A traffic-aware based industrial internet of things data processing method and system

Abstract

The application relates to the technical field of communication, in particular to an industrial internet of things data processing method and system based on traffic awareness, which comprises the following steps: obtaining data traffic information and corresponding data content of each device node in an industrial internet of things in a recent time period; extracting feature parameters of the data traffic information, and constructing a traffic feature matrix based on the feature parameters; determining a traffic load state of a current network based on the traffic feature matrix, and determining a priority queue of processing data content and a calculation resource allocation proportion according to the traffic load state and a priority attribute of the data content; processing the data content according to the priority queue and the calculation resource allocation proportion, and feeding back a processing result to a corresponding target device node; and the application can effectively improve the efficiency and reliability of industrial internet of things data processing.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and specifically to an industrial Internet of Things (IoT) data processing method and system based on flow awareness. Background Technology

[0002] In an Industrial Internet of Things (IIoT) environment, numerous devices interact via networks, generating data streams characterized by their massive scale, diverse types, and high real-time requirements. Traditional data processing methods often lack the ability to dynamically perceive network traffic conditions, making it difficult to flexibly adjust data processing strategies based on traffic changes. This leads to problems such as data congestion and increased processing latency during peak traffic periods, while potentially wasting computing resources during off-peak periods, failing to fully leverage the system's performance advantages. Furthermore, different types of industrial data have varying requirements for transmission quality and processing priority. Existing methods fail to effectively combine traffic awareness with data characteristics for differentiated processing, impacting data processing efficiency and reliability, and making it difficult to meet the application needs of complex IIoT scenarios. Summary of the Invention

[0003] To address the aforementioned issues, this invention provides an industrial IoT data processing method and system based on traffic awareness, aiming to improve the dynamic adaptability and resource utilization efficiency of industrial IoT data processing. By accurately sensing network traffic status and combining it with the priority attributes of data content, it achieves intelligent scheduling of data processing tasks and optimized allocation of computing resources.

[0004] To achieve the above objectives, the present invention provides the following technical solution: On one hand, embodiments of the present invention provide an industrial IoT data processing method based on flow awareness, the method comprising the following steps: S100: Obtain data flow information and corresponding data content of each device node in the Industrial Internet of Things within the most recent time period; S200, extract the feature parameters of the data traffic information, and construct a traffic feature matrix based on the feature parameters; the feature parameters include the mean feature, variance feature, peak feature and valley feature of the standardized traffic values ​​in each time window; S300, determine the current network traffic load status based on the traffic feature matrix, and determine the priority queue for processing data content and the computing resource allocation ratio according to the traffic load status and the priority attributes of data content. S400: The data content is processed according to the priority queue and the computing resource allocation ratio, and the processing result is fed back to the corresponding destination device node.

[0005] Optionally, in S100, obtaining the data flow information and corresponding data content of each device node in the industrial Internet of Things includes: S110, acquire the raw traffic data frames sent by each device node; S120, parse the original traffic data frame, extract the data transmission timestamp, data transmission rate, data packet size, traffic source device identifier and destination device identifier to form data traffic information; S130, identify the type of data content in the original traffic data frame, associate and store the data traffic information with the corresponding data content, and establish a mapping table containing traffic information and data content.

[0006] Optionally, in S200, the step of extracting feature parameters of the data traffic information and constructing a traffic feature matrix based on the feature parameters includes: S210, standardizes real-time traffic data to obtain standardized traffic values; S220, Set the time window length and sliding step size, extract standardized flow values ​​sequentially according to the sliding step size, and generate multiple flow datasets corresponding to continuous and non-overlapping time windows. Each flow dataset contains the standardized flow value of each node in the corresponding time window and the collection timestamp. S230, traverse the traffic dataset within each time window, perform feature extraction operation for each dataset, and obtain feature parameters; the feature parameters include the mean feature, variance feature, peak feature, and valley feature of the standardized traffic values ​​within each time window; S240, arrange the feature parameters of the same node in different time windows in sequence to form the feature parameter sequence of the node; concatenate the feature parameter sequences of all nodes horizontally according to the node identifier order to construct a traffic feature matrix. The row dimension of the traffic feature matrix corresponds to the time window number, the column dimension corresponds to the node identifier, and the depth dimension corresponds to the feature parameter type.

