Low-delay wireless screen projection dynamic spectrum allocation method based on star flash technology

Abstract

The invention discloses a low-delay wireless screen projection dynamic spectrum allocation method based on a star flash technology, and relates to the technical field of wireless video transmission. According to the low-time-delay wireless screen projection dynamic spectrum allocation method based on the star flash technology, when wireless screen projection is carried out on a video, communication interference data of multiple candidate frequency bands are input into a pre-training reinforcement learning model in real time to be analyzed to obtain a predicted transmission benefit index, and an optimal frequency band is selected as a current screen projection frequency band; according to the invention, the Polar code coding processing is executed based on the coding regulation and control index to obtain the coding data volume value, and the dynamic wireless transmission of the image data is completed in combination with the star flash protocol stack, so that the dynamic wireless transmission of the image data is realized, and the coding regulation and control index is generated based on the combination of the prediction transmission benefit index and the analysis result. Therefore, the stability of the high-complexity frame in the transmission process is remarkably improved, and the transmission efficiency and the presentation quality of the high-resolution video in the dynamic spectrum environment are improved.

Description

Technical Field

[0001] The present invention relates to the field of wireless video transmission technology, and specifically to a low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology. Background Art

[0002] As 4K / 8K ultra-high-definition video gradually becomes the mainstream display format, traditional screen projection technologies based on wireless protocols such as WiFi and Bluetooth have exposed obvious problems such as severe spectrum interference, unstable link delay, image freeze or frame loss during real-time video transmission with high bit rate and high frame rate. Especially in the environment of concurrent communication of multiple terminals, traditional screen projection solutions lack a dynamic matching mechanism between channel interference, device load and image complexity, and are difficult to meet the actual needs of low-latency, strong stability and high-fidelity transmission of high-definition video. As a new generation of short-range high-speed wireless communication protocol, Star Flash has the natural advantages of low power consumption, micro-latency, high concurrency and multi-band schedulability, providing a protocol basis for solving the wireless real-time transmission of high-resolution video streams in complex environments.

[0003] The limitations of the existing technology include at least the following problems. In the existing technology, the structural characteristics of the image frame are usually only used for compression in the local encoding stage, and the selection of the transmission channel depends on fixed rules, such as RSSI, packet loss rate or average bandwidth. The two are separated from each other in the system architecture and lack a linkage mechanism. Especially in scenarios with extremely high image loads such as 4K / 8K high-definition video, the existing technology sends highly complex frames directly to the default or nearest channel without analyzing whether the channel steady-state capability matches the image complexity, which can easily lead to transmission failure, frame retransmission or delay fluctuations. In addition, due to the lack of image load-driven encoding strategy adjustment capabilities, the existing technology cannot form a closed loop between channel switching, resource scheduling and encoding control, the scoring system is fragmented, and it is difficult to achieve dynamic resource allocation based on frame adaptation. As the complexity of the projection content increases and the sensitivity to transmission delay intensifies, it is easy to cause the transmission delay to be difficult to compress and the picture quality to fluctuate frequently during wireless projection, which in turn seriously affects the real-time transmission experience of high-resolution video. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides a low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology, which solves the problem that the existing technology is difficult to achieve coordinated matching of image complexity and channel capabilities, resulting in large delay and unstable image quality in high-resolution screen projection.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0006] A low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology includes the following steps: when wirelessly projecting a video, obtaining communication interference data of several candidate frequency bands in real time, and inputting the data into a pre-trained reinforcement learning model for predictive analysis to obtain a predicted transmission efficiency index for each candidate frequency band; performing a comprehensive analysis on the predicted transmission efficiency index of each candidate frequency band to obtain the current screen projection frequency band; obtaining image data of the current screen projection frame of the current screen projection frequency band, and performing feature analysis to obtain a screen projection visual complexity perception index of the current screen projection frequency band, and performing a comprehensive analysis in combination with the predicted transmission efficiency index of the current screen projection frequency band to obtain a coding control index of the current screen projection frame of the current screen projection frequency band; performing adaptive Polar code encoding processing on the image data of the current screen projection frame of the current screen projection frequency band based on the coding control index to obtain the coding data volume value of the current screen projection frequency band, and dynamically wirelessly transmitting the image data of the encoded current screen projection frame in combination with the Star Flash protocol stack.

[0007] Furthermore, the communication interference data includes an interference power intensity index, a frequency band electromagnetic field amplitude, a frequency band modulation index, a signal-to-noise ratio value, and an intermodulation interference intensity index. The reinforcement learning model is specifically a deep neural network, and the deep neural network includes an input layer, several fully connected layers, and an output layer. The specific steps for obtaining the predicted transmission efficiency index of each candidate frequency band are as follows: in the input layer of the deep neural network, the communication interference data of each candidate frequency band is received and preprocessed; in the fully connected layer of the deep neural network, the communication interference data of each candidate frequency band after preprocessing is subjected to multi-layer feature extraction processing to obtain a high-dimensional feature vector of each candidate frequency band; in the output layer of the deep neural network, the high-dimensional feature vector of each candidate frequency band is predicted to obtain a predicted transmission efficiency index for each candidate frequency band.

[0008] Furthermore, the specific steps to obtain the current projection frequency band are as follows: read the predicted transmission efficiency index of each candidate frequency band, and arrange them in descending order to generate a candidate frequency band priority table; mark the candidate frequency band in the first sequence in the candidate frequency band priority table as the current projection frequency band.

[0009] Furthermore, the current projection frame image data is specifically the pixel value and two-dimensional coordinates of each pixel point in the current projection frame image. The specific steps for obtaining the projection visual complexity perception index of the current projection frequency band are as follows: comprehensively analyze the pixel value and two-dimensional coordinates of each pixel point in the current projection frame image of the current projection frequency band to obtain the current image evaluation set of the current projection frame of the current projection frequency band, including the structural edge transition density index, the texture structure cross-intensity index, and the color jump aggregation index; comprehensively analyze the current image evaluation set of the current projection frame of the current projection frequency band based on the particle swarm optimization algorithm to obtain the projection visual complexity perception index of the current projection frame of the current projection frequency band.

[0010] Furthermore, the specific formula for calculating the coding control index of the current projection frame in the current projection frequency band is as follows: Among them, BmT is the coding control index of the current projection frame of the current projection frequency band, TsF is the projection visual complexity perception index of the current projection frame of the current projection frequency band, α1 is the visual adjustment coefficient stored in the database, YcX is the predicted transmission efficiency index of the current projection frequency band, α2 is the transmission efficiency adjustment coefficient stored in the database, and α3 is the interaction adjustment coefficient stored in the database.

[0011] Furthermore, the specific steps for obtaining the coded data volume value of the current projection frame of the current projection frequency band are as follows: read the coding control index of the current projection frame of the current projection frequency band, and perform a comprehensive analysis in combination with the preset minimum coding rate to obtain the Polar coding rate of the current projection frame of the current projection frequency band; perform Polar code encoding on the image data of the current projection frame of the current projection frequency band based on the Polar coding rate, and generate a bit coding sequence of the current projection frame of the current projection frequency band; perform statistical analysis on the bit coding sequence of the current projection frame of the current projection frequency band to obtain the coded data volume value of the current projection frame of the current projection frequency band.

[0012] Furthermore, the specific steps of dynamically wirelessly transmitting the image data of the current projection frame after encoding processing in combination with the Star Flash protocol stack are as follows: obtaining the channel behavior structure data of several transmission channels in the Star Flash protocol stack under the current projection frequency band, and performing a comprehensive analysis to obtain the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band; reading the encoded data volume value and projection visual complexity perception index of the current projection frame in the current projection frequency band, and performing a comprehensive analysis in combination with the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band to obtain the image negative collaborative adaptation index of each transmission channel in the Star Flash protocol stack under the current projection frequency band; wirelessly transmitting the image data of the current projection frame after encoding processing based on the image negative collaborative adaptation index of each transmission channel in the Star Flash protocol stack under the current projection frequency band.

