Dynamic routing optimization method, device, equipment and storage medium
By conducting in-depth path learning and node information collection on the device network topology structure, combining linear coding processing and multi-level data analysis, we identify and manage Tethering behavior, dynamic routing optimization is achieved, solving the problem that traditional routing strategies cannot adapt to network changes, and improving the problem that traditional routing strategies cannot adapt to network changes. The stability and security of the network are achieved.
Patent Information
- Application Number
- CN202410530726.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-29
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-04-29
AI Technical Summary
Traditional static routing strategies cannot adapt to real-time changes in network state and cannot effectively deal with network congestion, device failure and security threats. The existing routing optimization methods have shortcomings in data fusion, feature extraction and future state prediction, which limits network performance and stability. Improvement of sex.
By performing path learning and node information collection on the initial device network topology, combining linear coding processing and state estimation, multiple groups of candidate data transmission sequences are calculated, Tethering behavior is detected and managed, routing path interaction optimization is carried out, and multi-level data analysis and features are used to analyze and perform multi-level data. Extraction is dynamic routing optimization to realize global network status monitoring and performance analysis.
Improve the accuracy of dynamic routing optimization of equipment, reduce the high peak average power ratio of data transmission, ensure the stability and security of the network, ensure the fair distribution and efficiency of network resources, enhance the robustness of the network and reliability.
Smart Images

Figure CN118353827B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of dynamic routing optimization, and in particular to a dynamic routing optimization method, device, equipment and storage medium. Background Art
[0002] In today's era of increasing digitalization and networking, dynamic routing optimization, as an important part of network management and maintenance, is crucial to ensuring the stability and efficiency of network communications. With the rapid development of new technologies such as the Internet of Things, smart manufacturing, and smart cities, the number of devices has increased dramatically, and the complexity of interactions between devices has also increased. This requires the network to not only handle larger amounts of data, but also maintain efficient and reliable data transmission in a changing network environment. However, traditional static routing strategies have been unable to meet these requirements because they cannot adapt to real-time changes in network status and cannot effectively handle problems such as network congestion, device failures, and security threats.
[0003] In addition, as network security issues become increasingly prominent, how to ensure the security and stability of the network while ensuring data transmission efficiency has become an urgent problem to be solved. Tethering behavior that may exist in the network, that is, users share Internet connections through mobile devices, may lead to unfair use of the network and may even become a channel for network attacks. Therefore, how to effectively identify and manage such behaviors during the routing optimization process to ensure the fairness and security of the network is an important challenge in network management. The complexity and dynamics of the network environment require that the routing optimization strategy should not only consider the current state of the network, but also be predictive and be able to predict the future network state based on historical data and current trends, so as to make more scientific and reasonable routing decisions. However, existing routing optimization methods still have deficiencies in data fusion, feature extraction, and future state prediction, which limits the effectiveness of routing optimization strategies and cannot maximize the performance and stability of the network. Summary of the invention
[0004] The present application provides a dynamic routing optimization method, apparatus, device and storage medium, which are used to improve the accuracy of dynamic routing optimization of the device.
[0005] In a first aspect, the present application provides a dynamic routing optimization method, the dynamic routing optimization method comprising:
[0006] Perform path learning and neighbor node information collection on the preset initial device network topology structure to obtain a node state data set of each device node, and perform linear encoding processing and state estimation on the node state data set of each device node to obtain a node state evaluation index for each device node;
[0007] Calculate multiple groups of candidate data transmission sequences for each device node according to the node status evaluation index, and optimize the multiple groups of candidate data transmission sequences to obtain a target data transmission sequence for each device node;
[0008] Controlling each device node to transmit device signals through the target data transmission sequence, and performing Tethering behavior detection and identification on the initial device network topology structure to obtain a Tethering behavior identification result;
[0009] According to the Tethering behavior identification result, each device node is respectively optimized for routing path interaction topology, multiple sub-routing paths are generated for each device node, and a corresponding first device network topology structure is generated through the multiple sub-routing paths of each device node;
[0010] Performing multi-level data analysis and feature extraction on the first device network topology structure to obtain multi-level fusion features, and dynamically optimizing routing of the first device network topology structure by using the multi-level fusion features to obtain a second device network topology structure;
[0011] The network status of the second device network topology structure is monitored and the performance is analyzed to obtain a network status performance index, and the second device network topology structure is globally optimized according to the network status performance index to generate a target device network topology structure.
[0012] In a second aspect, the present application provides a dynamic routing optimization device, the dynamic routing optimization device comprising:
[0013] The acquisition module is used to perform path learning and neighbor node information acquisition on the preset initial device network topology structure, obtain the node state data set of each device node, and perform linear coding processing and state estimation on the node state data set of each device node to obtain the node state evaluation index of each device node;
[0014] A calculation module, used to calculate multiple groups of candidate data transmission sequences for each device node according to the node status evaluation index, and optimize the multiple groups of candidate data transmission sequences to obtain a target data transmission sequence for each device node;
[0015] an identification module, used to control each device node to perform device signal transmission through the target data transmission sequence, and perform Tethering behavior detection and identification on the initial device network topology structure to obtain a Tethering behavior identification result;
[0016] A generation module, configured to perform routing path interaction topology optimization on each device node according to the Tethering behavior identification result, generate multiple sub-routing paths for each device node, and generate a corresponding first device network topology structure through the multiple sub-routing paths of each device node;
[0017] an analysis module, configured to perform multi-level data analysis and feature extraction on the first device network topology structure to obtain multi-level fusion features, and dynamically optimize routing of the first device network topology structure by using the multi-level fusion features to obtain a second device network topology structure;
[0018] The optimization module is used to perform network status monitoring and performance analysis on the second device network topology structure to obtain network status performance indicators, and to globally optimize the second device network topology structure based on the network status performance indicators to generate a target device network topology structure.
