Parameter adjustment method and device for adaptive network environment and electronic equipment

Abstract

The application relates to a parameter adjustment method and device for an adaptive network environment and electronic equipment, the method comprising: acquiring real-time data of the network environment, and performing standardization processing and instant analysis on the data to capture key features of the network state; based on the preprocessed network environment data, identifying the network state and outputting a network state trend prediction result in a short future time; and according to the prediction result, using an SAC model to adaptively adjust network parameters. The parameter adjustment method and device for the adaptive network environment and the electronic equipment can collect real-time network environment data, use advanced data analysis technology to predict the network state, obtain a network state trend prediction result in a short future time, and then conveniently adjust the network environment parameters in real time and accurately according to the dynamic change of the network state, so as to optimize the network performance and improve the stability and efficiency of the network.

Description

Technical Field

[0001] The present invention relates to the field of network environment parameter adjustment, and in particular to a parameter adjustment method, device and electronic equipment for adaptive network environment. Background Art

[0002] Traditional network environment parameter adjustment technologies often rely on manual settings, which are not only time-consuming but also unable to respond to dynamic changes in network status in real time, resulting in unstable network performance. For example, in changing environmental conditions such as temperature fluctuations and electromagnetic interference, network equipment often cannot automatically adapt to these changes, thus affecting network speed and stability. Traditional network environment parameter adjustment technologies have limitations in data collection, processing, and analysis, making it difficult to achieve real-time and accurate environmental parameter adjustment. Therefore, improvements are needed. Summary of the Invention

[0003] Based on this, it is necessary to provide a parameter adjustment method, device and electronic device for adaptive network environment to address the problem that traditional network environment parameter adjustment technology has limitations in data collection, processing and analysis, and it is difficult to achieve real-time and accurate environment parameter adjustment.

[0004] The present invention provides a parameter adjustment method for an adaptive network environment, the method comprising:

[0005] Acquire real-time data of the network environment, perform standardization and instant analysis on the data to capture key characteristics of the network status;

[0006] Based on the pre-processed network environment data, the network status is identified and the network status trend forecast results for the short term are output;

[0007] According to the prediction results, the SAC model is used to adaptively adjust the network parameters.

[0008] In one embodiment, the acquisition of real-time data of the network environment, and the standardization and immediate analysis of the data to capture key characteristics of the network status include:

[0009] Obtain data collected by temperature sensors and electromagnetic interference detectors on network nodes;

[0010] Clean and format the acquired data;

[0011] Use high-precision timestamps to record data collection time, and adjust the data collection frequency to multiple times per second;

[0012] Apache Flink is used to process streaming data in real time, and sliding windows and aggregation functions are set to perform instant data analysis.

[0013] In one embodiment, the method of using a high-precision timestamp to record the data collection time and adjusting the data collection frequency to multiple times per second includes:

[0014] Add a high-precision timestamp to each data collected by the temperature sensor and electromagnetic interference detector;

[0015] Adjust the data collection frequency to multiple times per second;

[0016] Sort the data by timestamp;

[0017] Align data sources with different frequencies;

[0018] In response to missing data or data of different frequencies that are difficult to align completely, missing values ​​are supplemented based on interpolation or sampling techniques;

[0019] Correct timestamp precision based on missing values.

[0020] In one embodiment, the real-time processing of streaming data based on Apache Flink and setting of sliding windows and aggregation functions for instant data analysis include:

[0021] Build a stream data processing pipeline based on Apache Flink and configure data sources to receive real-time data from temperature sensors and electromagnetic interference detectors.

[0022] Divide the data stream into specific sliding windows that cover the time range required for analysis;

[0023] In each sliding window, the data in the window is summarized based on the aggregation function to capture the key characteristics of the network status;

[0024] In response to abnormal network status, an early warning signal is output and corresponding adjustment measures are triggered.

[0025] In one embodiment, the warning signal is used to push to users in the form of text messages and emails through the cloud platform, and is connected to the big data platform in real time through the data interface for management personnel to conduct on-site inspections and disposal.

[0026] In one embodiment, the identifying of the network status based on the pre-processed network environment data and outputting a prediction result of the network status trend in the short term in the future includes:

[0027] Obtain pre-processed network environment data;

[0028] The network environment data is input into the CNN model to output the network status trend prediction result in the short term in the future. The CNN model is trained using normal network environment data samples and abnormal network environment data samples as training data.