[0007] Optionally, in S300, determining the current network traffic load status based on the traffic feature matrix, and determining the priority queue for processing data content and the allocation ratio of computing resources according to the traffic load status and the priority attributes of the data content, includes: S310, Dimensionality reduction is performed on the traffic feature matrix, principal component features are extracted from the traffic feature matrix, eigenvalues ​​and eigenvectors of each principal component are calculated, and principal components with eigenvalues ​​greater than a preset threshold are selected as key feature dimensions. S320: Calculate the correlation coefficient of different nodes on key feature dimensions, and divide each node into multiple node traffic association groups based on the correlation coefficient. S330, calculate the overall traffic characteristics of each node traffic association group in the current time window, input the overall traffic characteristics into the traffic prediction sub-model, and obtain the short-term traffic change trend of the corresponding node traffic association group; the overall traffic characteristics include average traffic value, traffic fluctuation variance and peak frequency. S340, based on the short-term traffic change trend of each node traffic association group and the preset traffic load threshold, determine the current network traffic load status, which includes light load status, normal status and heavy load status. S350, Obtain the priority attribute of the data content; for node traffic association groups in a heavily loaded state, increase the priority of their data content by one level; for node traffic association groups in a lightly loaded state, decrease the priority of their data content by one level; the priority attribute is determined according to the type of data content and real-time requirements. S360 sorts all data content based on the adjusted priority attributes, constructs a priority queue for processing the data content, and determines the allocation ratio of computing resources according to the traffic load status of each node's traffic association group.

[0008] Optionally, in S330, calculating the overall traffic characteristics of each node traffic association group in the current time window includes: Iterate through the standardized flow values ​​of all nodes in the node flow association group within the current time window, calculate the arithmetic mean of the standardized flow values ​​of each node in the association group, and obtain the overall average flow value. The variance of traffic fluctuation for each node in the associated group within the current time window is calculated. The average of the sums of the variances of all nodes is then taken to obtain the overall variance of traffic fluctuation. Identify the number of times the standardized traffic value of each node in the associated group exceeds the preset peak threshold within the current time window, sum up the peak occurrence times of each node and divide by the number of nodes to obtain the overall peak occurrence frequency.

[0009] Optionally, in S400, the step of processing the data content according to the priority queue and the computing resource allocation ratio, and feeding back the processing result to the corresponding destination device node, includes: S410 allocates corresponding computing resources to each data item according to the priority queue order, and processes high-priority data first. S420 executes a preset data processing algorithm on the data content allocated to computing resources, including data cleaning, format conversion and feature extraction; S430 matches the processed data results with the corresponding destination device node information and generates a result feedback command; The S440 sends result feedback instructions to the target device node through the Industrial Internet of Things (IIoT) communication protocol, and records the data processing completion time and result status.

[0010] Optionally, the method further includes: The load of each data processing node is monitored in real time. When the load of a node exceeds a preset threshold, some of the data processing tasks of that node are migrated to other nodes with loads below the preset threshold.

[0011] On the other hand, embodiments of the present invention provide a flow-aware industrial IoT data processing system, comprising: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor performs the method described above.

[0012] On the other hand, embodiments of the present invention provide a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to perform the above-described method.

[0013] The beneficial effects of this invention are as follows: This invention discloses an industrial IoT data processing method and system based on traffic awareness. This invention acquires traffic information and data content from device nodes, extracts multi-dimensional traffic features to construct a matrix, and thereby determines the network load status. It dynamically adjusts data priorities based on the load status and allocates computing resources according to priority queues, ensuring that high-priority data is processed first during peak traffic periods to avoid congestion, while fully utilizing resources during off-peak periods to reduce waste. Furthermore, by dividing node traffic association groups and predicting short-term traffic trends, the system can adjust processing strategies in advance, enhancing the foresight and reliability of data processing. This effectively solves problems such as lack of traffic awareness, rigid resource allocation, and insufficient data processing efficiency in traditional methods, providing efficient and stable technical support for large-scale data interaction in complex industrial IoT scenarios. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a flowchart illustrating an industrial IoT data processing method based on flow perception, as described in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of an industrial IoT data processing system based on flow perception, according to an embodiment of the present invention. Detailed Implementation

[0016] The following will provide a clear and complete description of the concept, specific structure, and technical effects of the present invention in conjunction with embodiments and accompanying drawings, so as to fully understand the purpose, solution, and effects of the present invention. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of this invention; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this invention as detailed in the appended claims.

