Low-delay transmission method and system for realizing virtual-real fusion of intelligent internet of things and universe

Abstract

The invention provides a low-delay transmission method and system for realizing virtual-real fusion of intelligent Internet of Things and meta-universe, and relates to the technical field of meta-universe level, and the method comprises the steps: obtaining and classifying physical world data, building a mapping relation between a data stream and a virtual space, constructing a double-layer data processing architecture, and setting different transmission priorities. And monitoring network quality in real time and starting a data compensation mechanism. According to the method, real-time interaction between the physical world and the virtual space can be realized, the data transmission efficiency and stability are improved, the system delay is reduced, and the immersion experience of a user is enhanced.

Description

Technical Field

[0001] The present invention relates to metaverse technology, and in particular to a low-latency transmission method and system for realizing the virtual-real integration of intelligent Internet of Things and metaverse. Background Art

[0002] With the rapid development of the Internet of Things and Metaverse technology, low-latency transmission that integrates the virtual and real worlds of smart IoT and Metaverse has become a hot topic in current research. Smart IoT terminals can collect a large amount of physical world data, while Metaverse provides a virtualized display and interactive platform for these data. The organic combination of the two can achieve seamless connection between the physical world and the virtual world, bringing an immersive experience to users.

[0003] However, there are still some technical challenges in the process of realizing the integration of smart IoT and the metaverse. First, the massive amount of IoT data transmission will bring huge pressure to the network, which is easy to cause transmission delays and affect user experience. Secondly, there is a lack of effective mapping mechanism between physical world data and metaverse virtual space, making it difficult to achieve accurate correspondence and real-time interaction between the two worlds. Finally, the existing data processing architecture is often single and cannot perform differentiated processing for different types of data, resulting in inefficient resource utilization.

[0004] In order to solve these problems, a new low-latency transmission method is urgently needed that can effectively process and transmit massive amounts of IoT data, achieve accurate mapping of the physical world and the virtual space of the metaverse, and improve transmission efficiency through a multi-level data processing architecture, thereby providing users with a smooth and real-time virtual-reality fusion experience. Summary of the invention

[0005] The embodiments of the present invention provide a low-latency transmission method and system for realizing the virtual-real integration of smart Internet of Things and metaverse, which can solve the problems in the prior art.

[0006] According to a first aspect of the embodiments of the present invention, Provide a low-latency transmission method to achieve the virtual-real integration of smart IoT and Metaverse, including: Acquire physical world data sent by multiple smart IoT terminals, where the physical world data includes terminal identification information; Classify and sort the physical world data according to the terminal identification information, generate corresponding data streams, and establish a mapping relationship between the data streams and the metaverse virtual space; A two-layer data processing architecture is constructed based on the mapping relationship, in which a state prediction algorithm is used to obtain target state data in the first layer; a scene collaborative processing unit is set in the second layer to construct a scene fusion model in combination with the target state data; Different transmission priorities are set for the target status data and the scene fusion model, the target status data is divided into a first priority data stream, and the scene fusion model is divided into a second priority data stream; the network quality of each priority data stream is monitored in real time, and when network fluctuations are detected, a data compensation mechanism is started to compensate for different transmission priorities based on the prediction results of the state prediction algorithm.

[0007] Classifying and arranging the physical world data according to the terminal identification information, generating corresponding data streams, and establishing a mapping relationship between the data streams and the metaverse virtual space includes: Classifying the physical world data according to the terminal type and data type according to the terminal identification information to obtain an initial data set; Extracting data information from the initial data set to generate data feature labels, converting the data feature labels into corresponding semantic description information based on a preset semantic mapping rule, clustering the data in the initial data set according to the semantic description information, and generating a data stream with semantic association; The data stream is mapped to the metaverse virtual space to establish a mapping relationship between the physical world and the virtual space.

[0008] A two-layer data processing architecture is constructed based on the mapping relationship, in which a state prediction algorithm is used to obtain target state data in the first layer; and a scene collaborative processing unit is set in the second layer to construct a scene fusion model in combination with the target state data, including: Acquire a historical state sequence of the IoT terminal, wherein the historical state sequence includes a position vector, a velocity vector, and an acceleration vector; Divide the historical state sequence into multiple state sequence segments according to a preset time window size; input the state sequence segments into a long short-term memory network, the long short-term memory network generates a hidden state at the current moment according to the state sequence segments at the current moment and the hidden state at the previous moment, and obtains an initial prediction state according to the hidden state through a linear transformation of a weight matrix and a bias term; Acquire real-time observation data of the IoT terminal, construct a prediction error covariance matrix and an observation noise covariance matrix based on the real-time observation data, and perform a matrix operation on the product of the prediction error covariance matrix and the observation matrix to obtain a Kalman gain; Correcting the initial predicted state according to the Kalman gain specifically includes: calculating the deviation between the real-time observation data and the initial predicted state, and adding the product of the deviation and the Kalman gain to the initial predicted state to obtain corrected target state data; The corrected target state data is sent to the IoT terminal, the state at the next moment is updated, and a scene fusion model is obtained.

[0009] Constructing a prediction error covariance matrix and an observation noise covariance matrix based on the real-time observation data, and performing a matrix operation on the product of the prediction error covariance matrix and the observation matrix to obtain a Kalman gain includes: Obtaining the real state and predicted state of the IoT terminal, subtracting the predicted state from the real state to obtain a prediction error, multiplying the prediction error by the transpose of the prediction error, calculating an expected value of the result of the transpose multiplication, and obtaining a prediction error covariance matrix; Calculating the deviation between the real-time observation data and the predicted state to obtain an innovation sequence, and performing a cumulative summation operation on the innovation sequence within a preset sliding window to obtain an observation noise covariance matrix; Adding the prediction error covariance matrix to the observation noise covariance matrix to obtain an innovative covariance; dividing the prediction error covariance matrix by the observation noise covariance matrix to obtain an adaptive weight factor; The prediction error covariance matrix is ​​multiplied by the observation noise covariance matrix to obtain a seventh product, the seventh product is multiplied by the inverse matrix of the innovation covariance to obtain an initial Kalman gain, and the initial Kalman gain is multiplied by the adaptive weight factor to obtain a modified Kalman gain.

