A method and system for adaptive optimization of optical transmission quality of a video stream

Abstract

The present application relates to the technical field of optical communication link regulation, in particular to a video stream optical transmission quality adaptive optimization method and system.In the present application, by constructing a cross-layer mapping mechanism between the optical physical layer nonlinear characteristics and the video application layer encoding parameters, the degree of cross-phase modulation interference caused by the interaction of multiple wavelength channels can be quantified in real time, and the directed decoupling adjustment of chroma quantization step size is triggered when the channel degradation risk is perceived, the noise interference of non-sensitive chroma data is suppressed while the brightness details are preserved, and the key reference nodes are accurately identified according to the frequency of video frames in the time domain prediction sequence, the probability shaping modulation protection based on information entropy is implemented for high value data, the limited optical transmission energy is concentrated on ensuring the integrity of the core prediction frame, the global error code diffusion caused by single frame damage is effectively curbed, and the threshold dynamic closed-loop correction logic based on real-time symbol error rate is matched.

Description

Technical Field

[0001] This invention relates to the field of optical communication link control technology, and in particular to an adaptive optimization method and system for optical transmission quality of video streams. Background Technology

[0002] The field of optical communication link control technology mainly involves technologies related to the management and regulation of optical signal transmission in optical fiber transmission systems. It usually revolves around factors such as dispersion attenuation noise and bandwidth limitations that optical signals encounter during transmission. By coordinating and controlling the parameters of each link, the stable transmission and controllable operation of information in the optical communication network can be ensured.

[0003] Among them, the adaptive optimization method for optical transmission quality of video streams refers to a type of processing method that adjusts the data transmission quality of video streams in optical communication links. Typically, before video transmission, the video stream is layered or multi-rate encoded, and the corresponding video bitrate level is selected according to preset optical link parameters (such as fixed optical power level, fixed modulation format, and fixed channel bandwidth). During transmission, based on the bit error rate, packet loss statistics, or signal-to-noise ratio threshold information returned by the receiver, the video bitrate is switched or reselected, thereby achieving a matching adjustment between the video stream and optical transmission conditions.

[0004] Existing technologies mostly rely on preset static link budgets or passive feedback mechanisms based on average bit error statistics to adjust video stream levels. This operation mode, which relies on fixed parameters and hysteresis feedback, is difficult to capture transient nonlinear physical layer damage that occurs in fiber optic transmission. In particular, rapid optical power fluctuations caused by cross-phase modulation between multiple channels often destroy signal integrity before the compensation mechanism is triggered. Furthermore, the coarse-grained adjustment strategy of switching the overall bit stream level ignores the temporal correlation and hierarchical importance differences between video frames. This makes it impossible to provide differentiated protection for key reference frames when facing sudden interference, which can easily cause the spread effect of bit errors in the time domain. This results in unrecoverable mosaic or stuttering phenomena in the video picture at the receiving end, which not only reduces the efficiency of fine utilization of optical spectrum resources, but also fails to maintain a stable visual experience in dynamically changing channel environments. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose an adaptive optimization method and system for optical transmission quality of video streams.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an adaptive optimization method for optical transmission quality of video streams, comprising the following steps:

[0007] S1: Collect the instantaneous fiber-injected optical power of the current video stream wavelength channel and adjacent video stream wavelength channels, calculate the power fluctuation statistical variance, and calculate the cross-phase modulation interference index in combination with the set fiber nonlinear coefficient.

[0008] S2: Compare the cross-phase modulation interference index with the preset optical transmission interference warning threshold, configure the colorimetric parameters of the header information for the current video slice to be encoded according to the comparison result, and set the colorimetric step size decoupling configuration.

[0009] S3: Analyze the video compressed bitstream generated based on the color metric step size decoupling configuration encoding, identify the unique frame number identifier of the current video stream prediction frame in the video sequence, and count the total number of times the video stream prediction frame is referenced.

[0010] S4: Compare the total number of times the video stream prediction frame is referenced with the preset video frame importance threshold, extract the target frame compression data of the current video stream prediction frame based on the comparison result, and construct an optical symbol probability shaping mapping table in combination with the preset standard optical communication constellation diagram.

[0011] S5: Based on the optical symbol probability shaping mapping table, the target frame compressed data is symbol-mapped and optically modulated for transmission. The optical transmission interference warning threshold and the video frame importance threshold are corrected according to the transmission results to obtain the adaptive optimization result of video stream optical transmission quality.

[0012] As a further aspect of the present invention, the cross-phase modulation interference index is specifically obtained by multiplying the power fluctuation statistical variance, the instantaneous fiber input power of the current video stream wavelength channel, and the fiber nonlinear coefficient. The colorimetric quantization step size decoupling configuration includes the sum of the original blue differential component quantization parameters and the preset positive integer offset, and the sum of the original red differential component quantization parameters and the preset positive integer offset. The total number of times the video stream prediction frame is referenced is specifically calculated by counting the total number of references of the unique frame number identifier that directly points to the current video stream prediction frame through motion vectors in subsequent video stream frames. The optical symbol probability shaping mapping table includes the coordinate mapping relationship between the symbol to be transmitted and each optical communication constellation point in the standard optical communication constellation diagram and the occurrence probability of each optical communication constellation point. The adaptive optimization result of video stream optical transmission quality includes optimizing the optical transmission interference warning threshold and optimizing the video frame importance threshold.

[0013] As a further aspect of the present invention, the step of obtaining the cross-phase modulation interference index specifically includes:

[0014] S111: Set a sliding time window based on the video stream symbol rate period, obtain the material refractive index parameters and effective mode area of ​​the optical fiber transmission medium carrying the current video stream wavelength channel, set the optical fiber nonlinear coefficient, collect the instantaneous fiber-injected optical power of the current video stream wavelength channel and the instantaneous fiber-injected optical power of adjacent video stream wavelength channels according to the preset sampling period, and construct the basic physical parameter set of the optical fiber channel.

[0015] S112: Map the instantaneous fiber-in optical power of multiple adjacent video stream wavelength channels in the optical fiber channel basic physical parameter set to a sliding time window, and statistically analyze the adjacent channel power time-domain fluctuation variance between the instantaneous fiber-in optical power of each sampling point within the sampling period and the average instantaneous fiber-in optical power within the sliding time window.

[0016] S113: Calculate the product of the adjacent channel power time-domain fluctuation variance, the instantaneous fiber input power of the current video stream wavelength channel, and the fiber nonlinear coefficient, as the cross-phase modulation interference index.

[0017] As a further aspect of the present invention, the step of obtaining the colorimetric quantization step size decoupling configuration specifically includes:

[0018] S211: When the cross-phase modulation interference index exceeds the preset optical transmission interference warning threshold, configure the chromaticity quantization parameters of the header information for the current video slice to be encoded, read and keep the luminance component quantization parameters unchanged, identify the original blue difference component quantization parameters and the original red difference component quantization parameters in the video slice header information area, and generate a set of chromaticity component parameters to be corrected.

[0019] S212: Introduce a preset positive integer offset, calculate the sum of the original blue differential quantization parameter and the positive integer offset in the set of chromaticity component parameters to be corrected, and at the same time calculate the sum of the original red differential quantization parameter and the positive integer offset to obtain the chromaticity component offset correction value.