[0018] It is understood that the terms “first,” “second,” etc., used in this invention may be used herein to describe various concepts, but unless specifically stated otherwise, these concepts are not limited by these terms. These terms are used only to distinguish one concept from another. For example, first information may also be referred to as second information without departing from the scope of embodiments of the invention, and similarly, second information may also be referred to as first information. Depending on the context, the words “if,” “when,” or “in response to determination” as used herein may be interpreted as “when…” or “when…” or “in response to determination.”

[0019] The terms “at least one,” “multiple,” “each,” “any,” etc., used in this invention, “at least one” includes one, two, or more than two; “multiple” includes two or more than two; “each” refers to each of the corresponding multiple; and “any” refers to any one of the multiple.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0021] refer to Figure 1 ,like Figure 1 The figure shown is an industrial IoT data processing method based on flow awareness provided by an embodiment of the present invention. The method includes the following steps: S100: Obtain data flow information and corresponding data content of each device node in the Industrial Internet of Things within the most recent time period; The data traffic information includes data transmission rate, data packet size, traffic source device identifier and destination device identifier, and the data content includes industrial control command data, historical data and log data. S200, extract the feature parameters of the data traffic information, and construct a traffic feature matrix based on the feature parameters; the feature parameters include the mean feature, variance feature, peak feature and valley feature of the standardized traffic values ​​in each time window; S300, determine the current network traffic load status based on the traffic feature matrix, and determine the priority queue for processing data content and the computing resource allocation ratio according to the traffic load status and the priority attributes of data content. S400: The data content is processed according to the priority queue and the computing resource allocation ratio, and the processing result is fed back to the corresponding destination device node.

[0022] In steps S100 to S400 of this embodiment, the real-time traffic dynamics and data content characteristics of device nodes are accurately captured, laying the foundation for subsequent traffic analysis and resource scheduling. Through standardized processing of traffic data and multi-dimensional feature extraction, the constructed traffic feature matrix comprehensively reflects the spatiotemporal distribution patterns of network traffic, providing a quantitative basis for judging traffic load status. Combining the division of node traffic association groups with short-term traffic trend prediction, accurate identification of network load status is achieved, and resource allocation is made more aligned with actual needs by dynamically adjusting data priorities. Through the orderly execution of priority queues and differentiated allocation of computing resources, high-priority data is ensured to be processed first, while resource utilization is further optimized through a task migration mechanism, effectively improving the overall efficiency and system stability of industrial IoT data processing.

[0023] In some embodiments, S100, obtaining the data traffic information and corresponding data content of each device node in the industrial Internet of Things includes: S110, acquire the raw traffic data frames sent by each device node; S120, parse the original traffic data frame, extract the data transmission timestamp, data transmission rate, data packet size, traffic source device identifier and destination device identifier to form data traffic information; S130, identify the type of data content in the original traffic data frame, associate and store the data traffic information with the corresponding data content, and establish a mapping table containing traffic information and data content.

[0024] Specifically, the traffic acquisition module deployed at the network nodes monitors each device node in the Industrial Internet of Things (IIoT) in real time, collecting the raw traffic data frames sent by each device node; the data parsing unit parses the raw traffic data frames, extracting the data transmission timestamp, data transmission rate, data packet size, traffic source device identifier, and destination device identifier, and integrates this information into data traffic information; the data classification unit identifies the data content in the raw traffic data frames according to preset data type identification rules, determining whether it belongs to industrial control command data, historical data, or log data; the data association and storage unit associates and binds the data traffic information with the corresponding identified data content and stores it in the database, while establishing a mapping table containing traffic information and data content to achieve rapid query and matching of traffic information and data content.