[0010] Real-time monitoring of the network quality of each priority data stream, when a network fluctuation is detected, starting a data compensation mechanism, and performing compensation processing on different transmission priorities based on the prediction result of the state prediction algorithm includes: Constructing a network state vector including end-to-end delay and packet loss rate, adding the product of the network state vector and the smoothing factor to the product of the smoothing state value at the previous moment and the smoothing factor to obtain the smoothing state value at the current moment; Subtracting the smooth state value at the previous moment from the smooth state value at the current moment to obtain a difference vector, calculating the second norm of the difference vector to obtain a fluctuation detection value, adding the smooth state value at the current moment to the product of the fluctuation detection value and a preset adjustment coefficient, and constructing a multi-level fluctuation threshold; Inputting the fluctuation detection value and the multi-level fluctuation threshold into a compensation mapping function to calculate a compensation intensity; Based on the packet loss rate divided by the compensation strength to obtain a third product, the fluctuation detection value is multiplied by a preset fluctuation compensation coefficient to obtain a fourth product, the third product is added to the fourth product and rounded up to obtain the number of first priority compensated data packets; Dividing the end-to-end delay by the product of the fluctuation detection value and the preset adjustment coefficient to obtain a calculation result; comparing the calculation result with the minimum window value, and taking the product greater than the preset window value threshold as the compensation time window; Compensation operations are performed on data streams of different priorities according to the number of the first priority compensation data packets, the number of redundant packets, and the compensation time window.

[0011] Dividing the end-to-end delay by the product of the fluctuation detection value and the preset adjustment coefficient to obtain a calculation result; comparing the calculation result with the minimum window value, and taking the product greater than the preset window value threshold as the compensation time window includes: Performing a division operation using the product of the end-to-end delay, the fluctuation detection value and the preset adjustment coefficient as a divisor to obtain an initial compensation time; Compare the initial compensation time with a preset minimum window value, and when the initial compensation time is greater than the preset minimum window value, use the initial compensation time as a reference compensation window value; Multiplying the reference compensation window value by a preset smoothing coefficient to obtain a first product, multiplying the compensation window value at a previous moment by a complement of the preset smoothing coefficient to obtain a second product, and adding the first product to the second product to obtain a smoothing compensation window value; Multiplying the smoothing compensation window value by the reference compensation window value to obtain a correction compensation window value; calculating a current resource utilization rate and a weighted service quality value based on the correction compensation window value; The objective function is constructed by subtracting the product of the current resource utilization and the preset balance factor from the weighted service quality value, and the optimal compensation window value is obtained by solving the maximum value of the objective function under the constraint that the correction compensation window value is between the preset minimum window value and the preset maximum window value; A performance evaluation value is calculated based on the optimal compensation window value, and the preset adjustment coefficient is updated according to the performance evaluation value; and the optimal compensation window value is used as a final compensation time window value.

[0012] According to a second aspect of the embodiments of the present invention, Provide a low-latency transmission system that realizes the virtual-real integration of intelligent IoT and Metaverse, including: A first unit is used to obtain physical world data sent by multiple smart IoT terminals, where the physical world data includes terminal identification information; The second unit is used to classify and sort the physical world data according to the terminal identification information, generate a corresponding data stream, and establish a mapping relationship between the data stream and the metaverse virtual space; The third unit is used to construct a two-layer data processing architecture based on the mapping relationship, wherein a state prediction algorithm is used to obtain target state data in the first layer; and a scene collaborative processing unit is set in the second layer to construct a scene fusion model in combination with the target state data; The fourth unit is used to set different transmission priorities for the target status data and the scene fusion model, divide the target status data into a first priority data stream, and divide the scene fusion model into a second priority data stream; monitor the network quality of each priority data stream in real time, and when network fluctuations are detected, start the data compensation mechanism to compensate for different transmission priorities based on the prediction results of the state prediction algorithm.

[0013] According to a third aspect of the embodiments of the present invention, An electronic device is provided, comprising: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0014] A fourth aspect of the embodiments of the present invention is: A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0015] The beneficial effects of this application are as follows: By establishing a mapping relationship between the data flow of the physical world and the virtual space of the Metaverse and building a two-layer data processing architecture, an efficient integration of the physical world and the virtual world is achieved. This architecture can better process and integrate data from different smart IoT terminals, improving the efficiency and accuracy of data processing.

[0016] The use of state prediction algorithms and scene fusion models can effectively reduce data transmission delays. By predicting target state data and building scene fusion models, the system can respond to changes more quickly, reducing data processing and transmission time, thereby improving user experience.

[0017] Setting different transmission priorities and real-time monitoring of network quality, combined with data compensation mechanism, significantly improves the stability and reliability of the system. When the network fluctuates, the system can intelligently adjust the data transmission strategy to ensure the timely transmission of key data, effectively reducing the impact of network fluctuations on system performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic flow chart of a low-latency transmission method for realizing virtual-real integration of intelligent IoT and metaverse according to an embodiment of the present invention; Figure 2 This is a flow chart of the mapping relationship between physical world data and metaverse virtual space in an embodiment of the present invention; Figure 3 A flowchart of the dual-layer data processing architecture and scene fusion model construction according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0021] Figure 1 Schematic diagram of a low-latency transmission method for realizing virtual-real fusion of intelligent IoT and metaverse according to an embodiment of the present invention, as shown in FIG. Figure 1 As shown, the method includes: Acquire physical world data sent by multiple smart IoT terminals, where the physical world data includes terminal identification information; Classify and sort the physical world data according to the terminal identification information, generate corresponding data streams, and establish a mapping relationship between the data streams and the metaverse virtual space; A two-layer data processing architecture is constructed based on the mapping relationship, in which a state prediction algorithm is used to obtain target state data in the first layer; a scene collaborative processing unit is set in the second layer to construct a scene fusion model in combination with the target state data; Different transmission priorities are set for the target status data and the scene fusion model, the target status data is divided into a first priority data stream, and the scene fusion model is divided into a second priority data stream; the network quality of each priority data stream is monitored in real time, and when network fluctuations are detected, a data compensation mechanism is started to compensate for different transmission priorities based on the prediction results of the state prediction algorithm.

[0022] In an optional implementation, classifying and arranging the physical world data according to the terminal identification information, generating a corresponding data stream, and establishing a mapping relationship between the data stream and the metaverse virtual space includes: Classifying the physical world data according to the terminal type and data type according to the terminal identification information to obtain an initial data set; Extracting data information from the initial data set to generate data feature labels, converting the data feature labels into corresponding semantic description information based on a preset semantic mapping rule, clustering the data in the initial data set according to the semantic description information, and generating a data stream with semantic association; The data stream is mapped to the metaverse virtual space to establish a mapping relationship between the physical world and the virtual space.