[0020] S213: Summarize the chroma component offset correction values ​​and the luminance component quantization parameters, repackage them according to the format of the video coding standard, and use them as the chroma quantization step size decoupling configuration.

[0021] As a further aspect of the present invention, the step of obtaining the total number of times the video stream prediction frame is referenced specifically includes:

[0022] S311: Compress and encode the current video stream using the color metric step size decoupling configuration to generate a compressed video stream, parse the encoding strip header syntax structure of the compressed video stream, locate the network abstraction layer that carries the predicted frame data of the current video stream, extract the unique frame number identifier of the predicted frame of the current video stream in the video sequence from the network abstraction layer, and generate a unique sequence identifier code for the predicted frame.

[0023] S312: Retrieve the slice header information of each subsequent video stream frame located after the current video stream prediction frame in the video compressed bitstream, extract the reference image list defined in the slice header information, match and verify the unique sequence identifier code of the prediction frame with each reference image in the reference image list, and filter the reference index dataset of subsequent frames.

[0024] S313: Match entries in the subsequent frame reference index dataset that directly point to the unique sequence identifier of the predicted frame through motion vectors, accumulate the number of all matching entries, and count the frequency of subsequent video stream frames referencing the unique sequence identifier of the predicted frame to generate the total number of times the video stream predicted frame is referenced.

[0025] As a further aspect of the present invention, the step of obtaining the optical symbol probability shaping mapping table specifically includes:

[0026] S411: When the total number of times the video stream prediction frame is referenced is greater than the preset video frame importance threshold, extract the target frame compression data corresponding to the current video stream prediction frame from the video compression bitstream, determine the modulation order of the standard optical communication constellation diagram that adapts to the target frame compression data, calculate the theoretical maximum information entropy value based on the modulation order, and at the same time calculate the difference between the theoretical maximum information entropy value and the preset probability shaping redundancy adjustment amount as the target information entropy dispersion value;

[0027] S412: Based on the target information entropy dispersion value, the probability shaping and scaling coefficient is derived in reverse, the coordinate mapping relationship between the symbol to be transmitted and each optical communication constellation point in the standard optical communication constellation diagram is established, the position of each optical communication constellation point is determined in the complex plane, the geometric Euclidean distance of each optical communication constellation point relative to the origin of the standard optical communication constellation diagram is calculated, and the geometric distribution characteristic parameters of the constellation points are generated.

[0028] S413: Call the geometric distribution feature parameters of the constellation points, substitute the probability shaping scaling factor and the geometric Euclidean distance into the Maxwell-Boltzmann distribution function, calculate the occurrence probability of each optical communication constellation point under the non-uniform probability distribution, and construct the optical symbol probability shaping mapping table according to the mapping rules of the constellation points based on the occurrence probability.

[0029] As a further aspect of the present invention, the step of obtaining the adaptive optimization result of video streaming transmission quality specifically includes:

[0030] S511: Based on the optical symbol probability shaping mapping table, the target frame compressed data is symbol mapped and optically modulated and transmitted to obtain the real-time optical transmission symbol error rate returned by the optical receiver. When the real-time optical transmission symbol error rate is greater than the preset optical link security baseline, the difference between the real-time optical transmission symbol error rate and the optical link security baseline is calculated to generate the optical transmission symbol error rate over-limit amplitude.

[0031] S512: Calculate the product of the optical transmission interference warning threshold and the optical transmission symbol error rate exceeding the limit, determine the reduction range of the interference threshold, and at the same time calculate the product of the video frame importance threshold and the optical transmission symbol error rate exceeding the limit, determine the reduction range of the importance threshold, as the dynamic reduction correction amount of the dual threshold.

[0032] S513: Based on the dynamic reduction correction amount of the dual thresholds, calculate the difference between the optical transmission interference warning threshold and the corresponding interference warning threshold reduction correction amount, calculate the difference between the video frame importance threshold and the corresponding video frame importance threshold reduction correction amount, and obtain the optimized optical transmission interference warning threshold and the optimized video frame importance threshold respectively, and generate the adaptive optimization result of video stream optical transmission quality.

[0033] An adaptive optimization system for optical transmission quality of video streams, the system comprising:

[0034] The optical interference assessment module collects the instantaneous optical power of the current video stream wavelength channel and the adjacent video stream wavelength channels, calculates the statistical variance of power fluctuation, and calculates the cross-phase modulation interference index in combination with the set fiber nonlinear coefficient.

[0035] The quantitative decoupling configuration module compares the cross-phase modulation interference index with the preset optical transmission interference warning threshold, configures the color quantization parameters of the header information of the current video slice to be encoded according to the comparison result, and sets the color quantization step size decoupling configuration.

[0036] The prediction frame statistics module parses the video compressed bitstream generated based on the color metric step size decoupling configuration encoding, identifies the unique frame number identifier of the current video stream prediction frame in the video sequence, and counts the total number of times the video stream prediction frame is referenced.

[0037] The important frame mapping module compares the total number of times the video stream prediction frame is referenced with a preset video frame importance threshold, extracts the target frame compression data of the current video stream prediction frame based on the comparison result, and constructs an optical symbol probability shaping mapping table in combination with a preset standard optical communication constellation diagram.

[0038] The adaptive transmission optimization module performs symbol mapping and optical modulation transmission on the target frame compressed data based on the optical symbol probability shaping mapping table. It corrects the optical transmission interference warning threshold and the video frame importance threshold according to the transmission results to obtain the adaptive optimization result of video stream optical transmission quality.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0040] In this invention, by constructing a cross-layer mapping mechanism between the nonlinear characteristics of the photophysical layer and the coding parameters of the video application layer, the degree of cross-phase modulation interference caused by the interaction of multi-wavelength channels can be quantified in real time. When the risk of channel degradation is detected, the directional decoupling adjustment of the chromaticity quantization step size is triggered, which suppresses noise interference of non-sensitive chromaticity data while preserving luminance details. At the same time, key reference nodes are accurately identified based on the reference frequency of video frames in the temporal prediction sequence. Probabilistic shaping modulation protection based on information entropy is implemented for high-value data, concentrating limited optical transmission energy to ensure the integrity of the core prediction frame, effectively curbing the global bit error propagation caused by single-frame damage. Combined with threshold dynamic closed-loop correction logic based on real-time symbol error rate, fine-grained adaptive control of video transmission quality is achieved in complex optical link environments, significantly improving the robustness and decoding stability of the communication system under harsh conditions. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the workflow of the present invention;

[0042] Figure 2 This is a flowchart of the process for obtaining the cross-phase modulation interference index in this invention;

[0043] Figure 3 This is a flowchart of the process for obtaining the decoupled configuration of the colorimetric step size in this invention;

[0044] Figure 4 This is a flowchart of the process for obtaining the total number of times a predicted frame in a video stream is referenced in this invention.

[0045] Figure 5 This is a flowchart of the process for obtaining the optical symbol probability shaping mapping table in this invention;

[0046] Figure 6 This is a flowchart illustrating the process of obtaining the adaptive optimization results for video streaming transmission quality in this invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0048] Please see Figure 1 This invention provides a technical solution, an adaptive optimization method for optical transmission quality of video streams, comprising the following steps:

[0049] S1: Collect the instantaneous fiber-injected optical power of the current video stream wavelength channel and adjacent video stream wavelength channels, calculate the power fluctuation statistical variance, and calculate the cross-phase modulation interference index in combination with the set fiber nonlinear coefficient.

