Adaptive transmission method and system for underwater wireless optical communication based on safety threshold atlas

Abstract

This invention relates to an adaptive transmission method and system for underwater wireless optical communication based on a safety threshold map. The method includes the following steps: S2, generating a turbulent light intensity sequence based on turbulent light intensity statistical parameters and target turbulent coherence time; establishing a waveform-level link evaluation model based on the turbulent light intensity sequence; and obtaining a baseline signal-to-noise ratio and a dispersion-containing threshold anchor point through threshold search; S3, constructing a safety threshold predictor based on the dispersion-containing threshold anchor point; and generating a safety threshold map on a two-dimensional plane composed of candidate transmission rates and root-mean-square delay spread; S4, querying the safety threshold map to obtain safety thresholds and status labels, calculating the safety margin, and selecting the transmission rate based on the safety margin. This invention can simultaneously consider multipath dispersion and turbulent time fading, compressing the high-overhead waveform-level threshold search into a calibrable, interpretable, and online-queryable safety threshold map.

Description

Technical Field

[0001] This invention relates to the field of optical wireless communication technology, and in particular to an adaptive transmission method and system for underwater wireless optical communication based on a security threshold spectrum. Background Technology

[0002] Underwater wireless optical communication typically utilizes the relatively low attenuation window of blue-green light in water for information transmission. Compared to underwater acoustic communication, underwater wireless optical communication offers advantages such as high transmission rate, low latency, strong directionality, and good electromagnetic concealment, making it valuable for applications in ocean observation, submersible communication, underwater sensor networks, and high-speed data transmission from underwater platforms. A typical underwater wireless optical communication system employs an intensity modulation / direct detection structure, where the transmitter modulates the light intensity using a light-emitting diode or laser, and the receiver recovers the information using a photodetector or array receiver.

[0003] However, underwater wireless optical communication links are affected by factors such as water absorption, scattering, turbulence, bubbles, temperature and salinity gradients, and receiver noise, and their reliability varies significantly with the environment and link status. Absorption and scattering cause path loss, reducing received optical power; multipath scattering broadens the channel impulse response in time, causing the energy of transmitted symbols to leak into the decision window of adjacent symbols, thus generating inter-symbol interference; water turbulence and platform disturbances cause the received optical intensity to fluctuate over time, resulting in a time-dependent fading of the available signal-to-noise ratio. For high-speed links, the symbol period can be on the order of nanoseconds, while the turbulence correlation time can be on the order of milliseconds. These two mechanisms, operating on different time scales, jointly determine the bit error rate and available transmission rate.

[0004] Existing underwater wireless optical communication channel modeling techniques typically describe them based on a single mechanism. For example, some techniques obtain the channel impulse response through Monte Carlo ray tracing, radiative transfer equations, or empirical models to characterize absorption, scattering, and multipath spread; others use log-normal distribution, Weibull distribution, Gamma-Gamma distribution, or exponential-generalized gamma mixed distribution to describe turbulent light intensity statistics; still others employ link adaptive techniques to select the rate based on received signal-to-noise ratio, bit error rate feedback, or link quality estimation.

[0005] The above methods still have the following shortcomings in engineering implementation. For example, firstly, although the static channel impulse response model can describe multipath dispersion, it is difficult to simultaneously preserve the time-dependent fading of turbulent light intensity; while the standalone turbulence statistical model can describe the fluctuations in received light intensity, it cannot directly reflect symbol-level multipath memory. Secondly, if waveform-level bit error rate threshold search is performed point-by-point for all candidate transmission rates and all possible delay spread conditions, the computational load is large, making it difficult to use directly as an online control algorithm. Summary of the Invention

[0006] The purpose of this invention is to provide an adaptive transmission method and system for underwater wireless optical communication based on a security threshold map, in order to improve the problems in the prior art where dynamic underwater channels are difficult to simultaneously characterize multipath dispersion memory and turbulent time-related fading, and where the computational complexity of full-parameter spatial waveform-level threshold search is high. This invention can simultaneously consider multipath dispersion and turbulent time fading, and compress the high-overhead waveform-level threshold search into a calibrable, interpretable, and online-queryable security threshold map.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] An adaptive transmission method for underwater wireless optical communication based on a security threshold map includes the following steps:

[0009] S1, obtain the candidate transmission rate set, target bit error rate, equivalent channel impulse response, turbulent light intensity statistical parameters, target turbulent coherence time and receiver decision configuration parameters, and extract the root mean square delay spread;

[0010] S2, generate a turbulent light intensity sequence based on the statistical parameters of turbulent light intensity and the coherence time of the target turbulent flow, establish a waveform-level link evaluation model based on the turbulent light intensity sequence, and obtain the benchmark signal-to-noise ratio and the dispersion-containing threshold anchor point through threshold search;

[0011] S3, Construct a security threshold predictor based on the dispersion threshold anchor point, and generate a security threshold map on the two-dimensional plane formed by the candidate transmission rate and the root mean square delay spread;

[0012] S4. Query the security threshold map to obtain the security threshold and status label, calculate the security margin, and select the transmission rate based on the security margin.

[0013] In a further optimized scheme, step S4 includes the following sub-steps:

[0014] For each candidate rate R in the candidate rate set, query the safety threshold map to obtain the safety threshold and state label, and calculate the safety margin. , The currently available signal-to-noise ratio, The safety threshold corresponding to the candidate rate R

[0015] Exclude candidate rates whose status label is invalid;

[0016] For the remaining candidate rates, select the highest candidate rate whose status label is "normally available" and whose safety margin meets the preset requirements; if there is no normally available candidate rate that meets the conditions, select the candidate rate with the largest safety margin.

