the dynamic vision sensor array, each photosensitive pixel unit on the two-dimensional pixel target surface is equipped with an independent logarithmic photoelectric receiving circuit and differential comparator logic. Each pixel is spatially and temporally independent, independently monitoring the instantaneous evolution of the incident light intensity of the received light field projected onto its own photosensitive area without interference.

[0063] For the coordinates in the array For any independent pixel, its internal logarithmic amplifier circuit will physically sense the continuous incident light intensity at that point. It is instantly converted into a corresponding logarithmic voltage signal, thereby greatly expanding the sensor's ability to adapt to high dynamic range under complex underwater lighting conditions.

[0064] The differential circuit inside this pixel continuously calculates the current time. The logarithmic light intensity voltage value and the previous trigger time stored in the pixel. The deviation between the logarithmic light intensity voltage value. When this deviation, that is, the absolute value of the logarithmic light intensity change, first reaches and crosses the preset trigger threshold set by the system, the differential comparator flips, and the pixel is immediately activated.

[0065] The pixel activation process follows a rigorous mathematical triggering model:

[0066] In the above trigger determination formula, Represents the natural logarithm operation; The coordinates at the current time Instantaneous light intensity sensed on a pixel; This is the latched light intensity value of the pixel during the most recent successful output event; This indicates the preset trigger threshold set by the hardware, and its value determines the system's ability to suppress background ambient light gradient noise.

[0067] Once the above triggering conditions are met, the coordinate is The pixel unit no longer outputs analog light intensity grayscale values, but immediately sends a digital discrete event to the system bus. After this physical action is completed, the internal circuitry of the pixel quickly sets the current time... Write over the memory register and update it to the new reference time. This completes the reset loop for a single pixel cycle.

[0068] Output discrete events In terms of data structure, it is encapsulated as a structure containing multidimensional attributes:

[0069] In the data packet body, The pixel space physical coordinates that triggered the event; With a hardware timestamp accuracy ranging from nanoseconds to microseconds, it precisely records the absolute time of the light intensity change. This represents the polarity of the light intensity change, where +1 indicates that the light intensity is at the rising edge of an abrupt change. 1 indicates that the light intensity is at a sudden decrease edge.

[0070] During a continuous receiving period Within the dynamic visual sensor array target surface, tens of thousands of independent pixels, driven by the combined effects of high-frequency speckle scintillation of the received light field and the acousto-optic modulation envelope, frequently and randomly satisfy the aforementioned triggering conditions. This results in countless discrete events output by all activated pixels. According to timestamp The order of events is aggregated and output via the asynchronous event routing bus.

[0071] These continuous and highly sparse data packets are concatenated on the timeline, ultimately generating and outputting an asynchronous event stream with extremely high time resolution. The asynchronous event stream In mathematical sets, this is expressed as:

[0072] In the formula, Represents the entire reception period The total number of discrete events captured by the internal system. This asynchronous event stream, in an extremely concise data format, filters out a large amount of spatial static redundant information while fully preserving all high-frequency transient characteristics in the received optical field caused by scattering channels and acoustic field modulation, providing a high-quality raw data source for frequency domain locking and signal stripping in subsequent algorithm layers.

[0073] In this embodiment, because the emitted laser propagates through extremely harsh multipath scattering water, the asynchronous event stream output by the dynamic visual sensor array is filled with a massive amount of disordered background noise events caused by random water disturbances, suspended debris obstruction, and sensor dark current. To extract the signal that truly carries the original data from this high-density noise event stream, this invention deploys a lock-in filtering algorithm for a preset acoustic frequency in the processor.

[0074] The processor first reads the asynchronous event stream returned by the receiving device. Furthermore, it eliminates the spatial discreteness of events on the macroscopic time axis, focusing on the temporal density characteristics of event triggering. The algorithm smoothly reconstructs the discrete timestamp distribution by applying a small time sliding window statistical mechanism in the continuous time domain.