[0050] S2: Compare the cross-phase modulation interference index with the preset optical transmission interference warning threshold, configure the colorimetric parameters of the header information of the current video slice to be encoded according to the comparison result, and set the colorimetric step size decoupling configuration.

[0051] S3: Analyze the video compressed bitstream generated by color metric step size decoupling configuration encoding, identify the unique frame number identifier of the current video stream prediction frame in the video sequence, and count the total number of times the video stream prediction frame is referenced.

[0052] S4: Compare the total number of times the video stream prediction frame is referenced with the preset video frame importance threshold, extract the target frame compression data of the current video stream prediction frame based on the comparison result, and construct an optical symbol probability shaping mapping table in combination with the preset standard optical communication constellation diagram.

[0053] S5: Based on the optical symbol probability shaping mapping table, the target frame compressed data is symbol-mapped and optically modulated for transmission. The optical transmission interference warning threshold and the video frame importance threshold are corrected according to the transmission results to obtain the adaptive optimization result of video stream optical transmission quality.

[0054] The cross-phase modulation interference index is specifically obtained by multiplying the power fluctuation statistical variance, the instantaneous fiber input power of the current video stream wavelength channel, and the fiber nonlinear coefficient. The color quantization step size decoupling configuration includes the sum of the original blue differential component quantization parameters and the preset positive integer offset, and the sum of the original red differential component quantization parameters and the preset positive integer offset. The total number of times the video stream prediction frame is referenced is specifically calculated by counting the total number of references of the unique frame number identifier that directly points to the current video stream prediction frame through motion vectors in subsequent video stream frames. The optical symbol probability shaping mapping table includes the coordinate mapping relationship between the symbol to be transmitted and each optical communication constellation point in the standard optical communication constellation diagram and the occurrence probability of each optical communication constellation point. The adaptive optimization results of video stream optical transmission quality include optimizing the optical transmission interference warning threshold and optimizing the video frame importance threshold.

[0055] Please see Figure 2 The specific steps for obtaining the cross-phase modulation interference index are as follows:

[0056] S111: Set a sliding time window based on the video stream symbol rate period, obtain the material refractive index parameters and effective mode area of ​​the optical fiber transmission medium carrying the current video stream wavelength channel, set the optical fiber nonlinear coefficient, collect the instantaneous fiber-injected optical power of the current video stream wavelength channel and the instantaneous fiber-injected optical power of adjacent video stream wavelength channels according to the preset sampling period, and construct the basic physical parameter set of the optical fiber channel.

[0057] First, the symbol rate parameter used by the current video stream is read through the underlying monitoring interface of the optical transmission link. This parameter is usually determined by the clock frequency of the optical transmitter; for example, if the read symbol rate is 64 gigabits (GBd), the process defines a sliding time window containing a fixed number of symbol periods. The duration of this window is set to cover sufficient optical signal fluctuation characteristics, specifically calculated by multiplying the duration of a single symbol period by the number of symbols contained within the window. If the symbol period is 15.6 picoseconds, and a time period covering 1024 symbol periods is selected as the length of the sliding time window, then the duration of the sliding time window is set to 15.6 × 1024 = 15974.4 picoseconds, or approximately 15.97 nanoseconds. Simultaneously, the process accesses the static configuration database of the optical fiber link and queries the optical fiber transmission medium material properties corresponding to the wavelength channel number occupied by the current video stream, specifically including the refractive index parameter of the fiber core and the effective mode area parameter. For example, the query reveals that the refractive index of the fiber core corresponding to the current channel at a wavelength of 1550 nm is 1.46, and the effective mode area is 80 square micrometers. Next, the process calls a preset fiber nonlinearity coefficient configuration. This coefficient is a fixed value obtained through offline fiber characteristic testing experiments; for example, the fiber nonlinearity coefficient is set to 1.3 watts per kilometer. This parameter quantifies the degree of response of the fiber medium to changes in refractive index dependent on light intensity. After completing the setting of static and semi-static parameters, the process initiates a real-time data acquisition link, synchronously monitoring the optical power of the current video stream wavelength channel and adjacent video stream wavelength channels in the physical spectrum according to a preset high-frequency sampling period. The sampling period setting must satisfy the Nyquist sampling theorem and have sufficient oversampling factor to capture transient effects; for example, a sampling period of 0.05 nanoseconds is set. The acquisition device records the instantaneous input fiber optical power value at each sampling moment and adds precise timestamps to these discrete power data. For example, at a specific sampling moment, the instantaneous fiber-to-fiber optical power of the current video stream wavelength channel is 3.5 milliwatts, while the instantaneous fiber-to-fiber optical power of the adjacent video stream wavelength channel is 4.2 milliwatts. All the collected power data, together with the refractive index, mode area, and nonlinear coefficients obtained from the aforementioned query, constitute the set of fundamental physical parameters of the optical fiber channel.

[0058] S112: Map the instantaneous fiber-in optical power of multiple adjacent video stream wavelength channels in the basic physical parameters set of the fiber channel to the sliding time window, and statistically analyze the adjacent channel power time-domain fluctuation variance of the instantaneous fiber-in optical power of each sampling point within the sampling period and the average instantaneous fiber-in optical power within the sliding time window.

[0059] Based on the timestamp information of each sampling point, the power data in the basic physical parameter set of the fiber optic channel is mapped to the aforementioned sliding time window. This mapping operation is essentially a data truncation and alignment process, ensuring that the data involved in the calculation is strictly within the current analysis time interval, i.e., a window of 15.97 nanoseconds. The process performs statistical calculations on the instantaneous fiber-to-the-fiber optical power data of adjacent video stream wavelength channels. First, the process accumulates the instantaneous fiber-to-the-fiber optical power values ​​of adjacent channels for all sampling points within the sliding time window, and divides the sum by the total number of sampling points to calculate the average instantaneous fiber-to-the-fiber optical power within the window. For example, if the sliding time window contains 320 sampling points, and the sum of the power of all sampling points is 1280 milliwatts, then the calculated average instantaneous fiber-to-the-fiber optical power is 1280 ÷ 320 = 4 milliwatts. Subsequently, the process performs point-by-point calculation of the variance characteristics. For each sampling point within the window, the difference between its corresponding instantaneous fiber-injected optical power and the previously calculated average instantaneous fiber-injected optical power is calculated, and this difference is squared to quantify the degree of deviation of that point from the average level. The squaring operation is involved here, and the calculation formula is: Square milliwatts. The process accumulates the squared differences of all sampling points within the window and divides the accumulated result by the total number of sampling points to obtain the adjacent channel power time-domain fluctuation variance of adjacent video stream wavelength channels within the current sliding time window. This variance physically characterizes the severity of optical power jitter in adjacent channels and is a key factor in the strength of cross-phase modulation effects. After statistical calculation of 320 points, the final adjacent channel power time-domain fluctuation variance is 0.25 square milliwatts.

[0060] S113: Calculate the product of the adjacent channel power time-domain fluctuation variance, the instantaneous fiber-injected optical power of the current video stream wavelength channel, and the fiber nonlinear coefficient, as the cross-phase modulation interference index.