[0017] S4 further includes the following sub-steps:

[0018] Let the current rate be The current rate has been maintained for the number of windows. ;

[0019] Only when the candidate has a higher rate satisfy And satisfy ,and The rate upgrade is only allowed when the status label is normal and available; To upgrade the hysteresis threshold, Minimum stay time;

[0020] If the current rate enters a near-failure state or a failure state, or if the current safety margin is lower than the downlink protection threshold, a rate reduction operation will be performed.

[0021] Traditional greedy rate selection often prioritizes the highest instantaneous rate, which can lead to selecting excessively risky rates when approaching forward error correction thresholds or dispersion failure boundaries. This results in insufficient safety margins, frequent rate switching, or exposure to near-failure states. The proposed solution, through safety margins, prioritizing normal zone selection, hysteresis thresholds, and minimum dwell time, enables the system to maintain high transmission rates while reducing the risks of near-failure exposure and frequent rate switching.

[0022] An underwater wireless optical communication adaptive transmission system based on a security threshold map includes the following components:

[0023] The channel parameter acquisition module is used to acquire candidate transmission rate set, target bit error rate, equivalent channel impulse response, turbulent light intensity statistical parameters, target turbulent coherence time and receiver decision configuration parameters, and extract root mean square delay spread;

[0024] The channel state generation module is used to generate a turbulent light intensity sequence based on turbulent light intensity statistical parameters and target turbulent coherence time.

[0025] The waveform-level link evaluation module is used to establish a waveform-level link evaluation model based on the turbulent light intensity sequence, and to obtain the benchmark signal-to-noise ratio and dispersion-containing threshold anchor points through threshold search.

[0026] The security threshold map generation module is used to construct a security threshold predictor based on dispersion threshold anchor points and generate a security threshold map on a two-dimensional plane composed of candidate transmission rates and root mean square delay spread.

[0027] The adaptive rate control module is used to query the security threshold map to obtain the security threshold and status label, calculate the security margin, and select the transmission rate based on the security margin.

[0028] A computer program product includes computer-readable instructions that, when executed by a processor, implement the steps of the underwater wireless optical communication adaptive transmission method based on a security threshold map.

[0029] A computer-readable storage medium including computer-readable instructions that, when executed by a processor, implement the steps in the underwater wireless optical communication adaptive transmission method based on a security threshold map.

[0030] An electronic device includes: a memory storing program instructions; and a processor connected to the memory, executing the program instructions in the memory to implement the steps of the underwater wireless optical communication adaptive transmission method based on a security threshold map.

[0031] Compared with the prior art, the present invention has at least the following beneficial effects:

[0032] First, this invention combines the normalized equivalent channel impulse response and the time-dependent turbulent light intensity sequence, which can simultaneously describe the symbol-level memory caused by multipath dispersion and the slow time memory caused by turbulent fading.

[0033] Second, this invention uses the inter-symbol interference-free reference signal-to-noise ratio as the physical benchmark and the waveform-level threshold search result as the threshold anchor point to compress the high-overhead bit error rate search result into a semi-analytical secure threshold predictor, which has the characteristics of being interpretable, calibrable, and queryable.

[0034] Third, this invention unifies transmission rate and channel delay spread to the same risk scale through a dimensionless dispersion descriptor, so that dispersion risk under different rates and different channel spread conditions can be expressed in the same spectrum.

[0035] Fourth, this invention explicitly divides the normal usable state, near-failure state and failure state into safety threshold maps, avoiding extrapolation of finite threshold prediction results to high dispersion regions that are no longer suitable for operation.

[0036] Fifth, this invention enables the system to reduce the risk of near-failure exposure and frequent rate switching while maintaining a high transmission rate through safety margin, priority selection of normal areas, hysteresis threshold and minimum dwell time.

[0037] Sixth, the edge distribution, turbulence correlation time, bit error rate target, safety margin, near-failure boundary and failure boundary of the present invention can all be recalibrated according to specific sea conditions, receiver, modulation format and engineering safety level, and have good engineering adaptability.

[0038] Seventh, this invention is applicable to blue-green light underwater communication links, short-range high-speed underwater platform interconnection, underwater sensor network data backhaul, underwater vehicle short-range communication, and other optical wireless links affected by multipath dispersion and turbulent fading. Attached Figure Description

[0039] Figure 1 The flowchart illustrates an adaptive transmission method for underwater wireless optical communication based on a security threshold map, as provided in this embodiment.

[0040] Figure 2 The flowchart illustrates the equivalent channel impulse response processing and root mean square delay spread extraction provided in this embodiment of the invention.

[0041] Figure 3 A flowchart for generating dynamic turbulent light intensity sequences provided in an embodiment of the present invention.

[0042] Figure 4 A flowchart of waveform-level link evaluation and threshold anchor search provided for embodiments of the present invention.

[0043] Figure 5 A flowchart illustrating the construction process of the physical anchoring security threshold predictor provided in this embodiment of the invention.

[0044] Figure 6 This is a schematic diagram of the security threshold map and state region division provided for embodiments of the present invention.

[0045] Figure 7 The flowchart of online adaptive rate control provided in the embodiments of the present invention.