[0025] In some embodiments, in S200, extracting feature parameters of the data traffic information and constructing a traffic feature matrix based on the feature parameters includes: S210, standardizes real-time traffic data to obtain standardized traffic values; Through standardization, the traffic value of each node is converted to a preset value range. The standardization formula is: Standardized traffic value = (Original traffic value - Minimum traffic value) / (Maximum traffic value - Minimum traffic value). S220, Set the time window length and sliding step size, extract standardized flow values ​​sequentially according to the sliding step size, and generate multiple flow datasets corresponding to continuous and non-overlapping time windows. Each flow dataset contains the standardized flow value of each node in the corresponding time window and the collection timestamp. Specifically, a correspondence between time windows and traffic datasets is established by adding time window identifiers and corresponding node identifiers to each traffic dataset. The standardized traffic value sample represents the standardized device node traffic data within a specific time window. This sample preserves the relative magnitudes of the original traffic data while eliminating feature interference caused by differences in data range between different nodes. For example, when the original traffic value is between the minimum and maximum traffic values, the standardized traffic value is mapped to a decimal between 0 and 1. If the original traffic value equals the minimum traffic value, the standardization result is 0; if it equals the maximum traffic value, the standardization result is 1, facilitating subsequent cross-node traffic feature comparison and matrix construction.

[0026] S230: Iterate through the traffic dataset within each time window, perform feature extraction for each dataset, and obtain feature parameters; Specifically, the average flow rate for the time window is obtained by summing all standardized flow rates within the time window and dividing by the number of data samples. The variance of the flow rate fluctuation within the time window is obtained by dividing the sum of squared differences between each standardized flow rate and the mean by the number of data samples. The standardized flow rates within the window are compared, and the flow rate with the largest value is selected as the peak value, and the flow rate with the smallest value is selected as the trough value. The mean reflects the average flow rate level within the time window, the variance reflects the dispersion of the flow data, the number of peak values ​​counts the number of times the flow rate exceeds the preset flow threshold within the window, and the flow rate change rate is obtained by calculating the ratio of the difference between the standardized flow rates at the beginning and end of the window to the window duration. The extracted feature parameters are arranged according to the time window order and node identifier to form a flow feature matrix with dimensions of (number of nodes × number of time windows × feature dimension). Each element in the matrix corresponds to a specific feature parameter value of a specific node within a specific time window, providing a structured data foundation for the input of the subsequent flow prediction model.

[0027] S240, arrange the feature parameters of the same node in different time windows in sequence to form the feature parameter sequence of the node; concatenate the feature parameter sequences of all nodes horizontally according to the node identifier order to construct a traffic feature matrix. The row dimension of the traffic feature matrix corresponds to the time window number, the column dimension corresponds to the node identifier, and the depth dimension corresponds to the feature parameter type.

[0028] Specifically, the feature parameters of the same node in different time windows are arranged sequentially to form the feature parameter sequence of the node; the feature parameter sequences of all nodes are horizontally spliced ​​according to the node identifier order to construct a three-dimensional traffic feature matrix with dimensions of (number of time windows × number of nodes × feature dimension). The row dimension of the matrix corresponds to the time window number, the column dimension corresponds to the node identifier, and the depth dimension corresponds to the feature parameter type, so as to realize the structured representation of the traffic features of multiple nodes and multiple time windows in the Industrial Internet of Things.