[0061] The adjacent channel power time-domain fluctuation variance calculated in step S112 is called upon, and the instantaneous fiber-to-fiber optical power of the current video stream wavelength channel at the current moment, as well as the fiber nonlinear coefficient set in step S111, are read simultaneously. A multiplication operation is performed, multiplying the adjacent channel power time-domain fluctuation variance, the instantaneous fiber-to-fiber optical power of the current video stream wavelength channel, and the fiber nonlinear coefficient. The physical meaning of this calculation logic is that the phase noise caused by cross-phase modulation depends not only on the power fluctuation of adjacent channels (characterized by variance), but also on the degree of excitation of nonlinear effects by the current channel's own light intensity (characterized by instantaneous power) and the nonlinear response capability of the fiber medium itself (characterized by the nonlinear coefficient). For example, the adjacent channel power time-domain fluctuation variance of 0.25 (considered a dimensionless factor after normalization), the instantaneous fiber-to-fiber optical power of the current video stream wavelength channel of 3.5 milliwatts, and the fiber nonlinear coefficient of 1.3 per watt per kilometer are substituted into the calculation. It is important to note that the International System of Units (SI) should be used consistently in the actual calculation process. Therefore, the input optical power needs to be converted from milliwatts to watts, i.e., 3.5mW = 0.0035W. Based on this, if the normalized fluctuation variance factor is 0.25, the power factor is 0.0035, and the medium factor is 1.3, then the calculation process for the cross-phase modulation interference index is 0.25 × 0.0035 × 1.3 = 0.0011375. The calculated result of 0.0011375 is determined as the current cross-phase modulation interference index. This index is a dimensionless or scalar with specific dimensions, and its value range is usually between 0 and positive infinity. The larger the value, the more severe the cross-phase modulation interference caused by adjacent channels in the fiber optic link at the current moment, and the higher the risk of signal transmission quality degradation; the closer the value is to 0, the weaker the interference and the more ideal the channel condition.

[0062] Please see Figure 3 The specific steps for obtaining the colorimetric quantization step size decoupling configuration are as follows:

[0063] S211: When the cross-phase modulation interference index exceeds the preset optical transmission interference warning threshold, configure the chromaticity quantization parameters of the header information for the current video slice to be encoded, read and keep the luminance component quantization parameters unchanged, identify the original blue difference component quantization parameters and the original red difference component quantization parameters in the video slice header information area, and generate a set of chromaticity component parameters to be corrected.

[0064] The real-time calculated cross-phase modulation interference index is compared with a pre-set optical transmission interference warning threshold. This warning threshold is determined based on statistical analysis of the bit error rate (BER) in historical transmission experimental datasets. Specifically, it calculates the probability distribution of the BER exceeding the error correction limit under different interference indices, selecting the average interference index corresponding to the BER spike as the threshold; for example, this threshold is set to 1.0. If the currently calculated cross-phase modulation interference index is 0.0011375, which is less than the set warning threshold of 1.0, it indicates that the cross-phase modulation interference caused by adjacent channels is weak at the current moment, the channel state is relatively ideal, and the parameter adjustment mechanism is not triggered, maintaining the current encoding parameters unchanged. On the other hand, if the calculated cross-phase modulation interference index is greater than the set warning threshold of 1.0 at a certain moment, for example, if the index is 1.1375, then the parameter adjustment mechanism is triggered. The process immediately locates the data stream of the video slice to be encoded, which is currently being prepared for encoding or has been partially encoded, parses its network abstraction layer data structure header, and accurately identifies the video slice header information region. Within this region, the process reads the luminance component quantization parameter and explicitly marks it as remaining unchanged to preserve the video's basic luminance detail. Simultaneously, the process identifies and extracts the original blue and red differential quantization parameters defined in the video slice header information region. These two parameters control the compression fineness of the blue and red chrominance components in the video image, respectively. For example, the parsed original blue differential quantization parameter value is 28, and the original red differential quantization parameter value is 28. These two extracted parameters are combined to generate the set of chrominance component parameters to be corrected.

[0065] S212: Introduce a preset positive integer offset, calculate the sum of the original blue differential quantization parameter and the positive integer offset in the chromaticity component parameter set to be corrected, and at the same time calculate the sum of the original red differential quantization parameter and the positive integer offset to obtain the chromaticity component offset correction value.

[0066] A preset positive integer offset is introduced. The range and specific value of this offset are determined based on experiments on the sensitivity of chromaticity information to human visual perception. By calculating the average perceptible difference (JND) value of color difference perception by the human eye under different chromaticity quantization step sizes, the step size increment with insignificant changes in JND value is selected as the offset. It is usually set to a small positive integer, for example, a positive integer offset of 3. The purpose of the positive offset is to appropriately increase the chromaticity quantization parameters when interference is severe, that is, to increase the compression amplitude of chromaticity information, thereby reducing the bit occupation of chromaticity data and allocating more valuable bandwidth resources to luminance information or error correction coding. At the same time, high-frequency chromaticity noise is also suppressed during the quantization process. The process performs an addition operation on each element in the set of chromaticity component parameters to be corrected. Specifically, the process calculates the sum of the original blue differential component quantization parameters and the positive integer offset, for example, 28 + 3 = 31. At the same time, the process calculates the sum of the original red differential component quantization parameters and the positive integer offset, for example, 28 + 3 = 31. These two newly calculated values, 31 and 31, are defined as chromaticity component offset correction values.

[0067] S213: Summarize the chrominance component offset correction values ​​and luminance component quantization parameters, repackage them according to the video coding standard format, and use them as the chrominance quantization step size decoupling configuration.

[0068] The chroma component offset correction values ​​calculated in step S212 (e.g., the corrected blue differential quantization parameter 31 and red differential quantization parameter 31) are summarized with the luminance component quantization parameters (e.g., kept at 26) confirmed to remain unchanged in step S211. The process strictly follows the syntax specifications of international video coding standards such as H.264 or H.265 / HEVC, re-encapsulating these parameters and writing them into the corresponding syntax element fields in the video slice header. During encapsulation, the checksum or alignment padding bits in the slice header are updated to ensure the validity of the data structure. This set of data containing the original luminance parameters and the corrected chroma parameters is formally defined as the chroma quantization step size decoupling configuration. This configuration is then sent to the core quantization engine of the video encoder, which performs transform coefficient quantization operations on the current video slice according to this new configuration. Due to the increase in the chroma quantization parameters (from 28 to 31), the encoder uses a larger quantization step size when processing chroma residual data, thereby generating fewer coding bits and achieving adaptive bitrate allocation and anti-interference optimization in scenarios with high cross-phase modulation interference.

[0069] Please see Figure 4 The specific steps for obtaining the total number of times a predicted frame in a video stream is referenced are as follows:

[0070] S311: Use color quantization step size decoupling configuration to compress and encode the current video stream, generate a compressed video bitstream, parse the encoding strip header syntax structure of the compressed video bitstream, locate the network abstraction layer that carries the predicted frame data of the current video stream, extract the unique frame number identifier of the predicted frame of the current video stream in the video sequence from the network abstraction layer, and generate a unique sequence identifier code for the predicted frame.