[0046] Figure 8 This is a schematic diagram of the system module structure provided in an embodiment of the present invention.

[0047] Figure 9 A comparison chart of threshold anchor points and safety threshold prediction curves provided in the embodiments of the present invention.

[0048] Figure 10 This is a schematic diagram illustrating dynamic channel replay and online rate selection provided in an embodiment of the present invention.

[0049] Figure 11 This is a performance comparison chart of different rate control strategies provided in the embodiments of the present invention. Detailed Implementation

[0050] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0051] like Figure 1 As shown in the figure, this embodiment provides an adaptive transmission method for underwater wireless optical communication based on a security threshold map, which includes the following steps:

[0052] S1, Obtain the candidate transmission rate set and target bit error rate. Equivalent channel impulse response h[p], turbulent light intensity statistical parameters Turbulent coherence time and receiver decision configuration parameters (including oversampling factor) (Signal integration window, sliding threshold window length W, decision rules, and optional equalization parameters) to extract root mean square delay spread. ;

[0053] Among them, the candidate transmission rate set is determined by the rate level supported by the transceiver hardware and communication protocol; the target bit error rate is determined by the forward error correction threshold or system reliability requirements; the equivalent channel impulse response can be obtained from tank experiments, sea trials, Monte Carlo ray tracing, radiative transfer models, empirical models, or open channel data; the turbulent light intensity statistical parameters are obtained by fitting the edge distribution of measured or simulated light intensity samples; the turbulent coherence time can be estimated by the decay characteristics of the autocorrelation function of the light intensity time series; the receiver decision configuration parameters are determined by the specific receiver structure and digital signal processing scheme, and are actually used in the waveform-level link evaluation and bit error rate threshold search process to convert the received waveform into decision bits and calculate the bit error rate.

[0054] S2, based on turbulent light intensity statistical parameters Coherence time with turbulence A turbulent light intensity sequence I[n] is generated. A waveform-level link evaluation model is established based on the turbulent light intensity sequence I[n]. The benchmark signal-to-noise ratio and the dispersion-containing threshold anchor point are obtained through threshold search.

[0055] S3, Construct a security threshold predictor based on the dispersion threshold anchor point, and generate a security threshold map on the two-dimensional plane formed by the candidate transmission rate and the root mean square delay spread;

[0056] S4. Query the security threshold map to determine the current link status, and select the transmission rate based on the current link status.

[0057] Steps S1-S3 can be performed in the offline calibration stage, while step S4 is performed in the online adaptive stage. Steps S1-S3 compress the high-overhead waveform-level threshold search into a calibrable, interpretable, and online-queryable safe threshold map, which can then be queried and used online in step S4.

[0058] Please see Figure 2 The equivalent channel impulse response h[p] can be obtained from tank measurements, sea trial data, Monte Carlo ray tracing, numerical solutions to the radiative transfer equation, empirical models, open channel data, or other equivalent methods. To decouple the multipath dispersion effect from the static path loss, this embodiment normalizes the equivalent channel impulse response to satisfy: , Where p represents the discrete-time channel delay tap number of the equivalent channel impulse response, and h[p] represents the normalized channel response value corresponding to the p-th channel delay tap number. Normalization is used to decouple static path loss from multipath dispersion effects, so that h[p] mainly describes the temporal spread of signal energy and inter-symbol interference characteristics.

[0059] In continuous-time representation, τ is the propagation delay variable, representing the average channel delay. Root mean square delay spread They are defined as follows: , Where h(τ) represents the normalized equivalent channel impulse response in continuous-time form, Used to describe the time spread of received energy relative to the average time delay. If If the signal strength is relatively small, the signal energy is mainly concentrated in the target symbol decision window; if... If the energy leakage between adjacent symbols increases, the risk of inter-symbol interference will also increase.

[0060] Please see Figure 3 To simultaneously maintain the target edge distribution and temporal correlation of turbulent light intensity, this embodiment uses the Gaussian Copula autoregressive generation method to generate the turbulent light intensity sequence I[n].

[0061] First, construct the latent Gaussian autoregressive sequence (B): , , ; where "~" indicates "follows a certain probability distribution", () represents a standard normal distribution with a mean of 0 and a variance of 1; n represents the sampling time of the turbulent time series. Let ρ represent the latent Gaussian random variable at the nth sampling time; ρ represents the correlation coefficient between adjacent samples. These are mutually independent standard Gaussian random perturbations. . This represents the time step of the sequence.

[0062] Then, a probability integral transformation is performed using the standard normal distribution function: ;

[0063] Then, the inverse cumulative distribution function of the target edge distribution is used to generate a turbulent light intensity sequence through mapping: ,in, For the edge distribution of target light intensity, These are statistical parameters for turbulent light intensity. The edge distribution of the target light intensity can be selected from a log-normal distribution, a Weibull distribution, a Gamma-Gamma distribution, an exponential-generalized gamma mixture distribution, or it can be obtained by fitting measured underwater light intensity data.

[0064] This embodiment uses the Gaussian Copula autoregressive generation method to generate the turbulent light intensity sequence I[n]. Alternatively, an ARMA model, a higher-order autoregressive model, a filtered Gaussian process, or other time-related models can be used. The recursive expression will change accordingly, but for standard Gaussian latent variables, it can still be used. A probability integral transformation is performed, and then a light intensity sequence is generated through the inverse cumulative distribution function of the target edge distribution. If the latent variable does not follow a standard normal distribution, Φ(·) can be replaced with the cumulative distribution function Fz(·) of the latent variable itself, while the overall steps of inverse mapping of the target edge distribution through a uniform variable still apply.