[0023] When obtaining terminal identification information, the unique identifier of the terminal device (such as MAC address, device serial number, IMEI code, etc.), device type information (such as smartphones, VR headsets, smart watches, etc.) and data collection capabilities (such as camera parameters, sensor type, data sampling rate, etc.) can be collected through the device registration module. This information is stored in the terminal information database to form a terminal device portrait. For example, for a VR headset device equipped with a depth camera, its terminal identification information may include: "Device type: VR headset; Manufacturer: Brand A; Device ID: VR20250410001; Sensor: 6DoF position tracking, gesture recognition; Resolution: 4K×2; Refresh rate: 120Hz".

[0024] When classifying the physical world data according to the terminal identification information, the received data is preliminarily classified by the data classification engine. In the specific implementation, a terminal type mapping table and a data type classification rule base can be established. For example, data from a smartphone can be divided into geographic location data, acceleration data, image data, etc.; data from smart home devices can be divided into temperature data, humidity data, appliance status data, etc. By matching the identification information with the rules, an initial data set is generated, which contains data source information, data type labels, and raw data content. Taking the walking trajectory collected by a smartphone as an example, the initial data set may contain information such as: "Data source: smartphone A; data type: location trajectory; raw data: {timestamp, longitude, latitude, altitude, accuracy}".

[0025] When extracting features from the initial data set and generating data feature labels, the feature extractor is used to process the initial data. For location data, features such as moving speed, moving direction, and dwell time can be extracted; for image data, features such as object category, scene type, and character posture can be extracted; for audio data, features such as voice content and ambient sound classification can be extracted. Taking location trajectory data as an example, the extracted feature labels may be: "moving speed: 5km / h; moving mode: walking; trajectory shape: circular; dwell points: 3; duration: 30 minutes".

[0026] When converting data feature labels into semantic description information based on preset semantic mapping rules, the semantic mapping module converts machine-understood feature labels into description information with semantic understanding according to the predefined mapping rule base. For example, the location trajectory "moving speed: 5km / h; moving mode: walking; trajectory shape: circular; stop points: 3" can be converted into "the user did a 30-minute walking exercise in the park, circled the park, and stopped at the fitness area, rest pavilion, and lakeside on the way." The semantic mapping rule base realizes the mapping of features to semantics by associating data features with real-world semantic concepts.

[0027] When clustering the initial data set according to the semantic description information, the data clustering engine groups the data based on the semantic similarity calculation. In the specific implementation, the semantic distance measurement method can be used to calculate the semantic similarity between different data, and the hierarchical clustering or density clustering algorithms can be used to form data clusters with semantic associations. For example, the park walking data of multiple users in the same time period may be clustered into the "morning exercise activity in the park" data stream; the location data of the office area, computer usage records and conference room reservation information may be clustered into the "work activity" data stream.

[0028] When mapping the data stream to the virtual space of the metaverse to establish a mapping relationship, the virtual space mapping engine is responsible for converting the data stream into an expression in the virtual space. The specific implementation includes: spatial mapping, time mapping, entity mapping, and interaction mapping. Spatial mapping determines how the physical world's geographical location corresponds to the virtual world's coordinates; time mapping handles the relationship between the physical world's time and the virtual world's time; entity mapping defines how physical world objects are represented in the virtual world; and interaction mapping specifies how physical world behaviors are converted into virtual world interactions. Taking the user's park walking activity as an example, after mapping, the user's virtual image may be shown in the virtual world walking along the same path in the virtual park, with the stop points being enhanced and displayed, the walking speed being consistent with the actual speed, and even the weather and environmental sound effects in reality can be perceived.

[0029] In the mapping process, semantic-level mapping rules can be used to enhance the conversion from the physical world to the virtual world. For example, if an action of "a user raising his arm and waving" is detected in the physical world, not only can the same action be reproduced in the virtual world, but it can also be converted into a social behavior of "a virtual character saying hello" based on the contextual semantics; a semantic scene of multiple people gathering in a conference room in the physical world can automatically create a meeting environment and deploy related tools in the virtual space.

[0030] To ensure the accuracy and real-time nature of the mapping, the system uses an incremental update mechanism to process continuously generated physical world data. When new data arrives, the corresponding elements in the virtual space are quickly updated according to the established mapping rules, processing only the changed parts to reduce unnecessary calculation and transmission overhead.

[0031] Figure 2 This is a flowchart of the mapping relationship between the physical world data and the metaverse virtual space in an embodiment of the present invention: The figure shows a complete process of data processing and mapping. First, starting with terminal identification information and physical world data as input sources, the system classifies and organizes the data according to terminal type and data type based on the terminal identification information to form an initial data set. The initial data set is then divided into two processing branches: one branch extracts data information and generates data feature labels, and then converts these features into semantic description information based on semantic mapping rules; the other branch directly clusters the initial data set to generate semantically associated data streams. The processing results of these two branches are finally converged and mapped to the metaverse virtual space. The entire process reflects the conversion process from physical world data to virtual space. Through semantic mapping and feature extraction, structured data processing and virtual-real mapping are realized, forming a complete data processing and space mapping framework. This processing method not only ensures the integrity and relevance of the data, but also realizes the effective integration of the physical world and virtual space. It is a typical data-driven virtual-real fusion processing process.

[0032] The mapping between the physical world and the virtual world is usually based on direct geometric relationships and predefined rules, and lacks processing at the semantic understanding level. For example, traditional VR / AR systems mainly focus on the geometric correspondence between physical space and virtual space. The user's physical actions are directly mapped to the virtual character after being collected by sensors, lacking understanding and conversion of action semantics. Most existing metaverse platforms use preset rules to convert user behavior into virtual behavior, which makes it difficult to process multimodal data in complex scenarios and cannot achieve semantic association of cross-domain data.

[0033] The starting point of the improvement of this application is to introduce a semantic understanding layer, which elevates the processing of physical world data from pure data classification to the semantic understanding level. By establishing mapping rules between data features and semantic descriptions, a complete conversion link from physical world information to semantic understanding and then to virtual space expression is achieved. This data processing method based on semantic understanding can better understand the actual meaning of user behavior and environmental changes, thereby providing an expression in virtual space that is closer to real semantics.

[0034] The improvement effects of this application are mainly reflected in three aspects: first, the accuracy of physical-virtual mapping is improved through semantic understanding, making the expression in the virtual world more in line with user intentions; second, the semantic-based data clustering method realizes the association of cross-device and cross-modal data, and can integrate information from different terminals to form a more complete situational understanding; finally, the semantic-level mapping mechanism reduces the number of predefined rules, improves the adaptability and scalability of the system, and enables it to handle more diverse physical world inputs.