[0075] By introducing a continuous kernel density estimation method, discrete event time series are transformed into continuous signals characterizing the macroscopic activity level of the array, i.e., the event trigger rate function. The statistical model for this function is as follows:

[0076] In this nuclear density estimation formula, For the first asynchronous event stream The precise timestamp of each event; This represents the total capacity of the event stream; For the selected smoothing kernel function, in specific implementations, a Gaussian window function is typically used. Its time variance parameter is preferably set to 1 to 5 ultrasonic modulation cycles. This setting ensures both the continuity and smoothness of the trigger rate curve while retaining sufficient frequency resolution to accurately lock the preset acoustic frequency. Physically, it reflects the probability density of all array pixels within the field of view being collectively awakened by a sudden change in light intensity at any given moment.

[0077] Given that the received light field has been forcibly superimposed with a preset acoustic frequency by the acousto-optic conditioning device in the previous steps The global amplitude modulation envelope, which is the fundamental frequency, causes the macroscopic frequency of events generated by the dynamic visual sensor array to resonate strongly with this modulation envelope. Therefore, the algorithm then applies the event trigger rate function in the one-dimensional time domain. Perform high-resolution time-frequency analysis.

[0078] In practice, the trigger rate signal is mapped from the time domain to the frequency domain using a Fast Fourier Transform, and the corresponding event trigger spectral density matrix is ​​calculated. :

[0079] In the formula, The transformed frequency variable, The imaginary unit is used when obtaining the trigger spectral density matrix. Subsequently, the processing algorithm directly anchors the ultrasonic frequency previously physically set by the acoustic-optical conditioning device, i.e., the preset acoustic frequency, on the horizontal axis of the frequency domain. .

[0080] The spectrum near that specific physical frequency band It will exhibit extremely prominent spectral peaks due to the resonance of sound and light, while the remaining event energy caused by the random scattering channel of the ocean and the slow drift of the background light is extremely disordered and low-frequency scattered in the low-frequency broadband of the entire spectral domain.

[0081] Using the found location Based on the significant spectral peak center and its envelope width parameters, the processor constructs 4th to 8th order Butterworth digital bandpass filters. The center frequency of this filter is strictly locked at [value missing]. Its 3dB passband bandwidth is dynamically set to 1.2 to 1.5 times the full width at half maximum (FWHM) of the spectral peak, thereby filtering out out-of-band noise to the maximum extent while preserving the effective modulated signal.

[0082] Then, the entire asynchronous event stream The system imports this lock-in digital filter. In its execution logic, the system first filters the continuous event trigger rate function to extract a clean, same-frequency envelope signal. This envelope signal is then used as a time-domain mask threshold. Only discrete events whose trigger timestamps fall within the effective range of the mask threshold peak are retained, while those falling outside the range or deviating from the preset acoustic frequency are excluded. Isolated noise events exhibiting evolutionary patterns were all identified as invalid triggers caused by multipath scattering and were discarded. The remaining discrete events were then reassembled, successfully extracting a subset of effective signal events with high signal-to-noise ratios.

[0083] In addition to the aforementioned short-time Fourier transform combined with bandpass filtering, a digital phase-locked loop (PLL) algorithm can also be used to achieve frequency locking filtering: using a preset acoustic frequency as a reference signal, coherent demodulation of the event trigger rate function is performed, and the amplitudes of the in-phase and quadrature components are extracted to determine whether the event is a valid signal; this method has stronger noise resistance and is suitable for extreme near-shore scenarios with more complex background noise. Wavelet transform algorithms based on specific acoustic frequencies can also achieve equivalent frequency locking effects: constructing a complex wavelet basis with the preset acoustic frequency as the center frequency, performing a wavelet transform on the event trigger rate function, and extracting the amplitudes of the wavelet coefficients at the corresponding scale as the basis for determining a valid signal; this method has stronger time-frequency localization capabilities and is suitable for dynamic near-shore scenarios with rapidly changing channels. All of the above implementations fall within the scope of protection of this invention.