[0065] To ensure a consistent average power scale across different sea state benchmarks, the generated turbulent light intensity sequence was mean-normalized: N represents the total number of sampling points in the generated turbulent light intensity sequence, i.e., the sequence length; m is the summation index when calculating the sample mean. This processing makes the average value of the normalized turbulent light intensity sequence approximately 1, thus separating the description of average path loss from the relative fluctuations of turbulence.

[0066] Through the above processing, the resulting turbulent light intensity sequence I[n] has both a specified edge distribution and a specified time correlation, and can be used for dynamic link replay, bit error rate threshold search, and online available signal-to-noise ratio mapping.

[0067] Please see Figure 4 This embodiment uses an intensity modulation / direct detection underwater wireless optical communication link employing non-return-to-zero on-off keying modulation as an example for illustration. The transmitting end uses non-return-to-zero on-off keying modulation, and the candidate transmission rate is... The bit period is The oversampling factor is The transmitted waveform can then be represented as: Where k represents the sequence number of the transmitted symbol or transmitted bit; i represents the sequence number of the sampling point within one symbol period; b[k] represents the number of sampling points corresponding to each symbol; b[k] represents the k-th transmitted bit. .

[0068] Channel convolution: The convolution of the transmitted waveform with the equivalent channel impulse response h[n] is... ;

[0069] Turbulent multiplicative decay: , represents the received sequence before noise injection, and I[n] is the turbulent light intensity sequence.

[0070] Noise injection: Adding receiver noise The received sequence is then: .

[0071] Given an input signal-to-noise ratio When the noise variance is: , The input signal-to-noise ratio in decibels is: .

[0072] Integral decision: The receiver integrates the sampled values ​​within each symbol period to obtain the symbol decision statistic. .

[0073] To accommodate baseline changes caused by slow turbulence fading, this embodiment employs a sliding window averaging threshold: W represents the maximum number of signs contained in the sliding window, and j is the summation index within the sliding window. It is the adaptive decision threshold corresponding to the kth symbol.

[0074] The hard decision result at the receiving end is: b[k] is the kth original bit transmitted by the sender. This is the receiver's estimate of the bit. This is achieved by comparing b[k] with... The error rate is calculated based on whether the results are consistent.

[0075] The currently available signal-to-noise ratio can be mapped from the nominal signal-to-noise ratio and the turbulence intensity factor as follows: ,in, I²(t) represents the nominal received electrical signal-to-noise ratio when the normalized turbulent light intensity I=1; I(t) represents the normalized turbulent light intensity factor at the current moment, and I[n] is the discrete sampling form of I(t). I²(t) is used here because in IM / DD receivers, the photocurrent amplitude is approximately proportional to the received light intensity, while the received electrical signal power is approximately proportional to the square of the light intensity.

[0076] Primarily used for offline waveform-level threshold search, corresponding to offline threshold calibration; Primarily used for online spectrum lookup and rate adaptive decision-making. Transmit waveform, channel convolution, turbulent multiplicative fading, noise injection, integral decision, and hard decision are used for offline calculation of different input signal-to-noise ratios. The bit error rate (BER) under the given conditions is used to obtain the safety threshold that satisfies the target BER. Used to estimate the available signal-to-noise ratio of the current link online and compare it with the offline security threshold.

[0077] Please continue reading. Figure 4 To establish reliability benchmarks under different sea state conditions, this embodiment first sets up a condition free of inter-symbol interference: Or, in discrete time, it can be represented as: Under these conditions, the link retains only turbulent light intensity fading and receiver noise. The baseline signal-to-noise ratio (SNR) is determined by scanning the input SNR and calculating the bit error rate. .

[0078] This benchmark signal-to-noise ratio describes the minimum signal-to-noise ratio required to resist corresponding turbulent fading and achieve the target bit error rate when there is no multipath dispersion.

[0079] Under conditions involving multipath dispersion, several representative [links] were evaluated using a waveform-level link evaluation model. and A waveform-level threshold search is performed to determine the total signal-to-noise ratio requirement that satisfies the target bit error rate. The total signal-to-noise ratio requirement can be used as a threshold anchor point for the fitting and calibration of the subsequent safety threshold predictor.

[0080] With a candidate rate And a given root mean square delay spread For example, the waveform-level threshold search process can be performed according to the following steps:

[0081] (1) Fixed candidate rate And select the corresponding normalized channel impulse response h[p] or ;

[0082] (2) Generate the transmitted bit sequence and the turbulent light intensity sequence with target edge distribution and time correlation;

[0083] (3) Perform waveform shaping representation, channel convolution, turbulence multiplicative fading, and receiver noise injection;

[0084] (4) Obtained according to the receiving end's decision rules And compare it with b[k] to calculate the bit error rate;

[0085] (5) Change the input signal-to-noise ratio Repeat the link evaluation process of steps (2)-(4) above;

[0086] (6) Use a step-by-step scanning method or a "coarse scan plus binary search" method to find the solution that satisfies the condition. The lowest input signal-to-noise ratio, denoted as . .

[0087] Under the condition of no inter-symbol interference h[p]=δ[p], the same search can be performed to obtain... The threshold anchor point is obtained under conditions of multipath dispersion. The difference between the two is used to extract the inter-symbol interference cost caused by multipath dispersion.