[0035] In an optional implementation, a two-layer data processing architecture is constructed based on the mapping relationship, in which a state prediction algorithm is used to obtain target state data in the first layer; and in the second layer, a scene collaborative processing unit is set to construct a scene fusion model in combination with the target state data, including: Acquire a historical state sequence of the IoT terminal, wherein the historical state sequence includes a position vector, a velocity vector, and an acceleration vector; Divide the historical state sequence into multiple state sequence segments according to a preset time window size; input the state sequence segments into a long short-term memory network, the long short-term memory network generates a hidden state at the current moment according to the state sequence segments at the current moment and the hidden state at the previous moment, and obtains an initial prediction state according to the hidden state through a linear transformation of a weight matrix and a bias term; Acquire real-time observation data of the IoT terminal, construct a prediction error covariance matrix and an observation noise covariance matrix based on the real-time observation data, and perform a matrix operation on the product of the prediction error covariance matrix and the observation matrix to obtain a Kalman gain; Correcting the initial predicted state according to the Kalman gain specifically includes: calculating the deviation between the real-time observation data and the initial predicted state, and adding the product of the deviation and the Kalman gain to the initial predicted state to obtain corrected target state data; The corrected target state data is sent to the IoT terminal, the state at the next moment is updated, and a scene fusion model is obtained.

[0036] A two-layer data processing architecture is constructed based on the pre-established mapping relationship. The first layer sets up a state prediction algorithm to obtain the target state data. The second layer sets up a scene collaborative processing unit to build a scene fusion model in combination with the target state data.

[0037] Get the historical state sequence of the IoT terminal, including position vector, velocity vector and acceleration vector. For example, for a mobile robot, its historical state sequence can be: Position vector: (x,y,z)=(1,2,3),(2,3,4),(3,4,5); Velocity vector: (vx,vy,vz)=(0.1,0.2,0.3),(0.2,0.3,0.4),(0.3,0.4,0.5); Acceleration vector: (ax,ay,az)=(0.01,0.02,0.03),(0.02,0.03,0.04),(0.03,0.04,0.05).

[0038] The historical state sequence is divided into multiple state sequence segments according to the preset time window size. For example, if the time window size is set to 3, the state sequence segments can be obtained: [(1,2,3), (2,3,4), (3,4,5)].

[0039] The state sequence fragments are input into the long short-term memory network (LSTM). The input layer of the LSTM network contains 3 neurons, corresponding to the three features of position, speed, and acceleration. The hidden layer is set with 64 neurons, and the output layer contains 3 neurons, corresponding to the predicted position, speed, and acceleration at the next moment. LSTM generates the hidden state at the current moment based on the state sequence fragment at the current moment and the hidden state at the previous moment. Specifically, LSTM controls the flow of information through the input gate, forget gate, and output gate to update the cell state and hidden state. Then, LSTM obtains the initial predicted state based on the hidden state through the linear transformation of the weight matrix and the bias term.

[0040] Obtain the real-time observation data of the IoT terminal, and construct the prediction error covariance matrix and the observation noise covariance matrix based on the real-time observation data. For example, suppose the real-time observed position is (3.1, 4.1, 5.1), which deviates from the initial prediction state (3, 4, 5). Based on the historical prediction error, the prediction error covariance matrix can be constructed as a diagonal matrix with diagonal elements of 0.01, 0.01, and 0.01. The observation noise covariance matrix can also be set as a diagonal matrix with diagonal elements of 0.005, 0.005, and 0.005.

[0041] The Kalman gain is obtained by performing matrix operations on the product of the prediction error covariance matrix and the observation matrix. The observation matrix in this case is a 3x3 unit matrix. Through matrix operations, the Kalman gain is obtained as a diagonal matrix with diagonal elements of approximately 0.67, 0.67, 0.67.

[0042] The initial predicted state is corrected according to the Kalman gain. The specific steps include: calculating the deviation between the real-time observation data and the initial predicted state, which is (0.1, 0.1, 0.1) in this case. The product of the deviation and the Kalman gain is superimposed on the initial predicted state to obtain the corrected target state data. The corrected target state data is approximately (3.067, 4.067, 5.067).

[0043] The corrected target state data is sent to the IoT terminal to update the state at the next moment, thereby obtaining a scene fusion model.

[0044] The scene collaborative processing unit receives the corrected target state data from the first layer. For example, the received data is position (3.067, 4.067, 5.067), speed (0.3, 0.4, 0.5), and acceleration (0.03, 0.04, 0.05).

[0045] The scene collaborative processing unit obtains environmental perception data. This data may come from various sensors, such as cameras, lidar, etc. For example, the point cloud data of the surrounding environment is obtained, and an obstacle is identified nearby, with the position (5,6,7).

[0046] The scene collaborative processing unit fuses the target state data with the environmental perception data. Specifically, an occupancy grid map can be used to represent the environment. The current position and predicted trajectory of the IoT terminal are marked on the grid map, and the identified obstacles are also marked on the map.

[0047] Based on the fused data, the scene collaborative processing unit performs path planning and decision making. For example, considering that the predicted trajectory may collide with an obstacle, the system can plan a new path to avoid the obstacle. Assume that the planned new path is a series of position points: (3.067, 4.067, 5.067), (4, 5, 6), (6, 7, 8).

[0048] The scene collaborative processing unit generates control instructions and sends them to the IoT terminal. The control instructions may include speed adjustment, direction change, etc. For example, the generated control instructions are: adjust the speed to (0.2, 0.3, 0.4), and change the direction angle to 30 degrees.

[0049] After receiving the control command, the IoT terminal executes the corresponding action and sends its status data to the first layer again for prediction and correction in the next time step. This cycle repeats to continuously optimize the scene fusion model.

[0050] Through the above two-layer data processing architecture, the present invention realizes accurate prediction of the state of the IoT terminal and dynamic fusion of scenes. The state prediction algorithm of the first layer combines LSTM and Kalman filtering, which can effectively process time series data and correct the prediction results. The scene collaborative processing unit of the second layer combines the predicted state with the environmental perception data to achieve a higher level of scene understanding and decision control. This architectural design enables the system to adapt to complex and changing environments, and improves the intelligence level and operation efficiency of the IoT terminal.