[0029] In one embodiment, the adaptively adjusting network parameters using the SAC model according to the prediction results includes:

[0030] Receive the prediction results of the CNN model;

[0031] Construct an input vector based on the prediction results and the current network parameter configuration;

[0032] Input the input vector into the SAC model and output the action strategy, wherein the SAC model is based on a reinforcement learning algorithm and is trained using network data anomaly samples and corresponding action strategy samples as training data;

[0033] Based on the action strategy, adjust the network parameters.

[0034] The present invention also provides a parameter adjustment device for an adaptive network environment, comprising:

[0035] The acquisition module is used to obtain real-time data of the network environment, and perform standardization and real-time analysis on the data to capture the key characteristics of the network status;

[0036] The identification module is used to identify the network status based on the pre-processed network environment data and output the network status trend prediction results in the short term;

[0037] The adjustment module is used to adaptively adjust network parameters using the SAC model according to the prediction results.

[0038] The present invention also provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements any of the above-described parameter adjustment methods for an adaptive network environment when executing the computer program.

[0039] The present invention also provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the method for adjusting parameters of an adaptive network environment as described above is implemented.

[0040] The above-mentioned parameter adjustment method, device and electronic device for adaptive network environment, by collecting network environment data in real time and using advanced data analysis technology to predict network status, obtain the network status trend prediction results in the short term in the future, and then facilitate the real-time and accurate adjustment of network environment parameters according to the dynamic changes of network status, so as to optimize network performance and improve network stability and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 A schematic flow chart of a parameter adjustment method for an adaptive network environment in one embodiment;

[0043] Figure 2 Schematic diagram of data preprocessing process in one embodiment;

[0044] Figure 3 A schematic diagram of a process for recording data collection time using a high-precision timestamp in one embodiment;

[0045] Figure 4 Schematic diagram of a process for real-time processing of streaming data based on Apache Flink in one embodiment;

[0046] Figure 5 A schematic diagram of a network status identification process in one embodiment;

[0047] Figure 6 FIG. 1 is a flow chart of adaptively adjusting network parameters using the SAC model in one embodiment;

[0048] Figure 7 A schematic diagram of the structure of a parameter adjustment device for adaptive network environment in one embodiment;

[0049] Figure 8 FIG. 1 is a diagram showing the internal structure of an electronic device according to an embodiment. DETAILED DESCRIPTION

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying 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 making creative efforts shall fall within the scope of protection of the present invention.

[0051] The following combination Figures 1-8 The present invention describes a parameter adjustment method, device and electronic device for an adaptive network environment.

[0052] like Figure 1As shown, in one embodiment, a method for adaptive parameter adjustment of network environment includes the following steps:

[0053] Step S100, real-time data of network environment is obtained, and the data is standardized and analyzed in real time to capture key features of network state.

[0054] Temperature sensors and electromagnetic interference detectors are deployed at key locations in the network environment, which can continuously monitor temperature changes and electromagnetic interference levels in the environment. Through the establishment of wireless data transmission link, the sensors transmit the real-time collected data to the data processing center, so as to obtain real-time information flow about the network environment state, laying a solid foundation for further data analysis and parameter adjustment.

[0055] Step S200, based on the preprocessed network environment data, the network state is identified, and the network state trend prediction result in the near future is output.

[0056] It is convenient to predict the network state trend in the near future according to the real-time monitored network state, and then to adjust the network environment parameters in advance.

[0057] Step S300, according to the prediction result, the SAC model is used to adaptively adjust the network parameters.

[0058] SAC (Soft Actor-Critic) model is a reinforcement learning algorithm, which is mainly used to solve the control problem in continuous action space. SAC algorithm combines actor-critic method and maximum entropy reinforcement learning theory, aiming to optimize the policy while maximizing the entropy of the policy to improve the exploration ability and learning efficiency.

[0059] By using SAC model to adaptively adjust network parameters such as baud rate, frame length, TCP window size, etc. according to the prediction result, it is convenient to optimize network performance.

[0060] The adaptive parameter adjustment method of network environment, through real-time collection of network environment data and use of advanced data analysis technology to predict network state, obtains network state trend prediction result in the near future, and then adjusts network environment parameters in real time and accurately according to the dynamic change of network state, to optimize network performance, improve the stability and efficiency of network.