[0088] Under conditions involving multipath dispersion, the receiver integral decision factor, neglecting receiver noise, can be approximated as: Where k represents the current symbol number; Indicates the cursor or symbol interval relative to the current symbol; The slowly varying turbulent light intensity factor is approximately invariant around the k-th symbol. For the first The effective channel coupling coefficient corresponding to each symbol interval This indicates the adjacent transmitted symbols that affect the current decision.

[0089] For oversampling factors The effective channel coefficient can be expressed as: Among them, the main cursor Characterizing target symbol energy, non-zero vernier Characterizes leakage from the previous and next cursors of adjacent symbols.

[0090] The normalized inter-symbol interference leakage ratio can be defined as: .

[0091] To unify candidate transmission rates and channel delay spread to the same risk scale, this embodiment defines a dimensionless dispersion descriptor: .

[0092] when In Mb / s When the unit is ns, it can be used .

[0093] Furthermore, the total signal-to-noise ratio requirement can be decomposed as follows: ,in, This indicates the potential bandwidth cost at both the sender and receiver ends. This represents the true inter-symbol interference cost obtained from waveform-level threshold search. In embodiments where receiver, sampling, and bandwidth conditions are fixed, it can be set to... This allows us to focus on describing the cost of inter-symbol interference caused by dispersion.

[0094] Please see Figure 5 Directly performing waveform-level threshold search on all candidate transmission rates and all delay spread combinations would incur high computational overhead. Therefore, this embodiment utilizes a finite number of threshold anchors to construct a semi-analytical secure threshold predictor.

[0095] First, extract the inter-symbol interference cost reference value from the waveform-level threshold search results: ; Indicates the inter-symbol interference-free reference signal-to-noise ratio under a specific sea state or channel reference; Case indicates the sea state, turbulence statistics, or other reference channel scenario number. This represents the desired signal-to-noise ratio reference value obtained through waveform-level link simulation and bit error rate threshold search when the dimensionless dispersion descriptor is (D).

[0096] Then, a monotonically non-decreasing inter-symbol interference cost trend function is fitted within the stable operating region. Preferably, this function can be in logarithmic saturation form: , ,in, Let A represent the small dispersion bias, and let A represent the cost scale. This indicates a turning point; 'p' indicates the steepness of the ascent. This indicates that the dispersive boundary fails. Its derivative is: .

[0097] when When the above function is non-decreasing within the reachable region, it conforms to the physical constraint that the dispersion risk does not decrease as D increases.

[0098] To compensate for model residuals, this embodiment sets a safety margin for compensating for underestimation errors. Let the prediction error be: .

[0099] In the core verification area: .

[0100] Choose a safety margin that meets the following requirements: .

[0101] in, Indicates taking the positive part; An optional conservative margin is added. Ultimately, the safety threshold predictor is defined as: .

[0102] in, It is a dimensionless dispersion descriptor; Here is the dispersion descriptor corresponding to the i-th threshold anchor point; The boundary of the core stability verification region; The set of anchor points within the core verification area; The reference value for inter-symbol interference cost obtained from waveform-level threshold search; The trend function provides the predicted cost of inter-symbol interference. This represents the prediction error. This represents the conservative signal-to-noise ratio threshold after adding a safety margin, i.e., the safety threshold.

[0103] For operating points exceeding the dispersion failure boundary, the cost of inter-symbol interference can be expressed as: Let the cost be... This does not refer to an actual physical signal-to-noise ratio of infinitely large, but rather to mathematically marking the operating point as unavailable to prevent the online controller from continuing to select candidate rates that exceed the dispersion failure boundary.

[0104] With the above settings, the system will not extrapolate the finite safety threshold to high dispersion regions that are no longer suitable for operation.

[0105] Please see Figure 6 Based on the security threshold predictor, in the candidate transmission rate Root mean square delay spread A security threshold map is generated on the constructed two-dimensional parameter plane. The security threshold map can be generated offline using a discrete grid approach. That is, for each candidate transmission rate in the candidate transmission rate set... Root mean square delay spread The safety threshold for this grid point is calculated based on the safety threshold predictor. Then, according to , And optional The operating point is determined to be in an OK (normal availability), NEAR (near failure), or OUTAGE (failure) state. After traversing all candidate rate and delay spread combinations, a security threshold map is generated that can be queried or interpolated online. Therefore, each entry in the security threshold map must include at least one candidate transmission rate. Root mean square delay spread Dimensionless dispersion descriptor D, safety threshold and status labels.

[0106] Status labels are determined by the following rules: In this context, OK indicates a normal and available state, NEAR indicates a near-failure state, and OUTAGE indicates a failed state. This is the near-failure initiation threshold. The failure threshold This is the preset upper limit for the signal-to-noise ratio.

[0107] In an illustrative example, it is advisable to In other embodiments, the calibration can be recalibrated based on system hardware, target bit error rate, modulation format, receiver decision method, and engineering security level.

[0108] By using a safety threshold map, candidate operating points in the continuous parameter space can be mapped to queryable safety states, thus providing a clear engineering interface for online adaptive transmission. During the online operation phase, only the safety threshold map needs to be queried, eliminating the need to re-execute the expensive waveform-level BER threshold search.