[0084] In this embodiment, after bandpass-locked filtering based on a preset acoustic frequency, most of the random background noise caused by ocean multipath scattering in the system has been effectively removed. The remaining subset of effective signal events appears microscopically as a discrete point cloud with high physical correlation, which internally contains the initial symbolic features imparted by the transmitter through spatiotemporal joint modulation. The final step in the system processing architecture is to perform morphological analysis on this high-dimensional subset to restore the original binary data.

[0085] The processor first places a subset of valid signal events into a three-dimensional spatiotemporal coordinate system orthogonally formed by a two-dimensional physical space coordinate plane and a one-dimensional absolute time axis. Subsequently, the algorithm initiates a three-dimensional spatiotemporal density clustering mechanism to perform connectivity analysis and density scanning on all discrete event points within this space.

[0086] In the specific implementation of the algorithm, the three-dimensional spatiotemporal density clustering mechanism preferably adopts a density-based spatial clustering application with noise or its three-dimensional spatiotemporal extension variant. During the clustering process, the algorithm uses a set pixel spatial neighborhood radius (e.g., 2-5 pixels) and a microsecond-level temporal coherence window as core search neighborhood parameters, and statistically analyzes the density of discrete events within this three-dimensional search neighborhood. The system presets a minimum containment point threshold for the core object (preferably set to 10 to 50 events / cluster). When the density reaches the minimum containment point requirement, it is automatically determined that a spatiotemporally close event group has been found, and these discrete events clustered due to the flicker of the same light pulse are merged into an independent feature cluster; while detached events that do not reach the threshold are discarded as outlier noise. For each formed feature cluster, the system calculates its three-dimensional geometric centroid and accurately extracts the central spatial coordinates and average trigger timestamp that characterize the macroscopic physical properties of the cluster.

[0087] Assuming that within the current receive decoding cycle, the algorithm extracts a total of The processor serializes and concatenates the centroid parameters of these feature clusters in ascending order of timestamps, thereby constructing a spatiotemporal feature vector that can completely describe the current signal morphology at the receiving end. The system uses the transmission period of a single symbol as a decoding period, and the number of feature clusters obtained by clustering within a single period is... Theoretically, the number of activated light-emitting units should correspond to a single symbol at the transmitting end. One-to-one correspondence. If the number of clusters obtained by clustering exceeds... If the error exceeds the specified range, the current period is determined to be affected by strong noise interference. Data from that period is discarded, and inter-frame error correction is performed using a pre-set forward error correction code (such as a low-density parity-check code or a Reed-Solomon code) combined with valid data from adjacent periods to ensure data stream continuity. If the error is within the specified range, the data is sorted in ascending order by centroid timestamp, and the previous data is truncated. The main clusters are concatenated to form the spatiotemporal feature vector of the receiving end. This proves that its dimension matches the dictionary template vector. The mathematical structure of this feature vector is expressed as: .

[0088] Extracting the spatiotemporal feature vector at the receiving end Simultaneously, the processor calls the pre-built mapping dictionary in the non-volatile memory in parallel. It should be noted that the construction process of this mapping dictionary includes: in the offline phase before system deployment, dictionary calibration is completed in a calibration pool with clear water and no multipath scattering; firstly, time synchronization between the transmitter and receiver is achieved using a global hardware clock; then, the laser transmitter is driven to traverse and send all supported binary data block combinations, with each symbol repeatedly sampled 100 times; three-dimensional spatiotemporal clustering and centroid extraction are performed on the valid events of each sample, and the average of multiple results is taken to obtain distortion-free standard spatial coordinates and pulse trigger timing; finally, they are packaged into ideal symbol template feature vectors according to a unified dimensional order and stored in the non-volatile memory. This dictionary stores all binary data block combinations supported by the system and their corresponding ideal symbol template feature vectors. The dictionary is set to... The template feature vectors are Its data dimensions and To maintain consistency, the standard spatial emission coordinates and pulse triggering timing without distortion under ideal channel conditions were recorded.