[0071] The current video stream is compressed using a colorimetric quantization step-size decoupling configuration. This process involves transformation, quantization, and entropy coding operations, ultimately generating a compressed video bitstream. The process then initiates bitstream parsing logic, scanning the compressed video bitstream bit by bit to identify and lock the coded stripe header syntax structure. By parsing the type field of the network abstraction layer (BAL) data structure, the process filters out the specific BAL data structure carrying the predicted frame data for the current video stream. Within the internal data region of this structure, the process extracts frame number identifier data to uniquely distinguish the frame within the entire video sequence. This typically corresponds to the "image sequence count" or "frame number" field in the coding standard. The process formats this extracted identifier data to generate a unique sequence identifier code for the predicted frame. For example, if the count of the current predicted frame in the decoding order is 1054, the generated unique sequence identifier code for the predicted frame will be the value 1054. This identifier code is the core index key value for subsequently establishing inter-frame reference relationships.

[0072] S312: Retrieve the slice header information of each subsequent video stream frame after the current video stream prediction frame in the video compressed bitstream, extract the reference image list defined in the slice header information, match and verify the unique sequence identifier code of the prediction frame with each reference image in the reference image list, and filter the reference index dataset of subsequent frames.

[0073] The reference image list includes relative index addressing data based on the image playback order count and absolute position addressing data based on the image playback order count;

[0074] The relative index addressing data based on image playback order counting includes a parameter for calculating the difference between the sequence number of the current subsequent video stream frame and the sequence number of the reference frame stored in the decoded image buffer.

[0075] Absolute position addressing data based on image playback order counting includes the least significant bit of the sequence number or the sequence number corresponding to the unique sequence identifier of the predicted frame.

[0076] In the video compressed bitstream, starting from the current predicted video stream frame, each subsequent video stream frame is retrieved. For each subsequent frame, its slice header information is parsed, and a list of reference images defined therein is extracted. This list records in detail the index information of all reference frames required for the subsequent frame during the decoding prediction process. The process performs a matching verification operation, comparing the unique sequence identifier code of the predicted frame generated in step S311 (e.g., 1054) with each reference image item in the extracted list of reference images. The construction of the reference image list includes two addressing modes: relative index addressing data based on image playback order counting and absolute position addressing data based on image playback order counting. Relative index addressing data locates reference frames by calculating the difference. The process calculates the difference between the sequence number of the current subsequent video stream frame and the sequence number of the reference frame stored in the decoding image buffer. For example, if the current frame number is 1056, the difference parameter is 1056-1054=2, which points to frame number 1054. The absolute location addressing data directly contains the least significant bit of the sequence number or the complete sequence number. The process directly compares this value with the unique sequence identifier of the predicted frame. Through this dual-mode traversal comparison, the process filters out all subsequent frames that actually reference the current predicted frame.

[0077] S313: Match entries in the subsequent frame reference index dataset that directly point to the unique sequence identifier of the predicted frame through motion vectors, accumulate the number of all matched entries, and count the frequency of subsequent video stream frames referencing the unique sequence identifier of the predicted frame to generate the total number of times the predicted frame of the video stream is referenced.

[0078] The process iterates through the selected subsequent frame reference index dataset, deeply analyzing the macroblock or coded block level data of each subsequent frame. It checks each subsequent frame for entries that directly point to the unique sequence identifier of the predicted frame via motion vectors. This means that a specific image block in the subsequent frame is directly copied or interpolated from the corresponding region of the current predicted frame. An accumulator counter is set up, incrementing by one each time a matching entry is found or a valid reference relationship between a subsequent frame and the predicted frame is confirmed. The process counts the total frequency of references to the unique sequence identifier of the predicted frame throughout the entire subsequent video stream frame sequence. For example, if 8 out of the subsequent 12 video frames contain frame number 1054 in their reference lists, and all 8 frames call the data from this frame for motion compensation during the actual decoding process, the accumulated result is 8. This final statistical value is generated as the total number of times the predicted frame in the video stream is referenced. This number directly reflects the importance of the current predicted frame in the temporal correlation of the video sequence—the more times it is referenced, the wider the error propagation range will be if a transmission error occurs, and therefore the higher the importance level of the frame. This number can be 0 (no reference, non-reference frame) or a positive integer; the larger the value, the wider the range of influence.

[0079] Please see Figure 5 The specific steps for obtaining the optical symbol probability integer mapping table are as follows:

[0080] S411: When the total number of times the video stream prediction frame is referenced is greater than the preset video frame importance threshold, extract the target frame compression data corresponding to the current video stream prediction frame from the video compressed bitstream, determine the modulation order of the standard optical communication constellation diagram that adapts to the target frame compression data, calculate the theoretical maximum information entropy value based on the modulation order, and at the same time calculate the difference between the theoretical maximum information entropy value and the preset probability shaping redundancy adjustment amount as the target information entropy dispersion value.

[0081] The total number of times the predicted video stream frame is referenced obtained in step S313 is compared with a preset video frame importance threshold. This threshold is set based on the sensitivity analysis of video quality to bit errors. By calculating the average decrease in peak signal-to-noise ratio (PSNR) due to single-frame loss in historical datasets, the average minimum number of references that causes a PSNR drop exceeding 3dB is determined as the threshold; for example, a video frame importance threshold of 5 is set. If the current total number of references is 8, which is greater than the threshold of 5, the frame is determined to be a key frame. During this process, the target frame compressed data corresponding to the current predicted video stream frame is extracted from the video compressed bitstream, and the modulation order (e.g., 16th order QAM) of the standard optical communication constellation diagram adapted to this data is determined. Based on this, using the modified Shannon information entropy principle, a formula for calculating the target information entropy dispersion value is constructed:

[0082] ;

[0083] in, This represents the target information entropy dispersion value, in bits per symbol, and is the final information entropy target calculated by this method. This represents the modulation order of a standard optical communication constellation diagram, reflecting the theoretical maximum information carrying capacity in the unshaped state. Its value is determined by the modulation format, for example, in 16QAM. ; This represents the basic redundancy ratio coefficient, used to describe the proportion of channel reserved redundancy to the theoretical maximum information entropy. It is a dimensionless parameter, and its value is preset based on the average signal-to-noise ratio (SNR) of the optical channel, determined by analyzing the lower quartile of the historical channel SNR distribution. When the SNR is low, the reserved redundancy ratio should be appropriately increased, and the value range is usually 0.05~0.3, for example, it is set to 0.2 in this embodiment. The importance weighting coefficient is used to adjust the impact of the number of times a video frame is referenced on the degree of information entropy compression. It is a dimensionless parameter. This coefficient is used to control the protection strength of key frames. The larger the value, the stronger the protection of highly referenced important frames. Its value range is usually 0~1. For example, it is set to 0.3 in this embodiment. This indicates the total number of times the video stream prediction frame obtained in step S313 is referenced, for example, 8 in this embodiment; This represents a preset video frame importance threshold used to normalize the reference strength; for example, it is set to 5 in this embodiment. This is a mathematical correction constant used to ensure that the input to the logarithmic function is always greater than 1, thereby avoiding numerical anomalies when the reference ratio is small. Its value is usually 1.

[0084] In this embodiment, the relevant parameters are substituted into the formula for calculation:

[0085] Current modulation order Predict the total number of frame references Video frame importance threshold Basic redundancy ratio coefficient Importance weighting coefficient , corrected constant .