[0061] In this embodiment, referring to Figure 2 , real-time data of network environment is obtained, and the data is standardized and analyzed in real time to capture key features of network state, including the following steps:

[0062] Step S110, acquire the data collected by the temperature sensor and electromagnetic interference detector on the network node.

[0063] By installing temperature sensors and electromagnetic interference detectors at locations where network equipment may be affected by environmental factors, and configuring the wireless communication modules of the temperature sensors and electromagnetic interference detectors, real-time data transmission is ensured. Then the temperature sensors and electromagnetic interference detectors start working, real-time monitoring of environmental data, and sending to the data processing center through the preset data transmission protocol.

[0064] Among them, the deployment of temperature sensors and electromagnetic interference detectors is the starting point of data collection, real-time data transmission ensures the real-time nature of data, and data analysis and parameter adjustment is the next step to optimize network performance using these data.

[0065] Step S120, clean and format the acquired data.

[0066] The data processing center receives the data, performs preliminary cleaning and preprocessing, and prepares for subsequent analysis and adjustment.

[0067] Step S130, use high-precision timestamps to record data collection time, and adjust the data collection frequency to multiple times per second.

[0068] Each temperature sensor and electromagnetic interference detector will record a high-precision timestamp when collecting data, which provides an accurate time reference for subsequent data analysis. In addition, in order to improve the resolution of the data, the sampling frequency of the temperature sensor and electromagnetic interference detector is adjusted to ensure that it can collect data at a rate of multiple times per second. Such a collection frequency can capture rapid changes in the environment, providing a more detailed data basis for subsequent network state analysis and parameter adjustment.

[0069] Step S140, real-time processing of streaming data according to Apache Flink, and setting sliding windows and aggregation functions for immediate analysis of data.

[0070] Apache Flink is an open source stream processing framework, designed for processing real-time data streams and batch processing data. It can efficiently process massive amounts of data and provide low-latency and high-throughput processing capabilities. The working principle is as follows:

[0071] Stream processing: Data streams (unbounded streams) are continuously injected into Apache Flink, which processes data in real time according to the defined business logic and window configuration.

[0072] Batch processing: Batch processing (bounded data sets) tasks are loaded with data once, and the task execution stops after completion.

[0073] State management and fault tolerance: Apache Flink periodically creates checkpoints to save the state of each task. When a failure occurs, Apache Flink can resume execution from the most recent checkpoint.

[0074] Parallel execution: Apache Flink executes multiple sub-tasks in parallel through task slots on TaskManagers, improving processing efficiency. Each TaskManager can execute multiple task instances, enabling efficient utilization of the cluster.

[0075] Among them, TaskManager is the node that actually executes tasks in Apache Flink cluster, and each TaskManager is responsible for executing one or more tasks.

[0076] By using Apache Flink to process real-time data, it is convenient to realize real-time monitoring of network status.

[0077] In this embodiment, referring to Figure 3 , high-precision timestamps are used to record data collection times, and data collection frequency is adjusted to multiple times per second, including the following steps:

[0078] Step S131, a high-precision timestamp is attached to the data collected each time by the temperature sensor and the electromagnetic interference detector.

[0079] The timestamp can be accurate to the millisecond level, making it easy to accurately track the trend of data changes.

[0080] Step S132, adjust the data collection frequency to multiple times per second.

[0081] By increasing the data collection frequency, it is convenient to capture subtle changes in the network environment.

[0082] Step S133, sort the data according to the timestamp.

[0083] These data include timestamps and multiple environmental parameters. By sorting the data using timestamps, it is convenient to accurately analyze the environmental status.

[0084] Step S134, align different frequency data sources.

[0085] It is convenient to make all data sources at each time point synchronized on the same timeline.

[0086] Step S135, in response to data missing or different frequency data being difficult to completely align, missing values are supplemented according to interpolation or sampling techniques.

[0087] When data is missing or data of different frequencies cannot be completely aligned, interpolation or sampling techniques are used to supplement the missing values ​​to maintain the integrity of the time series.

[0088] Step S136: Correct the timestamp accuracy based on the missing value.

[0089] Missing values ​​can represent interruptions or errors in data collection. Through missing value analysis and processing, timestamps can be effectively corrected. By correcting the timestamp accuracy, clock drift or synchronization errors can be reduced, ensuring the consistency of data timing.