[0013] Furthermore, the channel behavior structure data includes the scheduling slot rate value, the time slot occupancy rate value, the network scheduling stability index, the network conflict pressure index, and the network occupancy saturation index. The specific steps for obtaining the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band are as follows: perform a comprehensive analysis on the channel behavior structure data of each transmission channel in the Star Flash protocol stack under the current projection frequency band to obtain the transmission evaluation set of each transmission channel in the Star Flash protocol stack under the current projection frequency band, including the resource configuration structure index and the channel load interference index; perform a comprehensive analysis on the transmission evaluation set of each transmission channel in the Star Flash protocol stack under the current projection frequency band to obtain the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band.

[0014] Furthermore, the specific steps for obtaining the transmission evaluation set of each transmission channel in the Star Flash protocol stack under the current projection frequency band are as follows: read the scheduling slot rate value, time slot occupancy value, and network scheduling stability index of each transmission channel in the Star Flash protocol stack under the current projection frequency band, and perform a comprehensive analysis in combination with the Bayesian confidence perception method to obtain the resource configuration structure index of each transmission channel in the Star Flash protocol stack under the current projection frequency band; read the network conflict pressure index and network occupancy saturation index of each transmission channel in the Star Flash protocol stack under the current projection frequency band, and perform a comprehensive analysis in combination with the Bayesian confidence perception method to obtain the channel load interference index of each transmission channel in the Star Flash protocol stack under the current projection frequency band.

[0015] Furthermore, the specific formula for calculating the image-negative cooperative transmission index of a transmission channel in the StarFlash protocol stack under the current projection frequency band is as follows: Among them, TxS is the image negative collaborative adaptation index of a transmission channel in the Star Flash protocol stack under the current projection frequency band, BaT is the coded data volume value of the current projection frame of the current projection frequency band, β1 is the coded volume adjustment coefficient stored in the database, TsF is the projection visual complexity perception index of the current projection frame of the current projection frequency band, β2 is the complex perception adjustment coefficient stored in the database, CsW is the transmission steady-state response index of a transmission channel in the Star Flash protocol stack under the current projection frequency band, β3 is the transmission steady-state adjustment coefficient stored in the database, and β4 is the projection visual buffer enhancement adjustment coefficient stored in the database.

[0016] The present invention has the following beneficial effects:

[0017] (1) The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology constructs a negative image collaborative adaptation index, integrates the coded data volume value of each frame of the image, the visual complexity perception index under particle swarm optimization, and the transmission steady-state response index of the corresponding transmission channel, thereby realizing a quantitative adaptation mechanism between the image structure load and the channel structure capability, and then dynamically selects the optimal channel to execute the scheduling transmission of the frame, ensuring that the image load and the channel capability are accurately matched. For example, the image jump degree and the channel resource availability can be combined for judgment, so that highly complex image frames can be preferentially allocated to channels with stronger load bearing capacity and more stable structure for transmission, thereby significantly improving the stability of highly complex frames during transmission and reducing the phenomena of frame misordering, delay fluctuation and screen tearing caused by channel mismatch, so that it is suitable for frame-level resource optimization scheduling in high-density video tasks.

[0018] (2) The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology introduces a prediction model driven by reinforcement learning and builds a deep neural network structure based on communication interference data to dynamically predict the transmission efficiency index of each candidate frequency band, thereby improving the intelligence level of frequency band resource scheduling. The neural network model receives real-time interference indicators of multiple dimensions in the input layer, such as interference power intensity, electromagnetic amplitude, modulation index, signal-to-noise ratio and intermodulation interference index, extracts high-dimensional feature vectors through a multi-layer fully connected structure, and performs regression prediction on the transmission efficiency in the output layer, so that it can adaptively learn the interference patterns in different screen projection environments and output accurate frequency band availability score results, thereby significantly improving the rationality of frequency band selection and the reliability of screen projection path, and then avoiding the impact of poor-quality frequency bands on video transmission stability.

[0019] (3) The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology realizes on-demand adjustment of the image frame coding intensity by constructing a nonlinear adaptation mechanism between the coding control index and the Polar code coding rate, thereby avoiding resource redundancy or insufficient compression under a fixed coding strategy. During the screen projection process, a coding control index is generated based on a joint analysis of the visual complexity perception index and the frequency band transmission efficiency index of the current frame, and the Polar code coding rate is adjusted accordingly. This automatically reduces coding redundancy when the image complexity is low, and increases coding intensity when the image jumps violently or the texture is dense, so as to ensure a balance between the restoration of image details and the ability to resist error. Combined with subsequent channel adaptation, adaptive coding and resource collaborative scheduling for the content characteristics of each frame are realized, and the transmission efficiency and presentation quality of high-resolution video in a dynamic spectrum environment are greatly improved.

[0020] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of a low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology in the present invention.

[0022] Figure 2 This is a flowchart of the specific steps for obtaining the predicted transmission efficiency index of each candidate frequency band in a low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology of the present invention.

[0023] Figure 3 This is a flowchart of the specific steps of the low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology of the present invention, which combines the Star Flash protocol stack to dynamically wirelessly transmit the image data of the current projection frame after encoding processing. DETAILED DESCRIPTION

[0024] See also Figure 1, an embodiment of the present invention provides a technical solution:

[0025] A low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology includes the following steps: when wirelessly projecting a video, obtaining communication interference data of several candidate frequency bands in real time, and inputting the data into a pre-trained reinforcement learning model for predictive analysis to obtain a predicted transmission efficiency index for each candidate frequency band; performing a comprehensive analysis on the predicted transmission efficiency index of each candidate frequency band to obtain the current screen projection frequency band; obtaining image data of the current screen projection frame of the current screen projection frequency band, and performing feature analysis to obtain a screen projection visual complexity perception index of the current screen projection frequency band, and performing a comprehensive analysis in combination with the predicted transmission efficiency index of the current screen projection frequency band to obtain a coding control index of the current screen projection frame of the current screen projection frequency band; performing adaptive Polar code encoding processing on the image data of the current screen projection frame of the current screen projection frequency band based on the coding control index to obtain the coding data volume value of the current screen projection frequency band, and dynamically wirelessly transmitting the image data of the encoded current screen projection frame in combination with the Star Flash protocol stack.

[0026] The specific formula for calculating the coding control index of the current projection frame in the current projection frequency band is as follows: Among them, BmT is the coding control index of the current projection frame of the current projection frequency band, TsF is the projection visual complexity perception index of the current projection frame of the current projection frequency band, α1 is the visual adjustment coefficient stored in the database, YcX is the predicted transmission efficiency index of the current projection frequency band, α2 is the transmission efficiency adjustment coefficient stored in the database, and α3 is the interaction adjustment coefficient stored in the database.

[0027] What needs to be explained is that the specific form of the tanh function is: Here, e is a natural constant and can be 2.71 in this embodiment, with a domain of (-∞, +∞) and a range of (-1, +1).

[0028] α1, α2, and α3 can be obtained through the following steps: using historical data, combined with the projection visual complexity perception index and the predicted transmission efficiency index, to conduct statistical regression analysis, quantify the specific impact of each factor on the coding control index, and thus fit the initial weight value. Secondly, using the sensitivity analysis method, adjust the value range of each coefficient, observe its impact on the coding control evaluation results, and ensure the stability and rationality of the model.