[0029] In some embodiments, S300, determining the current network traffic load status based on the traffic feature matrix, and determining the priority queue and computing resource allocation ratio for processing data content based on the traffic load status and the priority attributes of the data content, includes: S310, Dimensionality reduction is performed on the traffic feature matrix, principal component features are extracted from the traffic feature matrix, eigenvalues ​​and eigenvectors of each principal component are calculated, and principal components with eigenvalues ​​greater than a preset threshold are selected as key feature dimensions. S320: Calculate the correlation coefficient of different nodes on key feature dimensions, and divide each node into multiple node traffic association groups based on the correlation coefficient. Specifically, the Pearson correlation coefficient between nodes is obtained by calculating the ratio of the product of covariance and standard deviation of any two nodes on the key feature dimension; a correlation coefficient threshold is set, and when the absolute value of the correlation coefficient between two nodes is greater than or equal to the threshold, they are determined to be strongly correlated nodes and are assigned to the same correlation group. S330, calculate the overall traffic characteristics of each node traffic association group in the current time window, input the overall traffic characteristics into the traffic prediction sub-model, and obtain the short-term traffic change trend of the corresponding node traffic association group; the overall traffic characteristics include average traffic value, traffic fluctuation variance and peak frequency. Traffic trend prediction is performed for each node traffic association group. The traffic prediction sub-model within the group is trained by using historical traffic feature parameters. The key features of the current time window are input into the traffic prediction sub-model to obtain the short-term traffic change trend of each node traffic association group. S340, based on the short-term traffic change trend of each node traffic association group and the preset traffic load threshold, determine the current network traffic load status, which includes light load status, normal status and heavy load status. Specifically, when the predicted value of the short-term traffic change trend of the node traffic association group is less than the first threshold, it is determined to be a light load state; when the predicted value is greater than or equal to the first threshold and less than the second threshold, it is determined to be a normal state; when the predicted value is greater than or equal to the second threshold, it is determined to be a heavy load state. S350, Obtain the priority attribute of the data content; for node traffic association groups in a heavily loaded state, increase the priority of their data content by one level; for node traffic association groups in a lightly loaded state, decrease the priority of their data content by one level; the priority attribute is determined according to the type of data content and real-time requirements. Specifically, industrial control command data has a higher priority than historical data, and historical data has a higher priority than log data; S360 sorts all data content based on the adjusted priority attributes, constructs a priority queue for processing the data content, and determines the allocation ratio of computing resources according to the traffic load status of each node's traffic association group.

[0030] Specifically, the proportion of computing resources allocated to the associated groups under heavy load is higher than that under normal load, the proportion of computing resources allocated to the associated groups under normal load is higher than that under light load, and the sum of the computing resource allocation proportions of all associated groups is 100%. In some embodiments, S330, calculating the overall traffic characteristics of each node traffic association group in the current time window includes: Iterate through the standardized flow values ​​of all nodes in the node flow association group within the current time window, calculate the arithmetic mean of the standardized flow values ​​of each node in the association group, and obtain the overall average flow value. The variance of traffic fluctuation for each node in the associated group within the current time window is calculated. The average of the sums of the variances of all nodes is then taken to obtain the overall variance of traffic fluctuation. Identify the number of times the standardized traffic value of each node in the associated group exceeds the preset peak threshold within the current time window, sum up the peak occurrence times of each node and divide by the number of nodes to obtain the overall peak occurrence frequency.

[0031] Specifically, the standardized flow values ​​of all nodes within the node flow association group are iterated through in the current time window. The standardized flow values ​​of each node are summed and divided by the total number of nodes to obtain the overall average flow value of the association group. This value reflects the overall flow level of the association group in the current time window. The flow fluctuation variance of each node in the association group in the current time window is calculated. The variance values ​​of all nodes are arithmetically averaged to obtain the overall flow fluctuation variance, which reflects the overall dispersion of the flow data in the association group. The standardized flow values ​​of each node in the association group in the current time window are checked one by one. The number of times the standardized flow value of each node exceeds the preset peak threshold is counted. The peak occurrence frequency of all nodes is summed and divided by the number of nodes to obtain the overall peak occurrence frequency. This frequency is used to measure how frequently the flow peak of the association group occurs in the current time window.

[0032] In some embodiments, S400, the step of processing the data content according to the priority queue and the computing resource allocation ratio, and feeding back the processing result to the corresponding destination device node, includes: S410 allocates corresponding computing resources to each data item according to the priority queue order, and processes high-priority data first. S420 executes a preset data processing algorithm on the data content allocated to computing resources, including data cleaning, format conversion and feature extraction; S430 matches the processed data results with the corresponding destination device node information and generates a result feedback command; The S440 sends result feedback instructions to the target device node through the Industrial Internet of Things (IIoT) communication protocol, and records the data processing completion time and result status.

[0033] Specifically, when allocating computing resources, the resource ratios of CPU, memory, and network bandwidth are dynamically adjusted according to the computing resource allocation ratio of each node's traffic association group to ensure that high-priority data content receives sufficient processing resources. During data cleaning, outliers and duplicate data are filtered, and the mean imputation method is used to handle missing data to ensure data integrity and accuracy. The format conversion module converts heterogeneous data from different device nodes into JSON format for easy subsequent data interaction and parsing. The feature extraction stage extracts key features such as operation codes and parameter values ​​from industrial control instruction data, time series features from historical data, and event type features from log data based on the type of data content. The result feedback instruction includes processing result data, data source identifier, and timestamp information, and is transmitted to the destination device node via the MQTT protocol. A CRC check mechanism is used during transmission to ensure the reliability of data transmission. The data processing status recording module stores the status information of successful, failed, or timed-out processing in the local database and uploads it to the industrial IoT management platform so that managers can monitor the data processing process and device node operating status in real time.