[0051] Figure 3 A flowchart of the dual-layer data processing architecture and scene fusion model construction according to an embodiment of the present invention is as follows: The figure shows a two-layer state prediction and scene co-processing architecture. In the first-layer state prediction algorithm, starting from the mapping relationship, the target state data is obtained through processing and used to build a scene fusion model. The second-layer scene co-processing unit contains two parallel processing paths: the left path first obtains the historical state sequence of the IoT terminal (including position, speed and acceleration information), divides these sequences into state sequence fragments according to the preset time window, and then inputs the long short-term memory network (LSTM) for processing, and finally obtains the initial predicted state through linear transformation; the right path obtains the real-time observation data of the IoT terminal, constructs the prediction error covariance matrix and the observation noise covariance matrix, and calculates the Kalman gain. The processing results of the two paths will be used for deviation calculation, and the initial prediction state will be corrected based on the Kalman gain. Finally, the IoT terminal state is updated and the scene fusion model is generated. This two-layer architecture design not only ensures the accuracy of state prediction, but also realizes the dynamic co-processing of the scene, embodies the organic combination of prediction algorithm and real-time observation data, and constitutes a complete prediction-correction closed-loop system.

[0052] In an optional implementation, constructing a prediction error covariance matrix and an observation noise covariance matrix based on the real-time observation data, and performing a matrix operation on the product of the prediction error covariance matrix and the observation matrix to obtain a Kalman gain includes: Obtaining the real state and predicted state of the IoT terminal, subtracting the predicted state from the real state to obtain a prediction error, multiplying the prediction error by the transpose of the prediction error, calculating an expected value of the result of the transpose multiplication, and obtaining a prediction error covariance matrix; Calculating the deviation between the real-time observation data and the predicted state to obtain an innovation sequence, and performing a cumulative summation operation on the innovation sequence within a preset sliding window to obtain an observation noise covariance matrix; Adding the prediction error covariance matrix to the observation noise covariance matrix to obtain an innovative covariance; dividing the prediction error covariance matrix by the observation noise covariance matrix to obtain an adaptive weight factor; The prediction error covariance matrix is ​​multiplied by the observation noise covariance matrix to obtain a seventh product, the seventh product is multiplied by the inverse matrix of the innovation covariance to obtain an initial Kalman gain, and the initial Kalman gain is multiplied by the adaptive weight factor to obtain a modified Kalman gain.

[0053] Get the real state and predicted state of the IoT terminal. Taking a smartphone as an example, the real state can be the actual location coordinates (x, y, z) obtained by the GPS module, and the predicted state is the location coordinates (x', y', z') predicted based on the state at the previous moment and the motion model.

[0054] Calculate the prediction error by subtracting the predicted state from the actual state. For example, the prediction error is (Δx, Δy, Δz) = (x-x', y-y', z-z'). Then multiply the prediction error by itself to get the square term of the prediction error. For three-dimensional coordinates, 9 square terms can be obtained.

[0055] By finding the expected value of these square terms, we can get the prediction error covariance matrix. In actual operation, we can approximate the expected value by calculating the average value of these square terms through multiple sampling. For example, we can take 100 samples, sum the square terms each time and divide them by 100 to get a 3x3 matrix, which is the prediction error covariance matrix P.

[0056] Calculate the deviation between the real-time observation data and the predicted state to obtain the innovation sequence. Taking smartphone positioning as an example, the real-time observation data can be the position estimate (x_obs, y_obs, z_obs) obtained by converting the base station signal strength, and the innovation sequence is (x_obs-x', y_obs-y', z_obs-z').

[0057] The innovation sequence is cumulatively summed within the preset sliding window to obtain the observation noise covariance matrix R. The size of the sliding window can be set according to actual needs, for example, 10 sampling points are taken. The square sum of the innovation sequences of these 10 points is calculated to obtain a 3x3 matrix R.

[0058] Add the prediction error covariance matrix P to the observation noise covariance matrix R to get the innovation covariance S. This step actually adds the corresponding elements of the two 3x3 matrices to get a new 3x3 matrix S.

[0059] Calculate the adaptive weight factor. Divide each element of the prediction error covariance matrix P by the corresponding element of the observation noise covariance matrix R to obtain a new 3x3 matrix, which is the adaptive weight factor W.

[0060] Multiply the prediction error covariance matrix P by the observation matrix H to obtain an intermediate result matrix. The observation matrix H reflects the relationship between the state variables and the observed quantities, and in this case it can be simplified to a 3x3 unit matrix. Therefore, the result of this step is still the matrix P.

[0061] Multiply the intermediate result matrix obtained in the previous step by the inverse matrix of the innovation covariance S to obtain the initial Kalman gain K. Matrix inversion can be achieved through numerical calculation methods such as Gauss-Jordan elimination. The multiplication operation is a standard matrix multiplication, and a new 3x3 matrix K is obtained.

[0062] The initial Kalman gain K is multiplied by the corresponding element of the adaptive weight factor W to obtain the modified Kalman gain K'. This step is to adaptively adjust the Kalman gain to adapt to different observation noise environments.

[0063] The following is a specific data case to illustrate the above process: Suppose at a certain moment, the actual position of the smartphone is (10, 20, 30), the predicted position is (9, 21, 31), and the observed position is (11, 19, 29).

[0064] Calculate the prediction error: (1, -1, -1); Squared prediction error term: ((1, -1, -1), (-1, 1, 1), (-1, 1, 1)); Assume that the prediction error covariance matrix P obtained after multiple samplings is: ((1.2, -0.1, -0.1), (-0.1, 1.1, 0.1), (-0.1, 0.1, 1.3)); Innovation sequence: (2, -2, -2); Assume that the observed noise covariance matrix R accumulated in a 10-point sliding window is: ((0.8, 0, 0), (0, 0.9, 0), (0, 0, 1.0)); Innovation covariance S: ((2.0, -0.1, -0.1), (-0.1, 2.0, 0.1), (-0.1, 0.1, 2.3)); Adaptive weight factor W: ((1.5, -0.11, -0.1), (-0.11, 1.22, 0.1), (-0.1, 0.11, 1.3)); Initial Kalman gain K (assumed to be obtained by numerical calculation): ((0.6, -0.03, -0.026), (-0.03, 0.55, 0.022), (-0.026, 0.022, 0.565)); Corrected Kalman gain K': ((0.9, -0.0033, -0.0026), (-0.0033, 0.671, 0.0022), (-0.0026,0.00242, 0.7345)); In an optional implementation, the network quality of each priority data stream is monitored in real time, and when network fluctuation is detected, a data compensation mechanism is started, and compensation processing is performed on different transmission priorities based on the prediction result of the state prediction algorithm, including: Constructing a network state vector including end-to-end delay and packet loss rate, adding the product of the network state vector and the smoothing factor to the product of the smoothing state value at the previous moment and the smoothing factor to obtain the smoothing state value at the current moment; Subtracting the smooth state value at the previous moment from the smooth state value at the current moment to obtain a difference vector, calculating the second norm of the difference vector to obtain a fluctuation detection value, adding the smooth state value at the current moment to the product of the fluctuation detection value and a preset adjustment coefficient, and constructing a multi-level fluctuation threshold; Inputting the fluctuation detection value and the multi-level fluctuation threshold into a compensation mapping function to calculate a compensation intensity; Based on the packet loss rate divided by the compensation strength to obtain a third product, the fluctuation detection value is multiplied by a preset fluctuation compensation coefficient to obtain a fourth product, the third product is added to the fourth product and rounded up to obtain the number of first priority compensated data packets; Dividing the end-to-end delay by the product of the fluctuation detection value and the preset adjustment coefficient to obtain a calculation result; comparing the calculation result with the minimum window value, and taking the product greater than the preset window value threshold as the compensation time window; Compensation operations are performed on data streams of different priorities according to the number of the first priority compensation data packets, the number of redundant packets, and the compensation time window.