[0086] The target information entropy dispersion value is:

[0087] Bit / Symbol;

[0088] The calculation results show that when enhancing the protection of highly referenced important video frames, the information entropy of the optical signal needs to be compressed from the theoretical maximum of 4 bits / symbol to about 2.05 bits / symbol. This reduces the information density in exchange for higher transmission robustness, thereby improving the transmission reliability of critical video frames in optical communication links.

[0089] S412: Based on the target information entropy dispersion value, the probability shaping and scaling coefficient is derived in reverse. The coordinate mapping relationship between the symbol to be transmitted and each optical communication constellation point in the standard optical communication constellation diagram is established. The position of each optical communication constellation point in the complex plane is determined. The geometric Euclidean distance of each optical communication constellation point relative to the origin of the standard optical communication constellation diagram is calculated. The geometric distribution characteristic parameters of the constellation points are generated.

[0090] Based on the calculated target information entropy dispersion value, the probability shaping scaling coefficient in the Maxwell-Boltzmann distribution is derived in reverse using the entropy function in information theory. This derivation aims to find a specific coefficient such that, under the corresponding probability distribution, the entropy of the constellation diagram is exactly equal to the target value. Subsequently, the process establishes the coordinate mapping relationship between the symbols to be transmitted and each optical communication constellation point in the standard optical communication constellation diagram. In the complex plane, the process determines the real and imaginary coordinate positions of all 16 constellation points in the 16QAM constellation diagram; for example, the coordinates of one point are (1,1), and another point is (3,3). For each constellation point, the process calculates the square of its geometric Euclidean distance relative to the origin (0,0) of the standard optical communication constellation diagram. Specifically, the process calculates the sum of the squares of the real and imaginary coordinates of the point. This involves squaring operations; for the point with coordinates (1,1), the square of its geometric Euclidean distance is calculated as follows: For a point with coordinates (3,3), the squared distance is calculated as follows: These calculated distance values ​​constitute the geometric distribution characteristics of constellation points, revealing the differences in energy levels among different constellation points.

[0091] S413: Call the geometric distribution feature parameters of constellation points, substitute the probability shaping scaling factor and geometric Euclidean distance into the Maxwell-Boltzmann distribution function, calculate the occurrence probability of each optical communication constellation point under non-uniform probability distribution, and construct the optical symbol probability shaping mapping table according to the mapping rules of the constellation points based on the occurrence probability.

[0092] The process calls upon the geometric distribution feature parameters of the constellation points generated in step S412 and the probability scaling coefficients obtained through reverse calculation. Using the Maxwell-Boltzmann distribution function and normalization logic, the formula for calculating the probability of constellation point occurrence is constructed: In the formula, Indicates the first The probability of an optical communication constellation point appearing under a non-uniform probability distribution is the target of calculation. This represents the probability scaling factor, which determines the steepness of the probability distribution. The larger the factor, the more severely the probability of high-energy points is suppressed. The value range is a real number greater than 0. Here, we assume the inverse value is 0.3. Representing the The geometric Euclidean distance between points of the optical communication constellation (the square of the distance is taken as the energy term, obtained from S412); The first step in the summation traversal process The geometric Euclidean distance between the constellation points; This represents the total number of optical communication constellation points in the standard optical communication constellation diagram, and its value is 16 for 16QAM. This represents the natural exponential function. Substituting the parameters into the formula to perform the calculation: assuming the probability is obtained by reverse calculation using an integer scaling factor. The value is 0.3. For a low-energy point A (coordinates 1, 1) near the origin, the squared distance is... The value is 2; for a high-energy point B (coordinates 3, 3) far from the origin, the square of its distance is 2. The value is 18. Assume that after iterating and accumulating, the sum of the exponents of all points (i.e., the denominator) results in 4.5. The process for calculating the probability of point A is as follows: Calculate the probability of point B: Based on these precisely calculated occurrence probabilities, the process reallocates the constellation point mapping rules, constructing a light symbol probability shaping mapping table. This mapping table uses the logic of Huffman coding or a distribution matching algorithm to map frequently occurring bit combinations in the input bitstream to high-probability (low-energy) constellation points, while mapping less frequently occurring bit combinations to low-probability (high-energy) constellation points.

[0093] Please see Figure 6 The specific steps for obtaining the adaptive optimization results of video streaming transmission quality are as follows:

[0094] S511: Based on the optical symbol probability shaping mapping table, the target frame compressed data is symbol mapped and optically modulated for transmission. The real-time optical transmission symbol error rate returned by the optical receiver is obtained. When the real-time optical transmission symbol error rate is greater than the preset optical link security baseline, the difference between the real-time optical transmission symbol error rate and the optical link security baseline is calculated to generate the optical transmission symbol error rate over-limit amplitude.

[0095] Based on the constructed optical symbol probability shaping mapping table, symbol mapping is performed on the target frame compressed data of the current video stream, driving the optical modulator to generate optical signals and transmit them into the optical fiber link. At the optical receiver, the receiving device demodulates and makes decisions on the signal, and uses the forward error correction code verification mechanism to count the number of erroneous symbols during transmission, thereby obtaining the real-time optical transmission symbol error rate. For example, if the process counts 200 erroneous symbols during the transmission of 1 million symbols, then the real-time optical transmission symbol error rate is 200 ÷ 1,000,000 = 0.0002. The process then determines whether this real-time optical transmission symbol error rate is greater than the preset optical link safety baseline. The optical link safety baseline is set based on the video decoder's tolerance limit test, calculated by averaging the critical value of the bit error rate that would cause the decoder to crash or produce severe screen distortion, and determined after reserving a safety margin, for example, set to 0.0001. If the real-time symbol error rate of 0.0002 is greater than the safety baseline of 0.0001, then the transmission quality is deemed to have exceeded the limit. The process calculates the difference between the real-time optical transmission symbol error rate and the optical link security baseline, for example, 0.0002 - 0.0001 = 0.0001. This difference is generated as the optical transmission symbol error rate exceedance, which quantifies the specific degree of current link quality degradation.

[0096] S512: Calculate the product of the optical transmission interference warning threshold and the optical transmission symbol error rate exceeding the limit to determine the reduction range of the interference threshold. At the same time, calculate the product of the video frame importance threshold and the optical transmission symbol error rate exceeding the limit to determine the reduction range of the importance threshold, which serves as the dynamic reduction correction amount for the dual thresholds.

[0097] The decision threshold of the control logic is dynamically tightened based on the error magnitude. The process reads the current optical transmission interference warning threshold (e.g., the previously set 1.0) and the calculated optical transmission symbol error rate exceeding the limit (e.g., 0.0001). The process needs to correlate these two parameters with different dimensions, usually by introducing a normalized sensitivity coefficient (e.g., setting the coefficient to 1000 for magnitude matching, this coefficient is set based on the convergence speed requirements of the control loop), and calculates the product of the optical transmission interference warning threshold and the optical transmission symbol error rate exceeding the limit (after coefficient weighting) to determine the reduction amount of the interference threshold. For example, calculating 1.0 × 0.0001 × 1000 = 0.1 means that the interference threshold needs to be reduced by 0.1. At the same time, the process calculates the product of the video frame importance threshold (e.g., the previously set 5) and the optical transmission symbol error rate exceeding the limit (again, combined with the corresponding weighting coefficient, e.g., 4000, this coefficient is tuned through step response experiments) to determine the reduction amount of the importance threshold. For example, calculating 5 × 0.0001 × 4000 = 2 means the importance threshold needs to be lowered by 2. These two calculation results—0.1 and 2—are determined as the dynamic reduction correction amounts for the dual threshold. This calculation logic reflects the principle of negative feedback control: the higher the bit error rate, the greater the reduction in the judgment threshold, causing more frames to be judged as interfered or important frames, thereby triggering more aggressive protection measures.