[0034] In some embodiments, the method further includes: The load of each data processing node is monitored in real time. When the load of a node exceeds a preset threshold, some of the data processing tasks of that node are migrated to other nodes with loads below the preset threshold.

[0035] Specifically, a load monitoring module is deployed to periodically collect data on CPU utilization, memory usage, and task queue length of each data processing node, with a collection period of 5 seconds. Node load thresholds are set: CPU utilization at 80%, memory usage at 75%, and task queue length at 100 tasks. When any indicator exceeds its corresponding threshold, the node is identified as a high-load node. A task migration trigger mechanism is initiated, filtering out low-priority tasks that have not yet started processing from the task queues of high-load nodes and sorting them in ascending order of task size. Simultaneously, other data processing nodes in the network are scanned to identify all low-load nodes with CPU utilization below 40%, memory usage below 50%, and task queue length less than 20 tasks. The remaining resource capacity of each low-load node is calculated: remaining CPU processing capacity = (100% - current CPU utilization) × number of CPU cores, and remaining memory capacity = (100% - current memory usage) × total memory. The low-load nodes are then sorted in descending order of remaining resource capacity to form a target migration node list. A task migration channel is established, and low-priority tasks and their corresponding data context information are transmitted to the target migration node via a distributed task scheduling protocol. A breakpoint resumption mechanism is employed during transmission to ensure the integrity of the task data. After completing the task migration, high-load nodes update their local task queue status and send task migration logs to the industrial IoT management platform. The logs include the migration task ID, source node identifier, target node identifier, migration start time, and end time. Upon receiving the task data, the target migration node verifies the data context. If the verification is successful, the task is added to its local processing queue and scheduled for processing according to its original priority. After the task is completed, the processing result is synchronously fed back to the source node and the corresponding destination device node. This dynamic task migration mechanism achieves load balancing among data processing nodes, preventing data processing delays or task failures due to excessive load on a single node, and improving the stability and fault tolerance of the entire industrial IoT data processing system.

[0036] refer to Figure 2 This invention also provides a flow-aware industrial IoT data processing system, comprising: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor performs the method described above.

[0037] The content of the above method embodiments is applicable to this embodiment. The specific functions implemented in this embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments. Therefore, they will not be repeated here.

[0038] This invention also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described above. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0039] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0040] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0041] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0042] This invention also provides a computer program product, including a computer program or computer instructions, which are stored in a memory. A processor of a computer device reads the computer program or computer instructions from the memory and executes the computer program or computer instructions, causing the computer device to perform the above-described method.

[0043] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0044] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0045] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0046] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A data processing method for industrial IoT based on flow sensing, characterized in that, The method includes the following steps: S100: Obtain data flow information and corresponding data content of each device node in the Industrial Internet of Things within the most recent time period; S200, extract the feature parameters of the data traffic information, and construct a traffic feature matrix based on the feature parameters; the feature parameters include the mean feature, variance feature, peak feature and valley feature of the standardized traffic values ​​within each time window; the row dimension of the traffic feature matrix corresponds to the time window number, the column dimension corresponds to the node identifier, and the depth dimension corresponds to the feature parameter type; S300, dimensionality reduction is performed on the traffic feature matrix, principal component features are extracted from the traffic feature matrix, eigenvalues ​​and eigenvectors of each principal component are calculated, and principal components with eigenvalues ​​greater than a preset threshold are selected as key feature dimensions; Pearson correlation coefficients of different nodes on the key feature dimensions are calculated, and each node is divided into multiple node traffic association groups based on the Pearson correlation coefficients; the overall traffic characteristics of each node traffic association group in the current time window are calculated, and the overall traffic characteristics are input into the traffic prediction sub-model to obtain the short-term traffic change trend of the corresponding node traffic association group; the overall traffic characteristics include average traffic value, traffic fluctuation variance, and peak frequency. Based on the short-term traffic change trends of each node traffic association group and the preset traffic load threshold, the current network traffic load status is determined, including light load, normal, and heavy load status. The priority attributes of the data content are obtained; for node traffic association groups in a heavy load state, the priority of their data content is increased by one level; for node traffic association groups in a light load state, the priority of their data content is decreased by one level. The priority attributes are determined according to the type of data content and real-time requirements. Based on the adjusted priority attributes, all data content is sorted, a priority queue for processing data content is constructed, and the allocation ratio of computing resources is determined according to the traffic load status of each node traffic association group. S400: The data content is processed according to the priority queue and the computing resource allocation ratio, and the processing result is fed back to the corresponding destination device node.