[0065] Construct a network state vector. This vector contains two indicators: end-to-end delay and packet loss rate. For example, if the end-to-end delay at the current moment is 50ms and the packet loss rate is 2%, the network state vector can be expressed as [50, 2].

[0066] Next, calculate the smoothed state value. Select the smoothing factor α, for example, α = 0.8. Assuming that the smoothed state value at the previous moment is [48, 1.8], the smoothed state value at the current moment is calculated as follows: Current smoothing state value = α * [50, 2] + (1-α) * [48, 1.8] = [49.6, 1.96]; Calculate the fluctuation detection value. Subtract the previous smooth state value from the current smooth state value to obtain the difference vector [1.6, 0.16]. Calculate the second norm of this vector, that is, the fluctuation detection value is 1.61.

[0067] Construct a multi-level fluctuation threshold. Assuming the preset adjustment coefficient β = 0.5, the multi-level fluctuation threshold is calculated as follows: Multi-level fluctuation threshold = [49.6, 1.96] + 1.61 * 0.5 = [50.405, 2.765]; The fluctuation detection value and the multi-level fluctuation threshold are input into the compensation mapping function to calculate the compensation strength. The compensation mapping function can be a piecewise function that returns different compensation strengths according to the fluctuation detection value. For example, when the fluctuation detection value is less than 1, the compensation strength is 1; when the fluctuation detection value is between 1 and 2, the compensation strength is 1.5; when the fluctuation detection value is greater than 2, the compensation strength is 2. In this example, the fluctuation detection value is 1.61, and the corresponding compensation strength is 1.5.

[0068] Calculate the number of first priority compensation packets. Assuming the preset fluctuation compensation coefficient γ=0.8, then: Number of first priority compensation packets = ceil((2% / 1.5) + (1.61 * 0.8)) = ceil(1.33 + 1.288) = 3; Calculate the compensation time window. Assuming the minimum window value is 100ms and the preset window value threshold is 1.2, then: Calculation result = 50ms / (1.61 * 0.5) = 62.11ms; Compensation time window = max(62.11ms, 100ms) * 1.2 = 120ms; Based on the calculated number of compensation packets and compensation time window, compensation operations are performed on data streams of different priorities.

[0069] For the first priority data flow, three additional redundant data packets are sent within the 120ms time window.

[0070] For the second priority data stream, the compensation intensity can be reduced according to the actual situation, for example, two additional redundant data packets are sent within the 120ms time window.

[0071] For the third priority data stream, the compensation intensity can be further reduced, for example, an additional redundant data packet is sent within the 120ms time window.

[0072] Differentiated compensation processing can be performed on data streams of different priorities according to the real-time fluctuations in network quality, thereby improving the overall transmission quality and reliability.

[0073] The parameters can be adjusted according to specific scenarios and requirements. For example, the value of the smoothing factor α can be adjusted according to the stability of the network environment; the value of the preset adjustment coefficient β can be adjusted according to the sensitivity to network fluctuations; and the value of the preset fluctuation compensation coefficient γ can be adjusted according to the business requirements for transmission reliability.

[0074] The compensation mapping function can also be customized according to actual needs. For example, a more complex nonlinear function can be used to more accurately reflect the relationship between network fluctuations and compensation intensity.

[0075] In an optional implementation, dividing the end-to-end delay by the product of the fluctuation detection value and the preset adjustment coefficient to obtain a calculation result; comparing the calculation result with the minimum window value, and taking the product greater than the preset window value threshold as the compensation time window includes: Performing a division operation using the product of the end-to-end delay, the fluctuation detection value and the preset adjustment coefficient as a divisor to obtain an initial compensation time; Compare the initial compensation time with a preset minimum window value, and when the initial compensation time is greater than the preset minimum window value, use the initial compensation time as a reference compensation window value; Multiplying the reference compensation window value by a preset smoothing coefficient to obtain a first product, multiplying the compensation window value at a previous moment by a complement of the preset smoothing coefficient to obtain a second product, and adding the first product to the second product to obtain a smoothing compensation window value; Multiplying the smoothing compensation window value by the reference compensation window value to obtain a correction compensation window value; calculating a current resource utilization rate and a weighted service quality value based on the correction compensation window value; The objective function is constructed by subtracting the product of the current resource utilization and the preset balance factor from the weighted service quality value, and the optimal compensation window value is obtained by solving the maximum value of the objective function under the constraint that the correction compensation window value is between the preset minimum window value and the preset maximum window value; A performance evaluation value is calculated based on the optimal compensation window value, and the preset adjustment coefficient is updated according to the performance evaluation value; and the optimal compensation window value is used as a final compensation time window value.

[0076] Obtain the end-to-end delay, the fluctuation detection value, and the preset adjustment coefficient. The end-to-end delay can be measured by a network test tool, for example, it may be 100ms. The fluctuation detection value can be obtained by statistically analyzing the network jitter, for example, it may be 0.2. The preset adjustment coefficient can be set based on experience, for example, it may be 1.5.

[0077] The end-to-end delay is divided by the product of the fluctuation detection value and the preset adjustment coefficient to obtain the initial compensation time. Specifically, 100ms is divided by the product of 0.2 and 1.5 to obtain an initial compensation time of 333.33ms.

[0078] The initial compensation time is compared with a preset minimum window value. The preset minimum window value can be set according to system requirements, for example, it may be 200ms. Since 333.33ms is greater than 200ms, 333.33ms is used as the benchmark compensation window value.