[0029] Specifically, if Figure 2 As shown in the figure, the communication interference data includes the interference power intensity index, the frequency band electromagnetic field amplitude, the frequency band modulation index, the signal-to-noise ratio value, and the intermodulation interference intensity index. The reinforcement learning model is specifically a deep neural network, which includes an input layer, several fully connected layers, and an output layer. The specific steps for obtaining the predicted transmission efficiency index for each candidate frequency band are as follows:

[0030] In the input layer of the deep neural network, the communication interference data of each candidate frequency band is received and preprocessed. That is, the interference power intensity index, frequency band electromagnetic field amplitude, frequency band modulation value, signal-to-noise ratio value, and intermodulation interference intensity index of each candidate frequency band are standardized and converted to a range between 0 and 1 to ensure the consistency of the input of each channel of the neural network;

[0031] In the fully connected layer of the deep neural network, multi-layer feature extraction processing is performed on the communication interference data of each candidate frequency band after preprocessing (the communication interference data of each candidate frequency band after preprocessing is input into the first fully connected layer containing 64 neurons, and the original 5-dimensional input is expanded to a 64-dimensional feature representation. The primary combination features between the input parameters are extracted by combining weighted linear transformation and ReLU activation function, such as: the inverse coupling relationship between interference power and signal-to-noise ratio, the modulation enhancement behavior of electric field amplitude on modulation depth, etc., and then the 64-dimensional feature representation is input into the second fully connected layer containing 32 neurons, and the preliminary combination features are compressed and fused. The low-impact redundant features are removed through feature dimensionality reduction processing, and cross-feature fusion operations are introduced to capture the nonlinear interaction patterns between multi-dimensional physical parameters, such as: the local signal destruction effect of modulation depth under high intermodulation interference, the frequency band instability performance under high electric field and low RSSI, etc., and finally the second fully connected layer is used to extract the communication interference data of each candidate frequency band after preprocessing. The output feature vector is input to the third fully connected layer containing 16 neurons, which further refines and aggregates the fused features. While retaining cross-parameter fusion information, this layer uses a smaller-scale neuron setting to perform feature selection and combination extraction operations on the key influencing dimensions of frequency band transmission reliability, forming a final feature output with discriminative ability. Each neuron generates an output value based on the weighted combination result of the 32-dimensional fusion features of the previous layer. After processing with the ReLU activation function, the output forms a 16-dimensional high-dimensional feature vector. For example, the instantaneous availability score of the frequency band in a high interference power environment, the impact of modulation changes on channel output stability, the distortion sensitivity of the modulated signal under high electromagnetic field density, the signal retention ability to maintain effective transmission under low signal-to-noise conditions, the degree of damage to the net channel quality of the frequency band caused by intermodulation interference components, and the dynamic recovery ability of the channel under local RSSI degradation, etc., to obtain a high-dimensional feature vector for each candidate frequency band.

[0032] In the output layer of the deep neural network, the high-dimensional feature vector of each candidate frequency band is predicted (the output layer receives the 16-dimensional feature vector output by the aforementioned third fully connected layer, where each dimensional feature corresponds to the deep discrimination information of the current candidate frequency band in terms of channel stability, anti-interference ability, modulation adaptability, signal fidelity, etc., and the output layer structure is a single neuron fully connected output node. Its core processing steps are: performing a weighted sum operation on the input 16-dimensional feature vector according to the trained weight parameters, and superimposing the bias value to form the final score prediction output, that is, the predicted transmission efficiency index, which is used to reflect the transmission feasibility of the candidate frequency band at the current moment after comprehensive evaluation of factors such as communication interference status, signal stability, and anti-distortion ability), and obtain the predicted transmission efficiency index of each candidate frequency band.

[0033] Among them, the input layer is used to receive the communication interference data of each candidate frequency band, including the interference power intensity index, the frequency band electromagnetic field amplitude, the frequency band modulation value, the signal-to-noise ratio value and the intermodulation interference intensity index, and perform normalization preprocessing on the input data to ensure the stability and numerical consistency of the feature processing of the subsequent neural network.

[0034] The fully connected layer is used to perform multi-layer feature extraction processing on the input communication interference data, and sequentially performs dimensional expansion, feature fusion, and discriminant feature aggregation operations. It automatically learns from the original physical parameters to form a high-dimensional feature vector that represents the communication status of the candidate frequency band, which is used to reflect communication characteristics such as interference intensity, modulation adaptability, and channel stability.

[0035] The output layer is used to perform a weighted scoring operation on the high-dimensional feature vector generated by the fully connected layer and output the predicted transmission efficiency index corresponding to the candidate frequency band.

[0036] And the pre-training steps of the deep neural network are as follows:

[0037] A structured communication interference feature dataset is obtained. The dataset includes several candidate frequency band samples with known transmission performance labels. Each sample contains the communication interference feature input corresponding to the frequency band (including interference power intensity index, frequency band electromagnetic field amplitude, frequency band modulation value, signal-to-noise ratio value, intermodulation interference intensity index) and its corresponding transmission benefit label (such as transmission stability score, video frame loss rate, low-latency transmission success rate, etc.), and the structured communication interference feature dataset is divided into a training set and a validation set.

[0038] Initialize and configure the deep neural network to build a network architecture including an input layer, a three-layer fully connected structure, and an output layer. During the initialization phase, set the key hyperparameters of each fully connected layer, including the number of neurons in each layer (such as 64, 32, 16), the activation function type (such as ReLU), the parameter initialization strategy (using Kaiming initialization), and the output layer structure (single-node linear regression output).

[0039] Supervised training is performed based on the training set, and the number of training rounds is set (such as 100 rounds). The following steps are performed in each round: forward propagation stage: the communication interference input vector of each sample is passed into the neural network in turn, and the predicted transmission efficiency index of the frequency band is output; loss function calculation stage: a regression loss function based on the predicted value and the true transmission label is constructed, and the mean square error (MSE) is used as the main optimization target to measure the error between the predicted score and the label; back propagation and parameter optimization stage: gradient back propagation is performed on the total loss, and the Adam optimizer is used to iteratively update the network parameters. At the same time, batch normalization, gradient clipping and cosine annealing learning rate strategy are combined to improve the stability and convergence efficiency of training.

[0040] After each round of training, a complete inference evaluation process is performed based on the validation set, the predicted transmission benefit index is output, and the evaluation indicators (such as mean square error MSE, mean absolute error MAE, R 2 Fit), and draw the loss curve and evaluation index change trend chart with Epoch to monitor the convergence and generalization ability of the training process. If there is no obvious improvement in the verification index for multiple consecutive Epochs, the Early Stopping mechanism is triggered to terminate the training process early.

[0041] After the training is completed, the final converged neural network parameter file is saved, which contains the weight parameters and bias value configurations of each fully connected layer. At the same time, the final stable prediction model structure is exported for inference deployment and practical application of the frequency band selection module.

[0042] The interference power intensity index is the instantaneous total electromagnetic interference power generated by other wireless transmitting devices (such as WiFi, Bluetooth, radar, etc.) in addition to the current transmitting source. It can scan the candidate frequency band in real time through the built-in RF receiving module. After obtaining the instantaneous power spectrum density of the frequency band, it combines the preset device transmission power template to perform signal subtraction and perform weighted summation on the interference power corresponding to the frequency points other than the current device to obtain the instantaneous interference power intensity value of the frequency band.

[0043] The frequency band electromagnetic field amplitude is the electric field strength of the electromagnetic wave in the candidate frequency band at the receiving position, which can be obtained by an open-circuit dipole electric field sensor.