[0109] Please see Figure 7 During the online operation phase, the currently available signal-to-noise ratio is periodically acquired. And the current channel delay spread estimate. For the candidate rate set, for each candidate rate Query the security threshold map to obtain the security threshold and status label, and calculate the security margin: First, exclude candidate rates with an OUTAGE status. For the remaining candidate rates, prioritize the highest candidate rate with an OK status and a safety margin that meets preset requirements. If no normally usable candidate rate meets the conditions, select the lower rate with the largest safety margin, or maintain, reduce, or enter protection mode according to preset safety policies.

[0110] The current channel delay spread estimate can be obtained by the receiver periodically estimating the current discrete channel impulse response h[p] using pilot sequences, training sequences, or preambles. After nonnegating and normalizing the estimated channel impulse response, the average delay and root mean square delay spread formulas are applied as in the offline stage. Calculation. Channel impulse response estimation can employ methods such as correlation detection, least squares channel estimation, sparse channel estimation, training sequence matching, or receiver equalizer tap estimation. Since the rate of change of underwater channel delay spread is typically much lower than the symbol rate, it can be updated at a low frequency within a fixed time window, eliminating the need for re-estimation for each symbol. In implementations lacking real-time channel estimation capabilities, the corresponding delay spread estimate can be obtained by querying based on the current sea state classification, historical measurement data, link location parameters, or a pre-established channel database.

[0111] To avoid frequent rate switching, this embodiment introduces an upgrade hysteresis threshold. and minimum stay time Let the current rate be... The current rate has been maintained for the number of windows. Only when the candidate has a higher rate. satisfy And satisfy ,and The system is only allowed to upgrade when its status is normal and available. If the current rate enters a near-failure state or a failure state, or if the current safety margin is lower than the downlink protection threshold, the system will reduce its speed. This control strategy can be called the normal zone priority hysteresis strategy. This strategy excludes high-risk operating points through the safety threshold map and reduces rate jitter by using hysteresis and minimum dwell time.

[0112] In other implementations, the generation of turbulent light intensity sequences is not limited to the Gaussian Copula-AR(1) model, but may also employ ARMA model, higher-order autoregressive model, filtered Gaussian process, spectral synthesis method, measured sequence resampling method, or other models that can simultaneously describe edge distribution and related structure.

[0113] The edge distribution of target light intensity is not limited to log-normal distribution, Weibull distribution, Gamma-Gamma distribution and exponential-generalized gamma mixed distribution, but can also be obtained by fitting measured data from a specific sea area.

[0114] The equivalent channel impulse response is not limited to Monte Carlo ray tracing, but can also come from tank experiments, sea trials, numerical solutions of radiative transfer, empirical models, neural network estimations, or public datasets.

[0115] The inter-symbol interference cost trend function is not limited to the logarithmic saturation function; it can also be a piecewise monotonic function, spline function, monotonic regression model, physically constrained neural network, or lookup table interpolation model.

[0116] Online rate control is not limited to the normal zone priority hysteresis strategy. Risk penalty utility function, model predictive control, reinforcement learning strategy or other rule control methods can also be used, but it is preferred to retain the safety threshold map as the safety constraint interface.

[0117] This invention is not limited to non-return-to-zero on / off keying modulation, but can also be extended to other intensity modulation formats, pulse amplitude modulation, pulse position modulation, orthogonal frequency division multiplexing, or optical communication links with equalized reception. The output control quantities of this invention are not limited to transmission rate, but may also include transmit power, modulation order, coding rate, equalizer parameters, retransmission strategy, or receiver decision parameters.

[0118] Please see Figure 9 To verify the reliability of the security threshold predictor, the threshold anchor point obtained by waveform-level bit error rate search can be compared with the prediction curve obtained by the inter-symbol interference cost trend function. If the prediction curve is close to the threshold anchor point in the core verification region, and the security threshold curve after adding a security margin is not lower than the main threshold anchor point, it can be concluded that the security threshold predictor is not arbitrarily set by experience, but is calibrated by waveform-level link evaluation results.

[0119] Please see Figure 10 To verify the applicability of this invention in dynamic underwater optical channels, dynamic channel replay can be performed using a turbulent light intensity sequence with target edge distribution and time correlation. During dynamic channel replay, the system queries a safety threshold map based on the currently available signal-to-noise ratio trajectory, calculates the safety margin under different candidate transmission rates, and outputs the transmission rate that changes over time. By demonstrating the available signal-to-noise ratio trajectory, the safety threshold trajectory, and the output rate trajectory, it can be shown that the method of this invention can perform online adaptive control under time-correlated fading conditions.

[0120] Please see Figure 11 To verify the applicability of this invention near high-speed boundaries, the available signal-to-noise ratio (SNR) trajectories under different sea state benchmarks can be boundary aligned. For example, the available SNR trajectories can be shifted to the target speed. nearby: , .

[0121] Then, the following strategies were run: Greedy (selecting the highest rate directly from candidate rates that meet the current safety SNR threshold), OK-First (prioritizing the highest rate from candidate rates that are in OK state and meet the safety margin requirements; considering other available rates only when no suitable OK rate exists), OK-First-HYS (adding hysteresis margin and minimum dwell time to OK-First to reduce frequent rate switching), and AMC-HYS (adjusting the rate based on the relationship between available SNR and various rate thresholds, and adding hysteresis and minimum dwell time mechanisms). The average selected rate, average safety margin, number of rate switching times, and near-failure state exposure ratio were compared. Figure 11 This demonstrates that the method of the present invention can perform risk perception transmission control under operating conditions close to the common high-rate decision boundary.