[0090] In this embodiment, see Figure 4 , based on Apache Flink, real-time processing of streaming data, and setting sliding windows and aggregation functions to perform instant data analysis, including the following steps:

[0091] In step S141 , a stream data processing pipeline is constructed based on Apache Flink, and a data source is configured to receive data transmitted in real time by the temperature sensor and the electromagnetic interference detector.

[0092] By building a stream data processing pipeline through Apache Flink and configuring the data sources of temperature sensors and electromagnetic interference detectors, it is convenient to achieve real-time data processing and low-latency response; data cleaning, aggregation and analysis; anomaly detection and automatic alarm; multi-source data fusion and in-depth analysis; data storage and subsequent analysis.

[0093] Step S142: Divide the data stream into specific sliding windows, where the sliding windows cover the time range required for analysis.

[0094] In a stream data processing pipeline, the data stream is divided into specific sliding windows. The size of these windows can be adjusted according to actual needs to analyze the network status at different time granularities. By defining the size and sliding step of the sliding window, it can be ensured that the window can cover the time range required for analysis.

[0095] Step S143: In each sliding window, the data in the window is aggregated based on an aggregation function to capture key features of the network status.

[0096] These aggregation functions can calculate statistical indicators such as average, maximum, and minimum values, thereby capturing the key characteristics of the network status. In this way, they can quickly respond to changes in the network environment and take appropriate measures to maintain optimal network performance.

[0097] Step S144: In response to the abnormal network status, an early warning signal is output and corresponding adjustment measures are triggered.

[0098] It should be noted that the early warning signal is pushed to users in the form of text messages and emails through the cloud platform, and is connected to the big data platform in real time through the data interface for management personnel to conduct on-site inspections and disposal.

[0099] Specific application examples are as follows:

[0100] Assume that a 1-second sliding window is set, and the aggregation function is to calculate the average signal strength within the window. When new data flows in, Apache Flink will process it as follows:

[0101] 1. Data enters a sliding window, which contains all signal strength data collected in the last second.

[0102] 2. Apply the aggregate function to calculate the average of these data points.

[0103] 3. Output the average signal strength value and use it for network status monitoring.

[0104] 4. If the average signal strength falls below the preset threshold, the system will automatically trigger adjustments to enhance the signal.

[0105] In this embodiment, see Figure 5 , based on the pre-processed network environment data, identify the network status and output the network status trend prediction results in the short term, including the following steps:

[0106] Step S210: obtaining pre-processed network environment data.

[0107] Before inputting the data into the CNN (Convolutional Neural Network) model, the data is first standardized to ensure that each feature value is on the same scale, which helps the CNN model learn and converge. Data preprocessing also includes removing missing values ​​and outliers, and converting time series data into a format suitable for CNN processing.

[0108] Step S220, the network environment data is input into the CNN model, and the network status trend prediction result in the short term in the future is output. The CNN model is trained using normal network environment data samples and abnormal network environment data samples as training data.

[0109] In the present invention, a multi-layer convolutional neural network model is designed to effectively extract features from network status data. The model includes an input layer, multiple convolutional layers, multiple pooling layers, a fully connected layer, and an output layer. The convolutional layer is used to extract local features in the data; the pooling layer is used to reduce the dimension of the data while retaining important information; the fully connected layer is used to combine the features extracted by the convolutional layer and the pooling layer to facilitate the final prediction.

[0110] Feature extraction: Use convolutional layers and pooling layers to automatically learn complex feature representations from raw data.

[0111] Time Series Analysis: CNN models are able to process time series data because they can capture the local dependencies of data in the time dimension.

[0112] The following formula is used for CNN model training:

[0113]

[0114] Where L(θ) is the loss function, N is the number of training samples, and y i is the true value, f(x i ,θ) is the model prediction value, and θ is the model parameter.

[0115] The CNN model is trained by minimizing the loss function and the gradient descent algorithm is used to update the CNN model parameters:

[0116]

[0117] Here, α is the learning rate.

[0118] Prediction: Once the CNN model is trained, it can be used to predict network status trends in the short term.

[0119] The following are examples of specific implementations:

[0120] Assume there is a set of continuous network status data, including indicators such as signal strength, bandwidth utilization, and latency.

[0121] Input this data into the trained CNN model.

[0122] The CNN model will output a prediction of the network status for a period of time in the future, for example, predicting whether bandwidth utilization will increase in the next minute.