[0044] The frequency band modulation index value is the instantaneous modulation amplitude of the signal on the candidate frequency band. It can be obtained by acquiring the RF signal in the frequency band and performing complex baseband demodulation based on the analog-to-digital converter (ADC) to obtain the I / Q signal pair, that is, the in-phase component and the orthogonal component, and then performing a square root operation on the sum of the squares of the in-phase component and the orthogonal component to obtain the frequency band modulation index value.

[0045] The signal-to-noise ratio (SNR) is the ratio of the target signal power to the background noise power, which can be obtained through the RSSI (Received Signal Strength Indicator) sensor.

[0046] The intermodulation interference intensity index is the ratio of the intermodulation interference power generated by nonlinear mixing in the candidate frequency band to the main signal power. It can be obtained by using a spectrum analyzer to perform a frequency scan on the candidate frequency band and its upper and lower adjacent frequencies, setting the frequency coverage range to 100 MHz above and below the center frequency of the target frequency band, and obtaining the instantaneous power value of each frequency point within the frequency range through the actual measurement function of the spectrum analyzer to form a complete power spectrum data. Secondly, based on externally known or common wireless signal frequencies (such as WiFi channel frequencies, Bluetooth signal frequencies), calculate the possible third-order intermodulation frequency points, including the combination frequency of twice a certain frequency minus another frequency, the frequency of the sum of the two frequencies, etc. In the collected power spectrum, locate the theoretical positions of these intermodulation frequency points, and in their respective The power peak is extracted within the frequency window of ±1MHz of the corresponding center frequency. The power peak is the actual interference power of the intermodulation frequency point. Then, the main signal frequency point in the current frequency band is identified. If the candidate frequency band is the screen projection communication frequency band currently used by this device, the main signal frequency is the frequency value set by the transmitter of this device; if it is in the external spectrum sensing stage, the frequency point with the highest power amplitude in the current frequency band is selected as the main signal frequency point through the spectrum analyzer scanning results. Within the range of ±1MHz of the main signal frequency, the corresponding power peak is extracted as the main signal power value. Finally, the power peak of each intermodulation frequency point is converted with the main signal power value to obtain the interference ratio of each intermodulation frequency point, and weighted processing is performed to obtain the intermodulation interference intensity index of the candidate frequency band.

[0047] The specific steps to obtain the current projection frequency band are as follows: read the predicted transmission efficiency index of each candidate frequency band, and arrange them in descending order to generate a candidate frequency band priority table; mark the candidate frequency band in the first sequence in the candidate frequency band priority table as the current projection frequency band.

[0048] In this implementation scheme, by constructing a prediction model based on deep neural networks, it is possible to actively obtain the communication interference status of multiple candidate frequency bands before video projection, and extract multi-dimensional physical characteristics including interference power, electromagnetic field strength, modulation index, signal-to-noise ratio and intermodulation interference strength. The predicted transmission efficiency index of each frequency band is further output through the pre-training model, and then the frequency band priority table is automatically constructed by sorting in descending order of the index, thereby realizing the transition of frequency band selection from static rules to real-time performance prediction. This step can make a quantitative judgment on the feasibility of frequency band transmission according to the current trend of wireless environment changes, and ensure that the selected frequency band has better transmission performance at the current moment. It is particularly suitable for complex environments with multiple devices running concurrently and spectrum interference changing frequently. Finally, by giving priority to high-scoring frequency bands as projection channels, the risk of misjudgment in initial frequency band selection can be effectively reduced, and the delay jitter caused by frequent reselection can be reduced, thereby achieving more stable and efficient low-latency video transmission path planning.

[0049] Specifically, the current projection frame image data is the pixel value and two-dimensional coordinates of each pixel point in the current projection frame image. The specific steps to obtain the projection visual complexity perception index of the current projection frequency band are as follows: the pixel value and two-dimensional coordinates of each pixel point in the current projection frame image of the current projection frequency band are comprehensively analyzed to obtain the current image evaluation set of the current projection frame of the current projection frequency band, including the structural edge transition density index (characterizing the severity of the edge contour change in the current projection frame image, reflecting the complexity of the image in terms of spatial geometric structure), the texture structure cross intensity index (characterizing the local heterogeneity of the current projection frame image at the detail texture level), and the color jump aggregation index (characterizing the spatial concentration of the color mutation area in the current projection frame image).

[0050] Based on the particle swarm optimization algorithm, a comprehensive analysis is performed on the current image evaluation set of the current projection frame of the current projection frequency band (i.e., the structure edge transition density index, texture structure cross intensity index, and color jump aggregation index are first standardized, and the standardized structure edge transition density index, texture structure cross intensity index, and color jump aggregation index are weighted. In the weighted processing process, the weight coefficients corresponding to the structure edge transition density index, texture structure cross intensity index, and color jump aggregation index are obtained through the particle swarm optimization algorithm;

[0051] The details are as follows: initialize the particle swarm, where each particle corresponds to a set of ternary weighted coefficients, such as A, B, and C, which are respectively used as weighted coefficients for the structural edge transition density index, the texture structure cross intensity index, and the color jump aggregation index, and satisfy the constraint condition A+B+C=1, and A, B, and C are all between 0-1. Secondly, for each set of weighted coefficient combinations, calculate the error difference between the image perception score under the combination and the existing user perception feedback in the historical projection task, and use the error as the fitness function value of the current particle. In each round of iteration, the speed and position are updated based on the historical optimal solution of the particle and the optimal solution of the group, and gradually converge to the weighted combination with the minimum error. Finally, extract the optimal weighted coefficient group obtained by convergence during the particle swarm optimization process, and use it as the weighted coefficient corresponding to the structural edge transition density index, the texture structure cross intensity index, and the color jump aggregation index) to obtain the projection visual complexity perception index of the current projection frame of the current projection frequency band (used to measure the complexity of the current projection frame image).

[0052] The specific steps for obtaining the current image evaluation set for the current projection frequency band are as follows:

[0053] Read the pixel value of each pixel in the current projection frame image of the current projection frequency band, and perform grayscale processing to obtain the grayscale pixel value of each pixel in the current projection frame image of the current projection frequency band, and then combine the edge detection algorithm (such as Canny edge detection algorithm) to extract the edge of the entire image to obtain several edge pixels in the current projection frame image of the current projection frequency band, and then calculate the gradient direction angle value of each edge pixel in the current projection frame image of the current projection frequency band based on the Sobel operator (that is, based on the Sobel operator, the horizontal and vertical gradients of each edge pixel are calculated, and then the vertical gradient value is used as the input vertical axis of the angle function, and the horizontal gradient value is used as the input horizontal axis, and the inverse is performed. Shear transformation to obtain the gradient direction angle value), and with each edge pixel as the center, extract all adjacent edge points in the set local neighborhood window (for example, 3×3 or 5×5 pixel range), and count their corresponding gradient direction values, and calculate the difference with the gradient direction value of the central edge point; when the gradient direction difference between a pair of edge points is greater than the preset threshold (such as 30 degrees), the position is marked as an edge direction jump, and for each edge pixel point, the number of all jump events in its local window is counted to obtain the number of edge direction jumps corresponding to the edge point. After completing the full image traversal, the jump counts of all edge pixel points are weighted fused to obtain the current image evaluation set structure edge jump density index of the current projection frequency band;