2. The method according to claim 1, characterized in that, In S100, obtaining the data flow information and corresponding data content of each device node in the industrial Internet of Things includes: S110, acquire the raw traffic data frames sent by each device node; S120, parse the original traffic data frame, extract the data transmission timestamp, data transmission rate, data packet size, traffic source device identifier and destination device identifier to form data traffic information; S130, identify the type of data content in the original traffic data frame, associate and store the data traffic information with the corresponding data content, and establish a mapping table containing traffic information and data content.

3. The method according to claim 1, characterized in that, In S200, the step of extracting feature parameters of the data traffic information and constructing a traffic feature matrix based on the feature parameters includes: S210, standardizes real-time traffic data to obtain standardized traffic values; S220, Set the time window length and sliding step size, extract standardized flow values ​​sequentially according to the sliding step size, and generate multiple flow datasets corresponding to continuous and non-overlapping time windows. Each flow dataset contains the standardized flow value of each node in the corresponding time window and the collection timestamp. S230, traverse the traffic dataset within each time window, perform feature extraction operation for each dataset, and obtain feature parameters; the feature parameters include the mean feature, variance feature, peak feature, and valley feature of the standardized traffic values ​​within each time window; S240: Arrange the feature parameters of the same node in different time windows in sequence to form the feature parameter sequence of the node; then, horizontally concatenate the feature parameter sequences of all nodes according to the node identifier order to construct the traffic feature matrix.

4. The method according to claim 3, characterized in that, In S300, the calculation of Pearson correlation coefficients for different nodes on the key feature dimension, and the division of each node into multiple node traffic association groups based on the Pearson correlation coefficients, includes: The Pearson correlation coefficient between nodes is obtained by calculating the ratio of the product of the covariance and the standard deviation of any two nodes on the key feature dimension. Set a correlation coefficient threshold. When the absolute value of the Pearson correlation coefficient between two nodes is greater than or equal to the threshold, they are determined to be strongly correlated nodes and assigned to the same node traffic correlation group.

5. The method according to claim 4, characterized in that, In S300, calculating the overall traffic characteristics of each node's traffic association group within the current time window includes: Iterate through the standardized flow values ​​of all nodes in the node flow association group within the current time window, calculate the arithmetic mean of the standardized flow values ​​of each node in the association group, and obtain the overall average flow value. The variance of traffic fluctuation for each node in the associated group within the current time window is calculated. The average of the sums of the variances of all nodes is then taken to obtain the overall variance of traffic fluctuation. Identify the number of times the standardized traffic value of each node in the associated group exceeds the preset peak threshold within the current time window, sum up the peak occurrence times of each node and divide by the number of nodes to obtain the overall peak occurrence frequency.

6. The method according to claim 1, characterized in that, In S400, the step of processing the data content according to the priority queue and the computing resource allocation ratio, and feeding back the processing result to the corresponding destination device node, includes: S410 allocates corresponding computing resources to each data item according to the priority queue order, and processes high-priority data first. S420 executes a preset data processing algorithm on the data content allocated to computing resources, including data cleaning, format conversion and feature extraction; S430 matches the processed data results with the corresponding destination device node information and generates a result feedback command; The S440 sends result feedback instructions to the target device node through the Industrial Internet of Things (IIoT) communication protocol, and records the data processing completion time and result status.

7. The method according to claim 1, characterized in that, The method further includes: The load of each data processing node is monitored in real time. When the load of a node exceeds a preset threshold, some of the data processing tasks of that node are migrated to other nodes with loads below the preset threshold.

8. An industrial IoT data processing system based on flow sensing, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor performs the method as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images