[0089] To assess the physical similarity between the received damaged feature signals and each standard template, the processor traverses the mapping dictionary and calculates the spatiotemporal feature vector of the receiver one by one. With the feature vectors of each ideal symbol template in the dictionary Euclidean distance between The distance calculation process follows the following expansion formula:

[0090] In the above distance decision formula, Representing the The first dictionary template Standard three-dimensional parameters of each element; The dimensional weighting coefficient assigned to the system is used to balance the scale difference between spatial physical distance (pixel spacing) and temporal physical distance (microsecond time difference). In a typical configuration, this weighting coefficient ranges from 0.5 to 2.0 pixels per microsecond, and the system can dynamically adjust this weight based on the channel scattering intensity. Specifically, when the water turbidity is higher and scattering is stronger, leading to increased photon temporal dispersion, the temporal dimension weight is correspondingly reduced. (Euclidean distance) The smaller the value, the better the fit between the received features and the dictionary template.

[0091] After completing the full library traversal calculation, the processor performs a global minimum search on all distance outputs to find the specific value with the minimum Euclidean distance. The system directly outputs the dictionary template index corresponding to this minimum distance, using it as the decision symbol under the hard decision logic. Finally, according to the inverse mapping rule that corresponds one-to-one with the transmitter's mapping rule, the index value of the decision symbol is converted into a raw binary data block of the corresponding length, and then restored to a continuous raw binary bit stream through parallel-to-serial conversion logic. This completes the entire communication closed loop from physical channel perception coupled with near-shore multipath scattering and dynamic refractive index to digital information reconstruction.

[0092] To verify the technical effectiveness of the embodiments of the present invention, numerical simulations and theoretical comparisons were conducted in a simulated near-shore channel environment with an equivalent turbidity of 20-50 NTU (corresponding to an optical transmission attenuation length of 1.5-3.0) based on the aforementioned optical field evolution model and acousto-optic physical control model. Simulation results show that without the acousto-optic conditioning and event-driven reception mechanism of the present invention, the signal-to-noise ratio of the traditional coherent receiver deteriorates sharply to below 0 dB when the transmission distance exceeds 10 meters, and the theoretical bit error rate is as high as... The magnitude of the previous system caused a near-complete communication outage. However, by employing the system and method described in this embodiment (setting the laser wavelength to 532nm, the ultrasonic frequency to 1MHz, and the DVS threshold to 0.2), under the same water quality and distance, the simulation system was able to stably lock the acoustic resonance peak, improving the equivalent received signal-to-noise ratio by approximately 15dB to 20dB, and the expected bit error rate after decoding was stably reduced to [missing value]. The theoretical model and numerical simulations of this invention definitively demonstrate its ability to solve the communication interruption problem under severe multipath scattering in nearshore waters, and its great potential to achieve highly robust underwater data transmission.

[0093] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A near-shore long-distance data transmission method based on marine laser communication, characterized in that, include: Emit a laser beam carrying the raw data; An ultrasonic wave is emitted towards the water area in front of the receiving end to construct an ultrasonic field, causing a periodic change in the refractive index of the water medium in that area. This results in the superposition of a preset acoustic frequency amplitude modulation onto the speckle light field formed by the multipath scattering of the laser beam by the near-sea water to form a receiving light field. The received light field is monitored using a dynamic visual sensor array. When the absolute value of the logarithmic change in light intensity of a pixel reaches a preset trigger threshold, the corresponding pixel outputs a discrete event, generating an asynchronous event stream. The asynchronous event stream is subjected to locking filtering based on the preset acoustic frequency to extract a subset of valid signal events; The subset of valid signal events is decoded to restore the original data.

2. The near-shore long-distance data transmission method based on marine laser communication according to claim 1, characterized in that, The laser beam carrying the raw data includes: The original data in binary form is mapped to a symbol set, which is defined by the spatial coordinates of the activated transmitting unit and the light pulse trigger timestamp. A two-dimensional vertical cavity surface-emitting laser array is driven to output optical pulses according to the symbol set to obtain the laser beam.