[0054] The current projection frame image is divided into several non-overlapping image block areas of fixed size (for example, the size of each area is 16×16 pixels), and the grayscale pixel value of each pixel in each image block area in the current projection frame image of the current projection frequency band is read, and the grayscale pixel value of each pixel in each image block area in the current projection frame image of the current projection frequency band is analyzed with the grayscale pixel value of each pixel in its set neighborhood based on the local binary method to obtain the LBP value of each pixel in each image block area in the current projection frame image of the current projection frequency band, and perform standard deviation processing to obtain the LBP standard deviation value of each image block area in the current projection frame image of the current projection frequency band, and perform weighted processing to obtain the texture structure cross intensity index of the current projection frequency band;

[0055] The current projection frame image is converted from RGB color space to Lab color space, and the color difference between each pixel in the current projection frame image of the current projection frequency band and its surrounding pixels (such as 3×3 neighborhood) is calculated to obtain the average color difference value between each pixel in the current projection frame image of the current projection frequency band and the neighborhood. Secondly, a color difference mutation threshold (>15) is set, and all pixels that meet this condition are marked as high color difference pixels to form a high color difference point set in the image. Subsequently, spatial clustering processing is performed on the high color difference point set, that is, a density clustering algorithm (such as DBSCA based on distance radius and minimum number of neighbors) is used. N algorithm) clusters all high color difference pixels, and counts the average clustering density of all high color difference clusters (that is, for each cluster, calculate the number of its pixels and the surrounding boundary area of ​​the cluster in the image, which is achieved by constructing a minimum circumscribed rectangle or a two-dimensional bounding box. For each cluster, calculate the unit area pixel density of the cluster, that is, the total number of pixels in the cluster divided by the area of ​​the cluster enclosed area. Then, average the unit area density values ​​of all clusters to obtain the average clustering density of the high color difference area in the entire image) and the total number of clusters, and perform weighted processing to obtain the color jump aggregation index of the current projection frequency band.

[0056] In this implementation, by comprehensively evaluating the complexity of each frame of the image in spatial dimensions such as edge jumps, texture interlacing and color aggregation, a visual complexity perception index with quantifiable expression capabilities is generated, thereby significantly improving the ability to discriminate the image content load characteristics. Especially in the projection scene with high resolution and frequent switching of multi-detail frames, the complexity of different frames varies significantly. If a unified coding strategy is adopted, it is easy to lead to waste of coding resources or degradation of compression quality. Secondly, this method performs fine-grained perception processing on the internal structure of the image, and uses the particle swarm optimization algorithm to dynamically optimize the weighting coefficients of edge, texture and color features, so that the perception index has adaptive capabilities and can effectively reflect the level of coding resources required visually for the image frame. Finally, with the help of this perception indicator, in the subsequent coding control and channel selection process, image frames with higher transmission requirements can be accurately identified, thereby achieving accurate allocation and scheduling of resources, and improving the continuity and visual consistency of the overall video transmission.

[0057] Specifically, the specific steps for obtaining the coded data volume value of the current projection frame of the current projection frequency band are as follows: read the coding control index of the current projection frame of the current projection frequency band, and perform a comprehensive analysis in combination with the preset minimum coding rate to obtain the Polar coding rate of the current projection frame of the current projection frequency band; perform Polar code encoding on the image data of the current projection frame of the current projection frequency band based on the Polar coding rate, and generate a bit coding sequence of the current projection frame of the current projection frequency band; perform statistical analysis on the bit coding sequence of the current projection frame of the current projection frequency band (count the total number of all bits in the coding sequence to obtain the data volume required to transmit the frame image under the current coding strategy, and map the result to between 0-1 through the sigmoid function, that is, the coded data volume value of the current projection frame of the current projection frequency band) to obtain the coded data volume value of the current projection frame of the current projection frequency band.

[0058] Among them, the specific formula for calculating the Polar coding rate of the current projection frame of the current projection frequency band is as follows: LmB = max[ZxB, min(1,1-μ*BmT)]; among them, LmB is the Polar coding rate of the current projection frame of the current projection frequency band, ZxB is the preset minimum coding rate, BmT is the coding control index of the current projection frame of the current projection frequency band, and μ is the coding adjustment coefficient stored in the database.

[0059] It should be explained that μ can be obtained through the following steps: collecting image data of completed projection frames in multiple historical projection tasks, recording the image complexity index, transmission channel status, actual coding rate and coding error situation of the corresponding frames, and performing nonlinear fitting modeling on the mapping relationship between the coding control index and the final coding rate in different frame samples. Based on the minimum mean square error criterion or the target transmission success rate indicator, iteratively search for the optimal adjustment scaling factor μ in the sample set to minimize the coding rate prediction error of the fitting curve in multiple groups of scenarios. Then, the obtained optimal μ value is stored in the database as the adjustment coefficient for the Polar coding rate calculation process in subsequent tasks.

[0060] The specific steps for Polar code encoding are as follows: The number of original information bits in the current projection frame is obtained, and the target total codeword length is calculated based on the Polar code rate (i.e., the ceiling value obtained by dividing the original number of bits by the Polar code rate). Then, a ranked list of bit position channel reliabilities is constructed based on the target codeword length. (The reliability of the polarization subchannel corresponding to each bit position in the codeword is assessed using methods based on Bhattacharyya parameter estimation or channel capacity approximation. Next, all bit position channels are sorted from high to low reliability to generate a ranked list of position channel reliabilities.) Based on this ranked list, the bit positions with the highest reliability are selected as information bits, and the remaining positions are frozen bits and filled with fixed values. Next, the information bits and frozen bits are arranged and combined according to the polarization structure to form a coding input vector. Polar coding is then performed, and the coding input vector is recursively combined and linearly transformed at the bit level to complete the Polar code encoding operation. Finally, a polar-coded bit sequence is output. The length of the bit sequence is equal to the target codeword length and includes the positions mapped to the original image information bits, frozen bits, and optional parity bits. This sequence is used for subsequent transmission channel selection and bit encapsulation.

[0061] In this implementation, an adaptive coding control mechanism is introduced, and the complexity characteristics of the current frame image and the channel transmission conditions are combined to dynamically determine the Polar code coding rate, thereby accurately controlling the coding data volume of each frame image, thereby realizing intelligent adjustment of coding resources in the screen projection task. Secondly, the coding rate is calculated based on the coding control index, which can actively improve the coding accuracy when the image complexity is high, compress redundant data when the image structure is simple, and improve the overall coding efficiency. Subsequently, information bits are selected through the reliability sorting mechanism and polarization coding is performed to make the bit allocation more robust. Finally, the coding data volume is mapped to a standardized range, providing an accurate quantitative basis for subsequent channel adaptation and transmission path selection, thereby effectively avoiding the problem of resource waste or image quality fluctuation under the static coding strategy, thereby improving the transmission controllability and continuity of the projection picture in complex scenarios, especially suitable for multi-scene high-definition wireless screen projection tasks with drastic content changes.

[0062] Specifically, the channel behavior structure data includes the scheduling slot rate value, the time slot occupancy rate value, the network scheduling stability index, the network conflict pressure index, and the network occupancy saturation index. The specific steps for obtaining the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band are as follows: perform a comprehensive analysis on the channel behavior structure data of each transmission channel in the Star Flash protocol stack under the current projection frequency band to obtain the transmission evaluation set of each transmission channel in the Star Flash protocol stack under the current projection frequency band, including the resource allocation structure index and the channel load interference index; perform a comprehensive analysis on the transmission evaluation set of each transmission channel in the Star Flash protocol stack under the current projection frequency band to obtain the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band.

[0063] Among them, the scheduling slot rate value is the proportion of idle time slots in the scheduling structure of the transmission channel, reflecting the redundancy of the scheduling resources in the channel. It can monitor the target transmission channel through the receiving antenna array, and the RF receiving chip completes signal demodulation and sends the scheduling frame to the bit stream parser. The parser reads the time slot allocation field in the scheduling frame bit by bit. Each bit represents the status of a time slot (occupied / idle). The number of all time slots marked as idle is counted and divided by the total number of time slots to obtain the scheduling slot rate value.