[0123] In this way, we can monitor the network status in real time and take optimization measures in advance based on the prediction results. The following is an example:

[0124] If the CNN model predicts upcoming high bandwidth usage, network resources can be increased in advance to avoid possible performance degradation.

[0125] If it is predicted that the signal strength will drop, the location or parameters of the wireless access point can be adjusted in time to improve signal coverage.

[0126] The application of the above steps and algorithms can effectively predict network status trends and make adaptive adjustments accordingly to maintain stable network performance.

[0127] In this embodiment, see Figure 6 ,According to the prediction results, the SAC model is used to adaptively adjust the network parameters,including the following steps:

[0128] Step S310: Receive the prediction result of the CNN model.

[0129] Analyze the prediction results output by the CNN model, which include estimates of network status in the future, such as traffic trends and signal strength fluctuations.

[0130] Step S320: construct an input vector based on the prediction result and the current network parameter configuration.

[0131] The current network parameter configuration includes baud rate, frame length, and TCP window size. The input vector represents the current state of the network and the expected future state.

[0132] Step S330: Input the input vector into the SAC model and output the action strategy. The SAC model is based on a reinforcement learning algorithm and is trained using abnormal network data samples and corresponding action strategy samples as training data.

[0133] The SAC model is a reinforcement learning model based on deep learning, which can learn and output the optimal action strategy based on the current input vector.

[0134] Step S340: Adjust network parameters based on the action strategy.

[0135] Based on the decision of the SAC model, the network parameters are adaptively adjusted as follows:

[0136] Baud rate adjustment: If the forecast results show that network traffic will increase, the SAC model may recommend increasing the baud rate to handle the larger amount of data.

[0137] Frame length adjustment: Based on changes in network traffic, the SAC model can adjust the frame length to optimize data transmission efficiency and network throughput.

[0138] TCP window size adjustment: To control network congestion, the SAC model dynamically adjusts the TCP window size according to the network status.

[0139] Specific implementation steps:

[0140] 1. Receive the prediction results of the CNN model and convert them into an input format recognizable by the SAC model.

[0141] 2. Send the input vector to the SAC model, and the model will output the corresponding action strategy.

[0142] 3. Based on the decision of the SAC model, perform one or more of the following actions:

[0143] If a higher baud rate is recommended, increase the baud rate of the current network device through the network management interface.

[0144] If it is recommended to adjust the frame length, the configuration of the network protocol stack is modified to change the frame length of the data packet.

[0145] If TCP window size adjustment is recommended, the window size is adjusted by the network congestion control algorithm.

[0146] 4. Monitor the adjusted network performance indicators to verify the effectiveness of the SAC model decisions and make fine-tuning based on actual conditions.

[0147] Here are some examples:

[0148] Suppose a forecast shows that network traffic will increase by 30% over the next 5 minutes.

[0149] After receiving this prediction, the SAC model decides to increase the baud rate by 10% and the TCP window size by 20%.

[0150] The network administrator performs parameter adjustments based on the SAC model's recommendations.

[0151] After the adjustment, network performance indicators showed that network throughput increased by 25% while latency decreased by 15%, verifying the correctness of the SAC model decision.

[0152] In this way, network parameters can be automatically adjusted in real time according to changes in network status to maintain optimal network performance.

[0153] The parameter adjustment device for an adaptive network environment provided by the present invention is described below. The parameter adjustment device for an adaptive network environment described below and the parameter adjustment method for an adaptive network environment described above can refer to each other.

[0154] like Figure 7 As shown, in one embodiment, a parameter adjustment device for adaptive network environment includes an acquisition module 710 , an identification module 720 and an adjustment module 730 .

[0155] The acquisition module 710 is used to acquire real-time data of the network environment, and perform standardized processing and real-time analysis on the data to capture key features of the network status.

[0156] The identification module 720 is used to identify the network status based on the pre-processed network environment data and output the network status trend prediction result in the short term.

[0157] The adjustment module 730 is used to adaptively adjust network parameters using the SAC model according to the prediction results.

[0158] In this embodiment, the acquisition module 710 is specifically configured to:

[0159] Obtain data collected by temperature sensors and electromagnetic interference detectors on network nodes;

[0160] Clean and format the acquired data;

[0161] Use high-precision timestamps to record data collection time, and adjust the data collection frequency to multiple times per second;

[0162] Apache Flink is used to process streaming data in real time, and sliding windows and aggregation functions are set to perform instant data analysis.