[0098] S513: Based on the dynamic reduction correction amount of dual thresholds, calculate the difference between the optical transmission interference warning threshold and the corresponding interference warning threshold reduction correction amount, calculate the difference between the video frame importance threshold and the corresponding video frame importance threshold reduction correction amount, obtain the optimized optical transmission interference warning threshold and the optimized video frame importance threshold respectively, and generate the adaptive optimization result of video stream optical transmission quality.

[0099] Based on the calculated dual threshold dynamic reduction correction amount, the final threshold update operation is performed. The process calculates the difference between the current optical transmission interference warning threshold (1.0) and the corresponding interference warning threshold reduction correction amount (0.1), and obtains the optimized optical transmission interference warning threshold as 1.0-0.1=0.9. This means that in subsequent processing, as long as the interference index exceeds 0.9, the QP adjustment in S211 will be triggered, the anti-interference response will be more sensitive, and the colorimetric step size will be increased in advance, reserving more error correction bandwidth for transmission. At the same time, the process calculates the difference between the current video frame importance threshold (5) and the corresponding video frame importance threshold reduction correction amount (2), and obtains the optimized video frame importance threshold as 5-2=3. This means that more frames with fewer citations (a citation count greater than 3 is considered important) will also be included in the protection range of probability shaping. The two new values ​​obtained respectively - 0.9 and 3 - are generated as the adaptive optimization result of video stream optical transmission quality and are immediately written into the control register as a new judgment standard for the next frame of video data processing. This result serves a dual purpose: firstly, by lowering the interference judgment threshold (0.9), the video stream can initiate source-end load reduction protection when faced with minor physical layer disturbances; secondly, by lowering the importance judgment threshold (3), the frame set range for optical layer probability shaping protection is expanded, covering reference frames of moderate importance. This adaptive adjustment sacrifices some chroma detail and transmission spectral efficiency in exchange for higher signal error correction capability and video decoding stability under adverse conditions of increased bit error rate (0.0002), effectively curbing macroblock errors and video stuttering caused by bit error propagation, and realizing adaptive closed-loop optimization of video stream transmission quality in dynamic optical link environments.

[0100] An adaptive optimization system for optical transmission quality of video streams, the system comprising:

[0101] The optical interference assessment module collects the instantaneous optical power of the current video stream wavelength channel and the adjacent video stream wavelength channels, calculates the statistical variance of power fluctuation, and calculates the cross-phase modulation interference index in combination with the set fiber nonlinear coefficient.

[0102] The quantitative decoupling configuration module compares the cross-phase modulation interference index with the preset optical transmission interference warning threshold, configures the colorimetric parameters of the header information of the current video slice to be encoded based on the comparison result, and sets the colorimetric step size decoupling configuration.

[0103] The prediction frame statistics module parses the video compressed bitstream generated by color metric step size decoupling configuration encoding, identifies the unique frame number identifier of the current video stream prediction frame in the video sequence, and counts the total number of times the video stream prediction frame is referenced.

[0104] The important frame mapping module compares the total number of times the predicted video stream frame is referenced with the preset video frame importance threshold. Based on the comparison result, it extracts the target frame compression data of the current video stream predicted frame and constructs an optical symbol probability shaping mapping table in combination with the preset standard optical communication constellation diagram.

[0105] The adaptive transmission optimization module performs symbol mapping and optical modulation transmission on the target frame compressed data based on the optical symbol probability shaping mapping table. It corrects the optical transmission interference warning threshold and the video frame importance threshold according to the transmission results, and obtains the adaptive optimization result of video stream optical transmission quality.

[0106] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An adaptive optimization method for optical transmission quality of video streams, characterized in that, Includes the following steps: S1: Collect the instantaneous fiber-injected optical power of the current video stream wavelength channel and adjacent video stream wavelength channels, calculate the power fluctuation statistical variance, and calculate the cross-phase modulation interference index in combination with the set fiber nonlinear coefficient. S2: Compare the cross-phase modulation interference index with the preset optical transmission interference warning threshold, configure the colorimetric parameters of the header information for the current video slice to be encoded according to the comparison result, and set the colorimetric step size decoupling configuration. S3: Analyze the video compressed bitstream generated based on the color metric step size decoupling configuration encoding, identify the unique frame number identifier of the current video stream prediction frame in the video sequence, and count the total number of times the video stream prediction frame is referenced. S4: Compare the total number of times the video stream prediction frame is referenced with the preset video frame importance threshold, extract the target frame compression data of the current video stream prediction frame based on the comparison result, and construct an optical symbol probability shaping mapping table in combination with the preset standard optical communication constellation diagram. S5: Based on the optical symbol probability shaping mapping table, the target frame compressed data is symbol-mapped and optically modulated for transmission. The optical transmission interference warning threshold and the video frame importance threshold are corrected according to the transmission results to obtain the adaptive optimization result of video stream optical transmission quality. The specific steps for obtaining the cross-phase modulation interference index are as follows: S111: Set a sliding time window based on the video stream symbol rate period, obtain the material refractive index parameters and effective mode area of ​​the optical fiber transmission medium carrying the current video stream wavelength channel, set the optical fiber nonlinear coefficient, collect the instantaneous fiber-injected optical power of the current video stream wavelength channel and the instantaneous fiber-injected optical power of adjacent video stream wavelength channels according to the preset sampling period, and construct the basic physical parameter set of the optical fiber channel. S112: Map the instantaneous fiber-in optical power of multiple adjacent video stream wavelength channels in the optical fiber channel basic physical parameter set to a sliding time window, and statistically analyze the adjacent channel power time-domain fluctuation variance between the instantaneous fiber-in optical power of each sampling point within the sampling period and the average instantaneous fiber-in optical power within the sliding time window. S113: Calculate the product of the adjacent channel power time-domain fluctuation variance, the instantaneous fiber-injected optical power of the current video stream wavelength channel, and the fiber nonlinear coefficient, as the cross-phase modulation interference index; The specific steps for obtaining the colorimetric quantization step size decoupling configuration are as follows: S211: When the cross-phase modulation interference index exceeds the preset optical transmission interference warning threshold, configure the chromaticity quantization parameters of the header information for the current video slice to be encoded, read and keep the luminance component quantization parameters unchanged, identify the original blue difference component quantization parameters and the original red difference component quantization parameters in the video slice header information area, and generate a set of chromaticity component parameters to be corrected. S212: Introduce a preset positive integer offset, calculate the sum of the original blue differential quantization parameter and the positive integer offset in the set of chromaticity component parameters to be corrected, and at the same time calculate the sum of the original red differential quantization parameter and the positive integer offset to obtain the chromaticity component offset correction value. S213: Summarize the chroma component offset correction values ​​and the luminance component quantization parameters, repackage them according to the format of the video coding standard, and use them as the chroma quantization step size decoupling configuration.