[0064] The time slot occupancy rate value is the occupancy intensity of the scheduling resources of the transmission channel. It can receive the scheduling synchronization frame signal in the transmission channel through a directional RF receiving antenna set in the target frequency band. After being processed by a low-noise amplifier and a down-conversion mixer, it is converted into an intermediate frequency digital signal by a high-speed analog-to-digital converter (ADC). The digital signal enters the scheduling frame structure analysis circuit and extracts the slot status allocation field. This field is a bit sequence with the same length as the number of time slots. Each bit indicates whether the current time slot has been allocated for use. The number of bits marked as occupied in the bit sequence is counted through a counting and comparison logic circuit, and the ratio is calculated with the total number of time slots to output the normalized time slot occupancy rate value.

[0065] The network scheduling stability index is the degree of standardization of the scheduling structure in the transmission channel. It can receive the scheduling synchronization signal on the target transmission channel through a directional RF receiving antenna, and the signal amplitude is boosted by a low-noise amplifier (LNA). After down-conversion processing by a mixer and a bandpass filter, it is input into a high-speed analog-to-digital converter (ADC) for digital sampling. The digital signal is then sent to the frame structure decoding circuit, and the following fields are extracted from it: Slot Count (total number of time slots) field, which is used to indicate the number of time slots included in this scheduling cycle; Slot Interval (time slot period) field, which is used to indicate the interval between consecutive time slots under ideal conditions; Slot Position List (time slot start time list), which indicates the actual starting offset of each time slot relative to the starting point of the scheduling frame; Frame Header Sync Marker (frame synchronization flag bit), which is used to identify the starting reference time of the scheduling frame; then the above fields are input into the locally stored protocol standard template structure matching table through the scheduling structure analysis logic circuit, and the following processing steps are performed in sequence: the actual starting time interval between all adjacent time slots is calculated and compared with the Slot Interval is compared to obtain the time offset difference of each time slot pair; the standard deviation of all time interval offset differences is calculated to obtain the slot interval jitter index; the difference between Slot Count and the standard value specified by the protocol is counted; the weighted normalized score value of the above structural offset indicators is output as a network scheduling stability index in the range of 0-1.

[0066] The network conflict pressure index is the channel access pressure caused by concurrent competition in the transmission channel. It can monitor the signal field of the current channel through a full-band broadband receiving antenna. The signal is monitored for real-time energy fluctuations through a bandpass filter and a power detection circuit. When multiple devices try to access at the same time without scheduling control, multiple burst energy rising edges that partially overlap in time will be formed in the signal field. This phenomenon is captured by a high-precision energy edge identifier, and then the rising edge density is judged. The peak overlap judgment is completed by the conflict event identification circuit, and the number of valid conflict events is counted, which is regarded as the network conflict pressure index.

[0067] The network occupancy saturation index is the usage density of the scheduling resources of the transmission channel, reflecting the load intensity level of the channel. It can capture the scheduling frame structure of the transmission channel through the scheduling frame receiving channel, and extract the current status field of each time slot and the corresponding historical load flag bit (such as the slot continuous use flag bit and the slot data continuous frame bit) through the frame content parsing circuit. To ensure that it does not rely on the historical cache, only the continuous time slot occupancy flag field existing in the current frame is used to identify which time slots are currently in a valid continuous use state, and calculate the ratio between the number of continuous time slots occupied by valid data frames and the maximum total number of allocable slots defined by the scheduling structure, and output the normalized index as the network occupancy saturation index.

[0068] The specific formula for calculating the transmission steady-state response index of a transmission channel in the StarFlash protocol stack under the current projection frequency band is as follows: Among them, CsW is the transmission steady-state response index of a transmission channel in the Star Flash protocol stack under the current projection frequency band, ZyP is the resource configuration structure index of a transmission channel in the Star Flash protocol stack under the current projection frequency band, ω1 is the resource configuration adjustment coefficient stored in the database, TgR is the channel load interference index of a transmission channel in the Star Flash protocol stack under the current projection frequency band, ω2 is the load interference adjustment coefficient stored in the database, and ω3 is the coordination adjustment coefficient stored in the database.

[0069] It should be explained that ω1, ω2, and ω3 can be obtained through the following steps: based on historical data, the initial impact weights of each variable (resource allocation structure index, channel load interference index) on the transmission steady-state response index are determined through statistical regression analysis. Then, the value range of the coefficient is adjusted using the sensitivity analysis method to evaluate the stability and applicability of these parameters to the formula output. Next, the weights are further fitted through model optimization (such as machine learning algorithms) to ensure that the formula can accurately reflect the stability of the transmission channel.

[0070] The specific implementation example of calculating the transmission steady-state response index of a transmission channel in the Star Flash protocol stack under the current projection frequency band is as follows. The following data is available: including the resource configuration structure index and channel load interference index of 5 transmission channels (randomly selected) in the Star Flash protocol stack under the current projection frequency band, as shown in Table 1:

[0071] Table 1 Example of transmission channel sequence transmission evaluation set data in the Star Flash protocol stack under the current projection frequency band

[0072] Resource Allocation Structure Index Channel load interference index Transmission channel 1 0.754 0.415 Transmission Channel 2 0.643 0.327 Transmission Channel 3 0.884 0.397 Transmission channel 4 0.589 0.546 Transmission channel 5 0.829 0.485

[0073] The resource allocation adjustment coefficient ω1 stored in the database is approximately: 0.267;

[0074] The load interference adjustment coefficient ω2 stored in the database is approximately: 0.516;

[0075] The synergistic adjustment coefficient ω3 stored in the database is approximately: 0.217;

[0076] Substituting the data in Table 1 and the above adjustment coefficients into the specific formula for calculating the transmission steady-state response index of a transmission channel in the Star Flash protocol stack under the current projection frequency band, we obtain:

[0077] The transmission steady-state response index of the first transmission channel in the Star Flash protocol stack under the current projection frequency band = ln(1+0.754 0.267 ×(0.516 / 0.415))×(1+tanh(0.217×(0.754 / 0.415)))≈1.053;

[0078] The transmission steady-state response index of the second transmission channel in the Star Flash protocol stack under the current projection frequency band = ln(1+0.643 0.267 ×(0.516 / 0.327))×(1+tanh(0.217×(0.643 / 0.327)))≈1.229;

[0079] The transmission steady-state response index of the third transmission channel in the Star Flash protocol stack under the current projection frequency band = ln(1+0.589 0.267 ×(0.516 / 0.546))×(1+tanh(0.217×(0.589 / 0.546)))≈1.179;

[0080] The transmission steady-state response index of the fourth transmission channel in the Star Flash protocol stack under the current projection frequency band = ln(1+0.884 0.267 ×(0.516 / 0.397))×(1+tanh(0.217×(0.884 / 0.397)))≈0.737;

[0081] The transmission steady-state response index of the fifth transmission channel in the Star Flash protocol stack under the current projection frequency band = ln(1+0.829 0.267 ×(0.516 / 0.485))×(1+tanh(0.217×(0.829 / 0.485)))≈0.914.