[0163] In this embodiment, a high-precision timestamp is used to record the data collection time, and the data collection frequency is adjusted to multiple times per second, specifically for:

[0164] Add a high-precision timestamp to each data collected by the temperature sensor and electromagnetic interference detector;

[0165] Adjust the data collection frequency to multiple times per second;

[0166] Sort the data by timestamp;

[0167] Align data sources with different frequencies;

[0168] In response to missing data or data of different frequencies that are difficult to align completely, missing values ​​are supplemented based on interpolation or sampling techniques;

[0169] Correct timestamp precision based on missing values.

[0170] In this embodiment, Apache Flink is used to process streaming data in real time, and sliding windows and aggregation functions are set to perform real-time data analysis. Specifically, the following are used:

[0171] Build a stream data processing pipeline based on Apache Flink and configure data sources to receive real-time data from temperature sensors and electromagnetic interference detectors.

[0172] Divide the data stream into specific sliding windows that cover the time range required for analysis;

[0173] In each sliding window, the data in the window is summarized based on the aggregation function to capture the key characteristics of the network status;

[0174] In response to abnormal network status, an early warning signal is output and corresponding adjustment measures are triggered.

[0175] In this embodiment, the identification module 720 is specifically configured to:

[0176] Obtain pre-processed network environment data;

[0177] The network environment data is input into the CNN model, and the network status trend prediction results in the short term in the future are output. The CNN model is trained using normal network environment data samples and abnormal network environment data samples as training data.

[0178] In this embodiment, the adjustment module 730 is specifically configured to:

[0179] Receive the prediction results of the CNN model;

[0180] Construct an input vector based on the prediction results and the current network parameter configuration;

[0181] The input vector is input into the SAC model, and the action strategy is output. The SAC model is based on the reinforcement learning algorithm and is trained with network data anomaly samples and corresponding action strategy samples as training data;

[0182] Based on the action strategy, adjust the network parameters.

[0183] This parameter adjustment device for adaptive network environment collects network environment data in real time and uses advanced data analysis technology to predict network status, obtaining network status trend prediction results in the short term. It then facilitates real-time and accurate adjustment of network environment parameters according to dynamic changes in network status to optimize network performance and improve network stability and efficiency.

[0184] Figure 8 The following is a schematic diagram of the physical structure of an electronic device. The electronic device may be a smart terminal, and its internal structure diagram may be as follows: Figure 8 As shown. The electronic device includes a processor, a memory, and a network interface connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a parameter adjustment method for an adaptive network environment is implemented, the method comprising:

[0185] Acquire real-time data of the network environment, perform standardization and instant analysis on the data to capture key characteristics of the network status;

[0186] Based on the pre-processed network environment data, the network status is identified and the network status trend forecast results for the short term are output;

[0187] According to the prediction results, the SAC model is used to adaptively adjust the network parameters.

[0188] Those skilled in the art will understand that Figure 8 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the electronic device to which the solution of the present invention is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0189] On the other hand, the present invention also provides a computer storage medium storing a computer program, which, when executed by a processor, implements a parameter adjustment method for an adaptive network environment, the method comprising:

[0190] Acquire real-time data of the network environment, perform standardization and instant analysis on the data to capture key characteristics of the network status;

[0191] Based on the pre-processed network environment data, the network status is identified and the network status trend forecast results for the short term are output;

[0192] According to the prediction results, the SAC model is used to adaptively adjust the network parameters.

[0193] In another aspect, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and when the processor executes the computer instructions, implements a method for adjusting parameters of an adaptive network environment, the method comprising:

[0194] Acquire real-time data of the network environment, perform standardization and instant analysis on the data to capture key characteristics of the network status;

[0195] Based on the pre-processed network environment data, the network status is identified and the network status trend forecast results for the short term are output;

[0196] According to the prediction results, the SAC model is used to adaptively adjust the network parameters.

[0197] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

[0198] By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0199] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0200] The above-described embodiments merely illustrate several embodiments of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make various modifications and improvements without departing from the spirit of the present invention, and these modifications and improvements fall within the scope of the present invention. Therefore, the scope of the present invention shall be determined by the appended claims.