[0122] Please see Figure 8 This embodiment also provides an underwater wireless optical communication adaptive transmission system based on a security threshold map, comprising the following modules:

[0123] The channel parameter acquisition module is used to acquire candidate transmission rate set, target bit error rate, equivalent channel impulse response, turbulent light intensity statistical parameters, target turbulent coherence time and receiver decision configuration parameters, and extract root mean square delay spread;

[0124] The channel state generation module is used to generate a turbulent light intensity sequence based on turbulent light intensity statistical parameters and target turbulent coherence time.

[0125] The waveform-level link evaluation module is used to establish a waveform-level link evaluation model based on the turbulent light intensity sequence, and to obtain the benchmark signal-to-noise ratio and dispersion-containing threshold anchor points through threshold search.

[0126] The security threshold map generation module is used to construct a security threshold predictor based on dispersion threshold anchor points and generate a security threshold map on a two-dimensional plane composed of candidate transmission rates and root mean square delay spread.

[0127] The adaptive rate control module is used to query the security threshold map to obtain the security threshold and status label, calculate the security margin, and select the transmission rate based on the security margin.

[0128] For the specific execution methods and steps of each module, please refer to the relevant descriptions in the aforementioned method embodiments, which will not be repeated here.

[0129] This embodiment also provides an electronic device that may include a processor and a memory, wherein the memory is coupled to the processor. The electronic device may also include an input unit, a display unit, and a power supply.

[0130] A processor, sometimes also called a controller or operating control, may include a microprocessor or other processor device and / or logic device that receives input and controls the operation of various components of an electronic device.

[0131] The memory may be one or more of the following: a cache, flash memory, hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It can store information such as the processor's configuration information and the instructions executed by the processor. The processor can execute programs stored in the memory to perform information storage or processing. In one embodiment, the memory also includes a buffer memory, or buffer, to store intermediate information.

[0132] This invention also provides a computer program product including computer-readable instructions. When the computer-readable instructions are executed in an electronic device, the program product causes the electronic device to perform the operation steps included in the method of this invention.

[0133] This invention also provides a storage medium storing computer-readable instructions that cause an electronic device to perform the operation steps included in the method of this invention.

[0134] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0135] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0136] The above specific embodiments are merely several optional embodiments of the present invention. Based on the technical solutions of the present invention and the relevant teachings of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.

Claims

1. An adaptive transmission method for underwater wireless optical communication based on a security threshold map, characterized in that, Includes the following steps: S1, obtain the candidate transmission rate set, target bit error rate, equivalent channel impulse response, turbulent light intensity statistical parameters, target turbulent coherence time and receiver decision configuration parameters, and extract the root mean square delay spread; S2, generate a turbulent light intensity sequence based on the statistical parameters of turbulent light intensity and the coherence time of the target turbulent flow, establish a waveform-level link evaluation model based on the turbulent light intensity sequence, and obtain the benchmark signal-to-noise ratio and the dispersion-containing threshold anchor point through threshold search; S3, Construct a security threshold predictor based on the dispersion threshold anchor point, and generate a security threshold map on the two-dimensional plane formed by the candidate transmission rate and the root mean square delay spread; S4. Query the security threshold map to obtain the security threshold and status label, calculate the security margin, and select the transmission rate based on the security margin.

2. The underwater wireless optical communication adaptive transmission method based on security threshold maps according to claim 1, characterized in that, In step S2, the process of generating a turbulent light intensity sequence based on turbulent light intensity statistical parameters and turbulent coherence time includes: A model based on turbulent coherent temporal construction that simultaneously describes the target edge distribution and temporal correlation of turbulent light intensity; Perform a probability integral transformation on the model; Using the inverse cumulative distribution function of the target marginal distribution to transform the integral model Mapping is performed to obtain the initial turbulent light intensity sequence. ; For the edge distribution of target light intensity, These are statistical parameters of turbulent light intensity; The initial turbulent light intensity sequence is normalized to obtain the turbulent light intensity sequence. N represents the total number of sampling points in the generated turbulent light intensity sequence, and m is the summation index when calculating the sample mean.

3. The underwater wireless optical communication adaptive transmission method based on security threshold maps according to claim 1, characterized in that, In step S2, the process of establishing a waveform-level link evaluation model based on the turbulent light intensity sequence includes: Transmitted waveform representation: The bit period is defined as... The transmitted waveform is then represented as i represents the sampling point number within one symbol period; b[k] is the oversampling factor, representing the number of sampling points corresponding to each symbol; b[k] represents the k-th transmitted bit. Channel convolution: The convolution of the transmitted waveform with the equivalent channel impulse response h[n] is... ; Turbulent multiplicative decay: , which represents the received sequence before noise injection; Noise injection: Adding receiver noise Then, the received sequence is ; Integral decision: The receiver integrates the sampled values ​​within each symbol period to obtain the symbol decision statistic. ; Hard verdict: , It is the receiving end's estimation result of b[k]. W represents the maximum number of symbols contained in the sliding window; compare b[k] and Whether the bit error rate is consistently determined; Mapping: Maps the currently available signal-to-noise ratio to the nominal signal-to-noise ratio and the turbulence intensity factor. , The nominal signal-to-noise ratio is represented when the normalized turbulent light intensity I=1; the turbulent light intensity factor I[n] is the discrete sampling form of I(t).