[0082] The specific steps to obtain the transmission evaluation set of each transmission channel in the Star Flash protocol stack under the current projection frequency band are as follows: read the scheduling slot rate value, time slot occupancy value, and network scheduling stability index of each transmission channel in the Star Flash protocol stack under the current projection frequency band, and perform a comprehensive analysis in combination with the Bayesian confidence perception method (that is, first standardize the scheduling slot rate value, time slot occupancy value, and network scheduling stability index, and perform weighted processing based on the standardized processing results. In the weighted processing process, the weighted coefficients corresponding to the scheduling slot rate value, time slot occupancy value, and network scheduling stability index are obtained through the Bayesian confidence perception algorithm, which is as follows: First, preset each The ideal interval range of each structural parameter is used as the expected performance of a good channel resource structure. The ideal interval range can be obtained by selecting historical communication records with long-term stable operation, low transmission packet loss rate and few interference events, and performing central trend analysis and deviation range statistics on the corresponding scheduling slot rate value, time slot occupancy rate value and scheduling stability index. The main value distribution interval of each parameter in the high-quality transmission scenario is extracted. The obtained interval is normalized and used as the ideal value range of the structural parameter to constitute the structural prior model used. In each transmission channel, the fit degree of each measured parameter value with its ideal interval is analyzed to evaluate whether the parameter value is in the ideal range. The probability within the interval is used as its current confidence strength. A triangular membership function or an equivalent probability window function can be used to construct a confidence function for each parameter. The closer the value is to the center of the ideal interval, the higher its confidence is. Based on this, the instantaneous confidence corresponding to the three parameters is obtained. Subsequently, the confidence of each parameter is normalized so that its sum is equal to one, forming a weighted coefficient corresponding to the scheduling slot rate value, the time slot occupancy rate value, and the network scheduling stability index) to obtain the resource configuration structure index of each transmission channel in the Star Flash protocol stack under the current projection frequency band; read the network conflict pressure index, network saturation index, and network conflict pressure index of each transmission channel in the Star Flash protocol stack under the current projection frequency band. and degree index, and combined with the Bayesian confidence perception method for comprehensive analysis (that is, the network conflict pressure index and the network occupancy saturation index are first standardized, and the standardized network conflict pressure index and the network occupancy saturation index are weighted, and in the weighted processing process, the weighted coefficients corresponding to the network conflict pressure index and the network occupancy saturation index are obtained through the Bayesian confidence perception method, and the logic is consistent with the weighted coefficients corresponding to the vacancy rate value, the time slot occupancy rate value, and the network scheduling stability index obtained through the Bayesian confidence perception algorithm), and the channel load interference index of each transmission channel in the Star Flash protocol stack under the current projection frequency band is obtained.

[0083] In this implementation scheme, the behavioral structure data of each transmission channel in the Star Flash protocol stack is refinedly modeled to construct a resource configuration structure index and a channel load interference index, and further generate a transmission steady-state response index, thereby realizing multi-dimensional perception and quantifiable evaluation of the structural state of the transmission channel. In the specific implementation, parameters such as scheduling slot rate, time slot occupancy rate, scheduling stability, etc. are collected respectively, and the Bayesian confidence perception algorithm is introduced to perform probabilistic modeling of the degree of fit between each parameter and the ideal structural performance, so that the weight distribution of each indicator is more adaptive and reasonable. At the same time, combined with load information such as conflict pressure and saturation, a fusion indicator reflecting the channel state stability and anti-interference ability is formed, which provides a reliable channel evaluation basis for subsequent image transmission tasks, thereby significantly improving the accuracy and robustness of the channel scheduling process, and effectively avoiding problems such as frame transmission failure and delay jitter caused by structural bottlenecks or resource conflicts, thereby ensuring the continuity and low latency characteristics of the screen projection task.

[0084] Specifically, if Figure 3 As shown, the specific steps of dynamically wirelessly transmitting the image data of the current projection frame after encoding processing in combination with the Star Flash protocol stack are as follows: obtain the channel behavior structure data of several transmission channels in the Star Flash protocol stack under the current projection frequency band, and perform a comprehensive analysis to obtain the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band; read the encoded data volume value and projection visual complexity perception index of the current projection frame of the current projection frequency band, and perform a comprehensive analysis in combination with the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band to obtain the image negative coordination adaptation index of each transmission channel in the Star Flash protocol stack under the current projection frequency band; based on the image negative coordination adaptation index of each transmission channel in the Star Flash protocol stack under the current projection frequency band, wirelessly transmit the image data of the current projection frame after encoding processing (that is, arrange the image negative coordination adaptation index of each transmission channel in the Star Flash protocol stack under the current projection frequency band in descending order, and select the image negative coordination adaptation index). The transmission channel with the largest transmission suitability index is used as the target transmission channel, and based on the scheduling structure of the Star Flash protocol, the projection resource allocation instruction for the channel in the previous projection frame is generated. Then, according to the scheduling instruction, the encoded image data is wirelessly sent through the target transmission channel, and the data packet distribution and reception confirmation are completed within the projection cycle. This process is synchronously controlled by the Star Flash protocol scheduling beat to ensure that the data frame is transmitted according to the time slot. After the terminal device receives the complete encoded projection frame, the frame synchronization is reorganized according to the image frame sequence number and coding identifier, and it is decoded and rendered to complete the display of this round of wireless projection frame). Dynamic feedback monitoring is continuously performed during the projection process. That is, after the wireless projection task of each frame is completed, the candidate frequency band, image content and transmission channel of the next projection frame are re-analyzed, and the steps of prediction analysis, feature analysis, adaptive Polar code encoding processing and wireless transmission are repeated to directly complete the transmission of all projection frames of the video.

[0085] The specific formula for calculating the image-negative coordination suitability index of a transmission channel in the StarFlash protocol stack under the current projection frequency band is as follows: Among them, TxS is the image negative collaborative adaptation index of a transmission channel in the Star Flash protocol stack under the current projection frequency band, BaT is the coded data volume value of the current projection frame of the current projection frequency band, β1 is the coded volume adjustment coefficient stored in the database, TsF is the projection visual complexity perception index of the current projection frame of the current projection frequency band, β2 is the complex perception adjustment coefficient stored in the database, CsW is the transmission steady-state response index of a transmission channel in the Star Flash protocol stack under the current projection frequency band, β3 is the transmission steady-state adjustment coefficient stored in the database, and β4 is the projection visual buffer enhancement adjustment coefficient stored in the database.

[0086] It needs to be explained that the formula This item is used to respond and adjust the transmission steady-state capability of the transmission channel in the current projection frequency band in visually complex scenes, ensuring that the transmission channel has stability enhancement capabilities under low-complexity images, and exhibits adaptive suppression characteristics under high-complexity images.

[0087] β1, β2, β3, and β4 can be obtained through the following steps: using historical data, combined with the coded data volume value, the projection visual complexity perception index, and the transmission steady-state response index, statistical regression analysis is performed to quantify the specific impact of each factor on the image-negative collaborative adaptation index, thereby fitting the initial weight value. Secondly, the sensitivity analysis method is used to adjust the value range of each coefficient and observe its impact on the image-negative collaborative adaptation evaluation results to ensure the stability and rationality of the model. Based on the characteristics of the transmission channel and the actual situation, the preliminary fitted coefficients are corrected and optimized, and finally the coefficient values ​​applicable to the specific transmission channel are determined.

[0088] In this implementation scheme, by constructing the image-negative collaborative adaptation index, a deep fusion evaluation mechanism between the image content load and the transmission channel capacity is realized. With the support of the Star Flash protocol stack, it can be based on the encoded data volume value of each frame image and the visual complexity perception index, and then combined with the current steady-state response capability of each channel, accurately calculate the adaptation matching degree of each channel to the frame data. Secondly, in the scheduling process, by dynamically sorting the image-negative collaborative adaptation index and selecting the optimal channel, high-complexity frames are automatically guided to the transmission path with strong load-bearing capacity and high resource redundancy, thereby avoiding the screen freeze, frame delay or misordering problems caused by the allocation of complex images to inferior channels to the greatest extent. At the same time, this step cooperates with the beat synchronization mechanism and feedback perception mechanism of the Star Flash protocol, so that the system has frame-level dynamic scheduling capabilities, forming a closed-loop optimization path, thereby effectively improving the fluency, stability and picture quality consistency of the entire screen projection session, which is particularly suitable for bandwidth-sensitive application scenarios such as high-definition video and high-frequency frame rates.