3. The near-shore long-distance data transmission method based on marine laser communication according to claim 1, characterized in that, The process of emitting ultrasonic waves towards the water area in front of the receiver to construct an ultrasonic field includes: The ultrasonic waves are emitted using a ring-shaped ultrasonic transducer placed at the front end of the receiver to construct an ultrasonic standing wave. The refractive index of the water medium is fluctuated by the radiation pressure of the ultrasonic standing wave, thereby constructing a dynamic refractive index field with the preset acoustic frequency, and the speckle light field is superimposed with the periodic amplitude modulation.

4. The near-shore long-distance data transmission method based on marine laser communication according to claim 1, characterized in that, The step of monitoring the received light field using a dynamic visual sensor array and generating an asynchronous event stream includes: The pixels in the dynamic vision sensor array independently monitor the logarithm of the incident light intensity. When the absolute value of the logarithmic change in light intensity between the current time and the previous trigger time reaches the preset trigger threshold, the corresponding pixel outputs the discrete event containing spatial coordinates, timestamp, and polarity and updates the trigger time. All the discrete events generated within the receiving period constitute the asynchronous event stream.

5. The near-shore long-distance data transmission method based on marine laser communication according to claim 1, characterized in that, The step of performing lock-in filtering on the asynchronous event stream based on the preset acoustic frequency to extract a subset of valid signal events includes: The density of the asynchronous event stream on the time axis is statistically analyzed to obtain the event trigger rate function; Perform time-frequency analysis on the event trigger rate function to extract the spectral peaks in the frequency domain; The effective signal event subset is extracted by using bandpass filtering to remove noise by utilizing the significant spectral peak features at the preset acoustic frequency.

6. The near-shore long-distance data transmission method based on marine laser communication according to claim 5, characterized in that, Decoding the subset of valid signal events to reconstruct the original data includes: Three-dimensional spatiotemporal clustering is performed on the subset of effective signal events to construct the spatiotemporal feature vector of the receiver; Calculate the Euclidean distance between the spatiotemporal feature vector of the receiving end and the feature vector of each ideal symbol template in the preset mapping dictionary; The template index with the minimum Euclidean distance is output as the decision symbol and restored to the original data according to the inverse mapping rule.

7. A near-shore long-distance data transmission system based on marine laser communication, characterized in that, include: Laser emitting equipment used to emit laser beams into water bodies; The acoustic-optical conditioning device is positioned at the receiving end and is used to emit ultrasonic waves; Event-driven receiving device, comprising a dynamic vision sensor array; Processor, and memory storing computer programs; The processor is communicatively connected to the laser emitting device, the acousto-optic conditioning device, and the event-driven receiving device, respectively. When the processor executes the computer program, it controls the devices to work together to realize the near-shore long-distance data transmission method based on marine laser communication as described in any one of claims 1 to 6.

8. The near-shore long-distance data transmission system based on marine laser communication according to claim 7, characterized in that, The laser emitting device includes a two-dimensional vertical cavity surface-emitting laser array; the processor is specifically used to: map raw data in binary form into a symbol set defined by spatial coordinates and timestamps, and control the two-dimensional vertical cavity surface-emitting laser array to output light pulses according to the symbol set.

9. The near-shore long-distance data transmission system based on marine laser communication according to claim 7, characterized in that, The acoustic-optical conditioning device includes a ring-shaped ultrasonic transducer; the processor is specifically used to: control the ring-shaped ultrasonic transducer to emit ultrasonic waves to construct ultrasonic standing waves, thereby constructing a dynamic refractive index field with a preset acoustic frequency in the water body.

10. The near-shore long-distance data transmission system based on marine laser communication according to claim 7, characterized in that, The memory is pre-loaded with a mapping dictionary; the processor is specifically used to: acquire the asynchronous event stream generated by the event-driven receiving device; extract a subset of valid signal events by performing lock-in filtering on the asynchronous event stream based on the preset acoustic frequency; perform spatiotemporal clustering on the subset of valid signal events to obtain the spatiotemporal feature vector of the receiving end, and restore the original data by matching the decision symbol with the minimum Euclidean distance of the feature vector in the mapping dictionary.

Images