[0079] Smoothing is introduced. Set the preset smoothing coefficient to 0.8, and its complement is 0.2. Multiply the reference compensation window value 333.33ms by 0.8 to get the first product 266.66ms, multiply the previous moment compensation window value (assuming it is 300ms) by 0.2 to get the second product 60ms, and add the two to get the smooth compensation window value 326.66ms.

[0080] For further correction, the smoothing compensation window value is multiplied by the reference compensation window value to obtain the correction compensation window value of 108888.89ms².

[0081] Based on the correction compensation window value, calculate the current resource utilization and weighted service quality value. Assume that the current resource utilization is 0.7 and the weighted service quality value is 0.9.

[0082] Construct the objective function: subtract the product of the current resource utilization rate of 0.7 and the preset balance factor (assuming 0.5) from the weighted service quality value of 0.9, and obtain the objective function value of 0.55. Under the constraint that the correction compensation window value is between the preset minimum window value of 200ms and the preset maximum window value (assuming 500ms), the maximum value of the objective function is solved by the iterative optimization algorithm to obtain the optimal compensation window value, which may be 400ms.

[0083] Based on the optimal compensation window value of 400ms, the performance evaluation value is calculated. The resource utilization and service quality indicators can be comprehensively evaluated by weighted average to obtain a performance evaluation value, which may be 0.85, for example.

[0084] According to the performance evaluation value of 0.85, the preset adjustment coefficient is updated. An adaptive adjustment strategy can be adopted to appropriately reduce the adjustment coefficient when the performance evaluation value is higher than expected, and increase it otherwise. For example, the adjustment coefficient can be adjusted from 1.5 to 1.4.

[0085] The optimal compensation window value of 400ms is used as the final compensation time window value.

[0086] In order to improve the computational efficiency, parallel processing technology can be used. For example, while calculating the initial compensation time, the fluctuation detection value can be updated in parallel. In addition, cache technology can be used to store intermediate calculation results to avoid repeated calculations.

[0087] This method can still maintain stability under extreme network conditions. For example, when the end-to-end delay surges to 1000ms due to severe network congestion, a reasonable compensation time window can still be obtained by adjusting the fluctuation detection value and the adjustment coefficient to avoid system overreaction.

[0088] This method can also be combined with machine learning technology to train models with historical data and predict the optimal compensation window value. For example, a long short-term memory network (LSTM) model can be used to input historical network state data and output the predicted compensation window value, further improving the calculation accuracy and efficiency.

[0089] During the implementation process, attention should be paid to numerical precision. It is recommended to use double-precision floating-point numbers for calculations to ensure the accuracy of the calculation results. At the same time, appropriate exception handling should be performed for possible division by zero errors.

[0090] In order to facilitate system management and tuning, a visual interface can be designed to display the changing trends of various parameters in real time. For example, a line chart can be used to display the historical changes of indicators such as end-to-end delay, fluctuation detection value, and compensation time window to help administrators intuitively understand the system status.

[0091] In a multi-user scenario, different service quality weights can be set for different users to achieve differentiated compensation time window calculations. For example, for paying users, a higher service quality weight can be set to obtain a better compensation time window.

[0092] This method can also be used in combination with other network optimization technologies, such as congestion control, load balancing, etc., to form a comprehensive network optimization solution. For example, the congestion window size can be dynamically adjusted according to the calculated compensation time window to achieve more sophisticated flow control.

[0093] A distributed architecture can be used to distribute computing tasks to multiple nodes for execution, improving the scalability and fault tolerance of the system. For example, a distributed computing framework such as Hadoop can be used to parallelize large-scale compensation time window computing tasks.

[0094] This method can effectively improve network performance and user experience by dynamically calculating the optimal compensation time window by comprehensively considering multiple factors such as end-to-end delay, network fluctuations, resource utilization, and service quality. By introducing technologies such as smoothing, adaptive adjustment, and machine learning, the robustness and adaptability of the method are further enhanced, making it suitable for various complex network environments.

[0095] According to a second aspect of the embodiments of the present invention, Provide a low-latency transmission system that realizes the virtual-real integration of intelligent IoT and Metaverse, including: A first unit is used to obtain physical world data sent by multiple smart IoT terminals, where the physical world data includes terminal identification information; The second unit is used to classify and sort the physical world data according to the terminal identification information, generate a corresponding data stream, and establish a mapping relationship between the data stream and the metaverse virtual space; The third unit is used to construct a two-layer data processing architecture based on the mapping relationship, wherein a state prediction algorithm is used to obtain target state data in the first layer; and a scene collaborative processing unit is set in the second layer to construct a scene fusion model in combination with the target state data; The fourth unit is used to set different transmission priorities for the target status data and the scene fusion model, divide the target status data into a first priority data stream, and divide the scene fusion model into a second priority data stream; monitor the network quality of each priority data stream in real time, and when network fluctuations are detected, start the data compensation mechanism to compensate for different transmission priorities based on the prediction results of the state prediction algorithm.

[0096] According to a third aspect of the embodiments of the present invention, An electronic device is provided, comprising: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0097] A fourth aspect of the embodiments of the present invention is: A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0098] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A low-latency transmission method for realizing the virtual-real integration of intelligent IoT and metaverse, characterized in that: include: Acquire physical world data sent by multiple smart IoT terminals, where the physical world data includes terminal identification information; Classify and sort the physical world data according to the terminal identification information, generate corresponding data streams, and establish a mapping relationship between the data streams and the metaverse virtual space; A two-layer data processing architecture is constructed based on the mapping relationship, in which a state prediction algorithm is used to obtain target state data in the first layer; a scene collaborative processing unit is set in the second layer to construct a scene fusion model in combination with the target state data; Setting different transmission priorities for the target state data and the scene fusion model, dividing the target state data into a first priority data stream, and dividing the scene fusion model into a second priority data stream; The network quality of each priority data stream is monitored in real time. When network fluctuation is detected, a data compensation mechanism is started to perform compensation processing on different transmission priorities based on the prediction results of the state prediction algorithm.

2. The method according to claim 1, characterized in that Classifying and arranging the physical world data according to the terminal identification information, generating corresponding data streams, and establishing a mapping relationship between the data streams and the metaverse virtual space includes: Classifying the physical world data according to the terminal type and data type according to the terminal identification information to obtain an initial data set; Extracting data information from the initial data set to generate data feature labels, converting the data feature labels into corresponding semantic description information based on a preset semantic mapping rule, clustering the data in the initial data set according to the semantic description information, and generating a data stream with semantic association; The data stream is mapped to the metaverse virtual space to establish a mapping relationship between the physical world and the virtual space.