[0089] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0090] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology, characterized in that: The following steps are involved: When wirelessly projecting a video, the communication interference data of several candidate frequency bands is obtained in real time and input into a pre-trained reinforcement learning model for predictive analysis to obtain the predicted transmission efficiency index for each candidate frequency band. Comprehensively analyze the predicted transmission efficiency index of each candidate frequency band to obtain the current projection frequency band; Obtain the image data of the current projection frame of the current projection frequency band and perform feature analysis to obtain the projection visual complexity perception index of the current projection frequency band. Combined with the predicted transmission efficiency index of the current projection frequency band, a comprehensive analysis is performed to obtain the coding control index of the current projection frame of the current projection frequency band. Based on the coding control index, the image data of the current projection frame of the current projection frequency band is adaptively encoded with Polar code to obtain the coded data volume value of the current projection frequency band. The image data of the current projection frame after encoding is dynamically wirelessly transmitted in combination with the Star Flash protocol stack.

2. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 1 is characterized in that: The communication interference data includes an interference power intensity index, a frequency band electromagnetic field amplitude, a frequency band modulation index, a signal-to-noise ratio value, and an intermodulation interference intensity index. The reinforcement learning model is specifically a deep neural network, which includes an input layer, several fully connected layers, and an output layer. The specific steps of obtaining the predicted transmission efficiency index of each candidate frequency band are as follows: In the input layer of the deep neural network, the communication interference data of each candidate frequency band is received and preprocessed; In the fully connected layer of the deep neural network, multi-layer feature extraction is performed on the pre-processed communication interference data of each candidate frequency band to obtain a high-dimensional feature vector for each candidate frequency band; In the output layer of the deep neural network, the high-dimensional feature vector of each candidate frequency band is predicted to obtain the predicted transmission efficiency index of each candidate frequency band.

3. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 2 is characterized in that: The specific steps to obtain the current screen projection frequency band are as follows: Read the predicted transmission efficiency index of each candidate frequency band, sort them in descending order, and generate a candidate frequency band priority table; The candidate frequency band in the first sequence in the candidate frequency band priority table is marked as the current projection frequency band.

4. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 1 is characterized in that: The current projection frame image data specifically includes the pixel value and two-dimensional coordinates of each pixel point in the current projection frame image. The specific steps for obtaining the projection visual complexity perception index of the current projection frequency band are as follows: Comprehensively analyze the pixel value and two-dimensional coordinates of each pixel in the current projection frame image of the current projection frequency band to obtain the current image evaluation set of the current projection frame of the current projection frequency band, including the structure edge transition density index, texture structure cross intensity index, and color jump aggregation index; Based on the particle swarm optimization algorithm, a comprehensive analysis is performed on the current image evaluation set of the current projection frame in the current projection frequency band to obtain the projection visual complexity perception index of the current projection frame in the current projection frequency band.

5. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 4 is characterized in that: The specific formula for calculating the coding control index of the current projection frame in the current projection frequency band is as follows: Among them, BmT, TsF, and YcX are the coding control index, projection visual complexity perception index, and predicted transmission efficiency index of the current projection frame in the current projection frequency band, respectively. α1, α2, and α3 are the visual adjustment coefficient, transmission efficiency adjustment coefficient, and interaction adjustment coefficient stored in the database, respectively.

6. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 1 is characterized in that: The specific steps to obtain the coded data volume value of the current projection frame in the current projection frequency band are as follows: Read the coding control index of the current projection frame in the current projection frequency band, and perform a comprehensive analysis based on the preset minimum coding rate to obtain the Polar coding rate of the current projection frame in the current projection frequency band; Perform Polar code encoding on the image data of the current projection frame of the current projection frequency band based on the Polar code rate, and generate a bit code sequence of the current projection frame of the current projection frequency band; Statistical analysis is performed on the bit coding sequence of the current projection frame of the current projection frequency band to obtain the coded data volume value of the current projection frame of the current projection frequency band.

7. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 1 is characterized in that: The specific steps for dynamically wirelessly transmitting the encoded image data of the current projection frame in combination with the Star Flash protocol stack are as follows: Obtain the channel behavior structure data of several transmission channels in the StarFlash protocol stack under the current projection frequency band, and perform comprehensive analysis to obtain the transmission steady-state response index of each transmission channel in the StarFlash protocol stack under the current projection frequency band; Read the coded data volume value and projection visual complexity perception index of the current projection frame in the current projection frequency band, and perform a comprehensive analysis based on the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band to obtain the image-negative collaborative transmission adaptation index of each transmission channel in the Star Flash protocol stack under the current projection frequency band; Based on the image-negative cooperative adaptation index of each transmission channel in the StarFlash protocol stack under the current projection frequency band, the image data of the current projection frame after encoding is wirelessly transmitted.

8. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 7 is characterized in that: The channel behavior structure data includes the scheduling slot rate value, the time slot occupancy value, the network scheduling stability index, the network conflict pressure index, and the network occupancy saturation index. The specific steps for obtaining the transmission steady-state response index of each transmission channel in the Star Flash protocol stack under the current projection frequency band are as follows: Comprehensively analyze the channel behavior structure data of each transmission channel in the StarFlash protocol stack under the current projection frequency band to obtain the transmission evaluation set of each transmission channel in the StarFlash protocol stack under the current projection frequency band, including the resource configuration structure index and the channel load interference index; A comprehensive analysis is performed on the transmission evaluation set of each transmission channel in the StarFlash protocol stack under the current projection frequency band to obtain the transmission steady-state response index of each transmission channel in the StarFlash protocol stack under the current projection frequency band.

9. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 8 is characterized in that: The specific steps to obtain the transmission evaluation set of each transmission channel in the StarFlash protocol stack under the current projection frequency band are as follows: Read the scheduling slot rate value, time slot occupancy value, and network scheduling stability index of each transmission channel in the Star Flash protocol stack under the current projection frequency band, and perform a comprehensive analysis combined with the Bayesian confidence perception method to obtain the resource configuration structure index of each transmission channel in the Star Flash protocol stack under the current projection frequency band; The network conflict pressure index and network occupancy saturation index of each transmission channel in the StarFlash protocol stack under the current projection frequency band are read, and a comprehensive analysis is performed in combination with the Bayesian confidence perception method to obtain the channel load interference index of each transmission channel in the StarFlash protocol stack under the current projection frequency band.

10. The low-latency wireless screen projection dynamic spectrum allocation method based on Star Flash technology according to claim 7 is characterized in that: The specific formula for calculating the image-negative coordination suitability index of a transmission channel in the StarFlash protocol stack under the current projection frequency band is as follows: Among them, TxS and CsW are respectively the image negative coordination adaptation index and transmission steady-state response index of a transmission channel in the Star Flash protocol stack under the current projection frequency band; BaT and TsF are respectively the coded data volume value and projection visual complexity perception index of the current projection frame in the current projection frequency band; β1, β2, β3, and β4 are respectively the coding volume adjustment coefficient, complex perception adjustment coefficient, transmission steady-state adjustment coefficient, and projection visual buffer enhancement adjustment coefficient stored in the database.

Images