2. The adaptive optimization method for optical transmission quality of video streams according to claim 1, characterized in that, The cross-phase modulation interference index is specifically obtained by multiplying the power fluctuation statistical variance, the instantaneous fiber input power of the current video stream wavelength channel, and the fiber nonlinear coefficient. The colorimetric quantization step size decoupling configuration includes the sum of the original blue differential quantization parameters and the preset positive integer offset, and the sum of the original red differential quantization parameters and the preset positive integer offset. The total number of times the video stream prediction frame is referenced is specifically calculated by counting the total number of references in subsequent video stream frames that directly point to the unique frame number identifier of the current video stream prediction frame through motion vectors. The optical symbol probability shaping mapping table includes the coordinate mapping relationship between the symbol to be transmitted and each optical communication constellation point in the standard optical communication constellation diagram and the occurrence probability of each optical communication constellation point. The adaptive optimization result of video stream optical transmission quality includes optimizing the optical transmission interference warning threshold and optimizing the video frame importance threshold.

3. The adaptive optimization method for optical transmission quality of video streams according to claim 1, characterized in that, The steps for obtaining the total number of times the predicted video stream frame is referenced are as follows: S311: Compress and encode the current video stream using the color metric step size decoupling configuration to generate a compressed video stream, parse the encoding strip header syntax structure of the compressed video stream, locate the network abstraction layer that carries the predicted frame data of the current video stream, extract the unique frame number identifier of the predicted frame of the current video stream in the video sequence from the network abstraction layer, and generate a unique sequence identifier code for the predicted frame. S312: Retrieve the slice header information of each subsequent video stream frame located after the current video stream prediction frame in the video compressed bitstream, extract the reference image list defined in the slice header information, match and verify the unique sequence identifier code of the prediction frame with each reference image in the reference image list, and filter the reference index dataset of subsequent frames. S313: Match entries in the subsequent frame reference index dataset that directly point to the unique sequence identifier of the predicted frame through motion vectors, accumulate the number of all matching entries, and count the frequency of subsequent video stream frames referencing the unique sequence identifier of the predicted frame to generate the total number of times the video stream predicted frame is referenced.

4. The adaptive optimization method for optical transmission quality of video streams according to claim 3, characterized in that, The specific steps for obtaining the optical symbol probability shaping mapping table are as follows: S411: When the total number of times the video stream prediction frame is referenced is greater than the preset video frame importance threshold, extract the target frame compression data corresponding to the current video stream prediction frame from the video compression bitstream, determine the modulation order of the standard optical communication constellation diagram that adapts to the target frame compression data, calculate the theoretical maximum information entropy value based on the modulation order, and at the same time calculate the difference between the theoretical maximum information entropy value and the preset probability shaping redundancy adjustment amount as the target information entropy dispersion value; S412: Based on the target information entropy dispersion value, the probability shaping and scaling coefficient is derived in reverse, the coordinate mapping relationship between the symbol to be transmitted and each optical communication constellation point in the standard optical communication constellation diagram is established, the position of each optical communication constellation point is determined in the complex plane, the geometric Euclidean distance of each optical communication constellation point relative to the origin of the standard optical communication constellation diagram is calculated, and the geometric distribution characteristic parameters of the constellation points are generated. S413: Call the geometric distribution feature parameters of the constellation points, substitute the probability shaping scaling factor and the geometric Euclidean distance into the Maxwell-Boltzmann distribution function, calculate the occurrence probability of each optical communication constellation point under the non-uniform probability distribution, and construct the optical symbol probability shaping mapping table according to the mapping rules of the constellation points based on the occurrence probability.

5. The adaptive optimization method for optical transmission quality of video streams according to claim 4, characterized in that, The specific steps for obtaining the adaptive optimization results of video streaming transmission quality are as follows: S511: Based on the optical symbol probability shaping mapping table, the target frame compressed data is symbol mapped and optically modulated and transmitted to obtain the real-time optical transmission symbol error rate returned by the optical receiver. When the real-time optical transmission symbol error rate is greater than the preset optical link security baseline, the difference between the real-time optical transmission symbol error rate and the optical link security baseline is calculated to generate the optical transmission symbol error rate over-limit amplitude. S512: Calculate the product of the optical transmission interference warning threshold and the optical transmission symbol error rate exceeding the limit, determine the reduction range of the interference threshold, and at the same time calculate the product of the video frame importance threshold and the optical transmission symbol error rate exceeding the limit, determine the reduction range of the importance threshold, as the dynamic reduction correction amount of the dual threshold. S513: Based on the dynamic reduction correction amount of the dual thresholds, calculate the difference between the optical transmission interference warning threshold and the corresponding interference warning threshold reduction correction amount, calculate the difference between the video frame importance threshold and the corresponding video frame importance threshold reduction correction amount, and obtain the optimized optical transmission interference warning threshold and the optimized video frame importance threshold respectively, and generate the adaptive optimization result of video stream optical transmission quality.

6. The adaptive optimization method for optical transmission quality of video streams according to claim 3, characterized in that, The reference image list includes relative index addressing data based on the image playback order count and absolute position addressing data based on the image playback order count; The relative index addressing data based on image playback order counting includes a parameter for calculating the difference between the sequence number of the current subsequent video stream frame and the sequence number of the reference frame stored in the decoded image buffer. The absolute position addressing data based on the image playback order count includes the least significant bit of the sequence number or the sequence number corresponding to the unique sequence identifier of the predicted frame.

7. The adaptive optimization method for optical transmission quality of video streams according to claim 4, characterized in that, The formula for the target information entropy dispersion value is as follows: ； In the formula, This represents the target information entropy dispersion value. This indicates the modulation order of a standard optical communication constellation diagram. This represents the basic redundancy ratio coefficient. This represents the importance weighting coefficient. This indicates the total number of times the predicted frame of the video stream is referenced. This represents the preset video frame importance threshold. This is a mathematical correction constant.

8. An adaptive optimization system for optical transmission quality of video streams, characterized in that, The adaptive optimization method for optical transmission quality of video streams according to any one of claims 1-7, wherein the system comprises: The optical interference assessment module collects the instantaneous optical power of the current video stream wavelength channel and the adjacent video stream wavelength channels, calculates the statistical variance of power fluctuation, and calculates the cross-phase modulation interference index in combination with the set fiber nonlinear coefficient. The quantitative decoupling configuration module compares the cross-phase modulation interference index with the preset optical transmission interference warning threshold, configures the color quantization parameters of the header information of the current video slice to be encoded according to the comparison result, and sets the color quantization step size decoupling configuration. The prediction frame statistics module parses the video compressed bitstream generated based on the color metric step size decoupling configuration encoding, identifies the unique frame number identifier of the current video stream prediction frame in the video sequence, and counts the total number of times the video stream prediction frame is referenced. The important frame mapping module compares the total number of times the video stream prediction frame is referenced with a preset video frame importance threshold, extracts the target frame compression data of the current video stream prediction frame based on the comparison result, and constructs an optical symbol probability shaping mapping table in combination with a preset standard optical communication constellation diagram. The adaptive transmission optimization module performs symbol mapping and optical modulation transmission on the target frame compressed data based on the optical symbol probability shaping mapping table. It corrects the optical transmission interference warning threshold and the video frame importance threshold according to the transmission results to obtain the adaptive optimization result of video stream optical transmission quality.

Images