4. The underwater wireless optical communication adaptive transmission method based on security threshold maps according to claim 1, characterized in that, In step S2, the process of obtaining the baseline signal-to-noise ratio and the dispersion-containing threshold anchor point through threshold search includes: (1) Fixed candidate rate and ; (2) Generate the transmitted bit sequence and the turbulent light intensity sequence with target edge distribution and time correlation; (3) Perform waveform shaping, channel convolution, turbulent multiplicative fading, and receiver noise injection; (4) Obtained according to the hard decision rule of the receiving end And compare it with b[k] to calculate the bit error rate (BER); (5) Change the input signal-to-noise ratio Repeat the link evaluation process of steps (2)-(4) above; (6) Find satisfaction The lowest input signal-to-noise ratio is denoted as the dispersion threshold anchor point; BER is the bit error rate. The target bit error rate; (7) The reference signal-to-noise ratio can be obtained by performing the same search under the condition of no inter-symbol interference h[p]=δ[p]. .

5. The underwater wireless optical communication adaptive transmission method based on security threshold maps according to claim 1, characterized in that, In step S3, the process of constructing a safety threshold predictor based on the dispersion threshold anchor point includes: Extract inter-symbol interference cost reference values ​​from waveform-level threshold search results. , Indicates the reference signal-to-noise ratio without inter-symbol interference; Within the stable operating region, a monotonically non-decreasing inter-symbol interference cost trend function is fitted. , ,in, Let A represent the small dispersion bias, and let A represent the cost scale. This indicates a turning point; 'p' indicates the steepness of the ascent. This indicates that the dispersive boundary has failed; Set a safety margin Let the prediction error be ; In the core verification area there are ; Choose a safety margin that meets the following requirements: , Indicates taking the positive part; An optional conservative margin is added; The safety threshold predictor is defined as: ,in, It is a dimensionless dispersion descriptor; Here is the dispersion descriptor corresponding to the i-th threshold anchor point; The boundary of the core stability verification region; The set of anchor points within the core verification area; The reference value for inter-symbol interference cost obtained from waveform-level threshold search; The trend function provides the predicted cost of inter-symbol interference. For prediction error, This represents the conservative signal-to-noise ratio threshold after adding a safety margin.

6. The underwater wireless optical communication adaptive transmission method based on security threshold maps according to claim 1, characterized in that, In step S3, the process of generating a security threshold map on a two-dimensional plane composed of candidate transmission rates and root mean square delay spread includes: For each candidate transmission rate in the candidate transmission rate set Root mean square delay spread The safety threshold for this grid point is calculated based on the safety threshold predictor. Then, based on the near-failure initiation threshold Failure threshold and the preset upper limit of signal-to-noise ratio Determine whether the operating point is in a normal, usable, or near-failure state (OK, NEAR, NEAR, or OUTAGE). After traversing all candidate rates and delay spread combinations, a security threshold map is formed; each entry in the security threshold map includes a candidate transmission rate. Root mean square delay spread Dimensionless dispersion descriptor D, safety threshold and status labels; Status label defined as .

7. The underwater wireless optical communication adaptive transmission method based on security threshold maps according to claim 1, characterized in that, S4 includes the following sub-steps: For each candidate rate in the candidate rate set Query the safety threshold map to obtain the safety threshold and status label, and calculate the safety margin. , The currently available signal-to-noise ratio, Candidate rate The corresponding security threshold; Exclude candidate rates whose status label is invalid; For the remaining candidate rates, select the highest candidate rate whose status label is "normally available" and whose safety margin meets the preset requirements; if there is no normally available candidate rate that meets the conditions, select the candidate rate with the largest safety margin.

8. The underwater wireless optical communication adaptive transmission method based on security threshold maps according to claim 7, characterized in that, S4 further includes the following sub-steps: Let the current rate be The current rate has been maintained for the number of windows. ; Only when the candidate has a higher rate satisfy And satisfy ,and The rate upgrade is only allowed when the status label is "normal and available". To upgrade the hysteresis threshold, Minimum stay time; If the current rate enters a near-failure state or a failure state, or if the current safety margin is lower than the downlink protection threshold, a rate reduction operation will be performed.

9. An underwater wireless optical communication adaptive transmission system based on a security threshold map, characterized in that, It includes the following components: The channel parameter acquisition module is used to acquire candidate transmission rate set, target bit error rate, equivalent channel impulse response, turbulent light intensity statistical parameters, target turbulent coherence time and receiver decision configuration parameters, and extract root mean square delay spread; The channel state generation module is used to generate a turbulent light intensity sequence based on turbulent light intensity statistical parameters and target turbulent coherence time. The waveform-level link evaluation module is used to establish a waveform-level link evaluation model based on the turbulent light intensity sequence, and to obtain the benchmark signal-to-noise ratio and dispersion-containing threshold anchor points through threshold search. The security threshold map generation module is used to construct a security threshold predictor based on dispersion threshold anchor points and generate a security threshold map on a two-dimensional plane composed of candidate transmission rates and root mean square delay spread. The adaptive rate control module is used to query the security threshold map to obtain the security threshold and status label, calculate the security margin, and select the transmission rate based on the security margin.

10. A computer program product comprising computer-readable instructions, characterized in that, When executed by a processor, the computer-readable instructions implement the steps of the underwater wireless optical communication adaptive transmission method based on a security threshold map as described in any one of claims 1-8.

Images