Claims

1. A parameter adjustment method for an adaptive network environment, characterized in that: The method comprises: Real-time data on the network environment collected by temperature sensors and electromagnetic interference detectors on network nodes is obtained, and the data is standardized and analyzed in real time to capture key characteristics of the network status. The standardization and real-time analysis include: using high-precision timestamps to record data collection time, adjusting the data collection frequency to multiple times per second, time-aligning the different frequency data sources of the temperature sensors and electromagnetic interference detectors, and supplementing missing values ​​through interpolation or sampling techniques; building a stream data processing pipeline based on Apache Flink, configuring the data source to receive the real-time data stream transmitted by the temperature sensors and electromagnetic interference detectors, dividing the data stream into sliding windows covering the analysis time range, summarizing the data within each sliding window based on an aggregation function, and outputting warning signals in response to abnormal network status and triggering corresponding adjustment measures; Based on the preprocessed network environment data, the time series data is converted into a format suitable for CNN processing, the network status is identified, and a prediction result of the network status trend in the short term is output. The preprocessed network environment data is input into the CNN model, and the prediction result of the network status trend in the short term is output. The CNN model is trained with normal and abnormal network environment data samples and includes convolutional layers, pooling layers, and fully connected layers. Based on the prediction results, the SAC model is used to adaptively adjust network parameters. Specifically, an input vector is constructed based on the prediction results and the current network parameter configuration, and the input vector is input into the SAC model trained based on the reinforcement learning algorithm. The network parameter adjustment strategy is output to dynamically adjust the baud rate, frame length or TCP window size. The SAC model is trained using network data anomaly samples and corresponding action strategy samples as training data.

2. The parameter adjustment method for an adaptive network environment according to claim 1, characterized in that: The method uses a high-precision timestamp to record the data collection time and adjusts the data collection frequency to multiple times per second, including: Add a high-precision timestamp to each data collected by the temperature sensor and electromagnetic interference detector; Adjust the data collection frequency to multiple times per second; Sort the data based on timestamp.

3. The parameter adjustment method for an adaptive network environment according to claim 2, characterized in that: The time alignment of different frequency data sources of the temperature sensor and the electromagnetic interference detector and supplementation of missing values ​​by interpolation or sampling techniques include: Align data sources with different frequencies; In response to missing data or data of different frequencies that are difficult to align completely, missing values ​​are supplemented based on interpolation or sampling techniques; Correct timestamp precision based on missing values.

4. The parameter adjustment method for an adaptive network environment according to claim 3, characterized in that: The warning signal is used to be pushed to users in the form of text messages and emails through the cloud platform, and is connected to the big data platform in real time through the data interface for management personnel to conduct on-site inspections and disposal.

5. A parameter adjustment device for adaptive network environment, characterized in that: include: An acquisition module is used to acquire real-time network environment data collected by temperature sensors and electromagnetic interference detectors on network nodes, and perform standardization and real-time analysis on the data to capture key characteristics of the network status. The standardization and real-time analysis include: using high-precision timestamps to record data collection time, adjusting the data collection frequency to multiple times per second, time-aligning the different frequency data sources of the temperature sensors and electromagnetic interference detectors, and supplementing missing values ​​through interpolation or sampling techniques; building a stream data processing pipeline based on Apache Flink, configuring the data source to receive the real-time data stream transmitted by the temperature sensors and electromagnetic interference detectors, dividing the data stream into sliding windows covering the analysis time range, summarizing the data within each sliding window based on an aggregation function, and outputting warning signals in response to abnormal network status and triggering corresponding adjustment measures; An identification module is used to convert the time series data into a format suitable for CNN processing based on the preprocessed network environment data, identify the network status, and output a prediction result of the network status trend in the short term. The preprocessed network environment data is input into a CNN model, and the prediction result of the network status trend in the short term is output. The CNN model is trained with normal and abnormal network environment data samples and includes convolutional layers, pooling layers, and fully connected layers. An adjustment module is configured to adaptively adjust network parameters using a SAC model based on prediction results. An input vector is constructed based on the prediction results and the current network parameter configuration, and the input vector is input into a SAC model trained based on a reinforcement learning algorithm. The module then outputs a network parameter adjustment strategy to dynamically adjust the baud rate, frame length, or TCP window size. The SAC model is trained using network data anomaly samples and corresponding action strategy samples as training data.

6. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images