3. The method according to claim 1, characterized in that A two-layer data processing architecture is constructed based on the mapping relationship, in which a state prediction algorithm is used to obtain target state data in the first layer; and a scene collaborative processing unit is set in the second layer to construct a scene fusion model in combination with the target state data, including: Acquire a historical state sequence of the IoT terminal, wherein the historical state sequence includes a position vector, a velocity vector, and an acceleration vector; Divide the historical state sequence into multiple state sequence segments according to a preset time window size; input the state sequence segments into a long short-term memory network, the long short-term memory network generates a hidden state at the current moment according to the state sequence segments at the current moment and the hidden state at the previous moment, and obtains an initial prediction state according to the hidden state through a linear transformation of a weight matrix and a bias term; Acquire real-time observation data of the IoT terminal, construct a prediction error covariance matrix and an observation noise covariance matrix based on the real-time observation data, and perform a matrix operation on the product of the prediction error covariance matrix and the observation matrix to obtain a Kalman gain; Correcting the initial predicted state according to the Kalman gain specifically includes: calculating the deviation between the real-time observation data and the initial predicted state, and adding the product of the deviation and the Kalman gain to the initial predicted state to obtain corrected target state data; The corrected target state data is sent to the IoT terminal, the state at the next moment is updated, and a scene fusion model is obtained.

4. The method according to claim 3, characterized in that Constructing a prediction error covariance matrix and an observation noise covariance matrix based on the real-time observation data, and performing a matrix operation on the product of the prediction error covariance matrix and the observation matrix to obtain a Kalman gain includes: Obtaining the real state and predicted state of the IoT terminal, subtracting the predicted state from the real state to obtain a prediction error, multiplying the prediction error by the transpose of the prediction error, calculating an expected value of the result of the transpose multiplication, and obtaining a prediction error covariance matrix; Calculating the deviation between the real-time observation data and the predicted state to obtain an innovation sequence, and performing a cumulative summation operation on the innovation sequence within a preset sliding window to obtain an observation noise covariance matrix; Adding the prediction error covariance matrix to the observation noise covariance matrix to obtain an innovative covariance; dividing the prediction error covariance matrix by the observation noise covariance matrix to obtain an adaptive weight factor; The prediction error covariance matrix is ​​multiplied by the observation noise covariance matrix to obtain a seventh product, the seventh product is multiplied by the inverse matrix of the innovation covariance to obtain an initial Kalman gain, and the initial Kalman gain is multiplied by the adaptive weight factor to obtain a modified Kalman gain.

5. The method according to claim 1, characterized in that Real-time monitoring of the network quality of each priority data stream, when a network fluctuation is detected, starting a data compensation mechanism, and performing compensation processing on different transmission priorities based on the prediction result of the state prediction algorithm includes: Constructing a network state vector including end-to-end delay and packet loss rate, adding the product of the network state vector and the smoothing factor to the product of the smoothing state value at the previous moment and the smoothing factor to obtain the smoothing state value at the current moment; Subtracting the smooth state value at the previous moment from the smooth state value at the current moment to obtain a difference vector, calculating the second norm of the difference vector to obtain a fluctuation detection value, adding the smooth state value at the current moment to the product of the fluctuation detection value and a preset adjustment coefficient, and constructing a multi-level fluctuation threshold; Inputting the fluctuation detection value and the multi-level fluctuation threshold into a compensation mapping function to calculate a compensation intensity; Based on the packet loss rate divided by the compensation strength to obtain a third product, the fluctuation detection value is multiplied by a preset fluctuation compensation coefficient to obtain a fourth product, the third product is added to the fourth product and rounded up to obtain the number of first priority compensated data packets; Dividing the end-to-end delay by the product of the fluctuation detection value and the preset adjustment coefficient to obtain a calculation result; comparing the calculation result with the minimum window value, and taking the product greater than the preset window value threshold as the compensation time window; Compensation operations are performed on data streams of different priorities according to the number of the first priority compensation data packets, the number of redundant packets, and the compensation time window.

6. The method according to claim 5, characterized in that Dividing the end-to-end delay by the product of the fluctuation detection value and the preset adjustment coefficient to obtain a calculation result; Comparing the calculation result with the minimum window value, and taking the product greater than the preset window value threshold as the compensation time window includes: Performing a division operation using the product of the end-to-end delay, the fluctuation detection value and the preset adjustment coefficient as a divisor to obtain an initial compensation time; Compare the initial compensation time with a preset minimum window value, and when the initial compensation time is greater than the preset minimum window value, use the initial compensation time as a reference compensation window value; Multiplying the reference compensation window value by a preset smoothing coefficient to obtain a first product, multiplying the compensation window value at a previous moment by a complement of the preset smoothing coefficient to obtain a second product, and adding the first product to the second product to obtain a smoothing compensation window value; Multiplying the smoothing compensation window value by the reference compensation window value to obtain a correction compensation window value; calculating a current resource utilization rate and a weighted service quality value based on the correction compensation window value; The objective function is constructed by subtracting the product of the current resource utilization and the preset balance factor from the weighted service quality value, and under the constraint that the correction compensation window value is between the preset minimum window value and the preset maximum window value, the maximum value of the objective function is solved to obtain the optimal compensation window value; A performance evaluation value is calculated based on the optimal compensation window value, and the preset adjustment coefficient is updated according to the performance evaluation value; and the optimal compensation window value is used as a final compensation time window value.

7. A low-latency transmission system for realizing the virtual-real integration of intelligent Internet of Things and Metaverse, used to implement the method described in any one of claims 1 to 6, characterized in that: include: A first unit is used to obtain physical world data sent by multiple smart IoT terminals, where the physical world data includes terminal identification information; The second unit is used to classify and sort the physical world data according to the terminal identification information, generate a corresponding data stream, and establish a mapping relationship between the data stream and the metaverse virtual space; The third unit is used to construct a two-layer data processing architecture based on the mapping relationship, wherein a state prediction algorithm is used to obtain target state data in the first layer; and a scene collaborative processing unit is set in the second layer to construct a scene fusion model in combination with the target state data; A fourth unit is used to set different transmission priorities for the target state data and the scene fusion model, divide the target state data into a first priority data stream, and divide the scene fusion model into a second priority data stream; The network quality of each priority data stream is monitored in real time. When network fluctuation is detected, a data compensation mechanism is started to perform compensation processing on different transmission priorities based on the prediction results of the state prediction algorithm.

8. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images