[0019] The third aspect of the present application provides a dynamic routing optimization device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the dynamic routing optimization device executes the above-mentioned dynamic routing optimization method.
[0020] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned dynamic routing optimization method.
[0021] In the technical solution provided by the present application, by conducting in-depth path learning and node information collection on the initial device network topology, combined with linear coding processing and minimum variance unbiased estimator, the state of each device node can be accurately evaluated. This not only improves the accuracy of node state evaluation, but also provides a reliable data basis for subsequent route optimization. By calculating and optimizing multiple sets of candidate data transmission sequences for each device node, the peak-to-average power ratio of data transmission is effectively reduced, and the quality and efficiency of data transmission are significantly improved. This optimization minimizes the distortion of the signal during transmission, thereby ensuring the stability and reliability of data transmission in the network. Not only can the device nodes be controlled to perform effective signal transmission, but also the Tethering behavior in the network can be effectively identified and managed through advanced traffic data collection and analysis technology. Such a management mechanism significantly improves the security and stability of the network and ensures the fair allocation of network resources. According to the Tethering behavior identification results, the device nodes can be targeted for routing path optimization to generate a more reasonable device network topology. This dynamic and flexible optimization strategy fully considers the real-time status and changes in the network, ensuring the efficiency and adaptability of the network structure. By using multi-level data analysis and feature extraction technology, we can understand the network status more comprehensively and deeply, thus providing richer and more accurate decision support for route optimization. Through real-time monitoring of network status and comprehensive performance analysis, we can timely discover potential problems in the network and perform global optimization based on network status performance indicators. Such global optimization not only improves the performance of the network, but also enhances the robustness and reliability of the network, thereby improving the accuracy of dynamic route optimization of the device. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0023] Figure 1 A schematic diagram of an embodiment of a dynamic routing optimization method in an embodiment of the present application;
[0024] Figure 2 Schematic diagram of an embodiment of a dynamic routing optimization device in an embodiment of the present application. DETAILED DESCRIPTION
[0025] The embodiments of the present application provide a method, apparatus, device and storage medium for dynamic routing optimization. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0026] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 , an embodiment of the dynamic routing optimization method in the embodiment of the present application includes:
[0027] Step S101: perform path learning and neighbor node information collection on a preset initial device network topology structure to obtain a node state data set of each device node, and perform linear coding processing and state estimation on the node state data set of each device node to obtain a node state evaluation index of each device node;
[0028] It is understandable that the execution subject of the present application may be a dynamic routing optimization device, or a terminal or a server, which is not limited here. The present application embodiment is described by taking a server as the execution subject as an example.
[0029] Specifically, the connection status and physical location of each target device are obtained through network detection technology and spatial positioning system. The initial device network topology structure is constructed according to the principles of graph theory and network topology, which can clearly show the connection mode and mutual relationship between devices. The initial topology structure is path-learned to simulate the flow path of data in the network, so as to determine the key nodes and potential communication bottlenecks in the network. In this process, multiple device nodes in the network are identified, and the role and importance of these nodes in the network are analyzed. Neighbor node information is collected for device nodes, and key indicators such as signal strength, energy usage, and data transmission rate of each node are collected. The node status data set of each device node is processed by a preset linear encoder, and the status data is converted into a format that is more suitable for analysis and processing, which also helps to reduce the loss and error in the data transmission process. After encoding processing, the node linear encoding set of each device node is obtained. The linear encoding set is estimated by the preset minimum variance unbiased estimator. By analyzing the encoding data, the operating status and performance indicators of each node are accurately predicted, and the node status evaluation indicators of each device node are generated. These indicators determine the effect and accuracy of the subsequent dynamic routing optimization process.
[0030] Step S102: Calculate multiple groups of candidate data transmission sequences for each device node according to the node status evaluation index, and optimize the multiple groups of candidate data transmission sequences to obtain a target data transmission sequence for each device node;
[0031] Specifically, multiple phase rotation sequences are generated for each device node according to the node status evaluation index. Phase rotation sequence is a signal processing method that can effectively adjust and optimize the transmission characteristics of the signal. The generation of the sequence is based on a comprehensive analysis of the status of each node, such as signal strength, energy usage and data transmission rate, so as to ensure that each sequence can adapt to the current status of its corresponding node to the greatest extent. The phase rotation sequence is multiplied with the target transmission data, and the adjusted signal sequence is combined with the data actually required to be transmitted to generate multiple groups of candidate data transmission sequences for each device node. Based on the calculation of the peak-to-average power ratio of each group of candidate data transmission sequences, these candidate data transmission sequences are evaluated and compared to select the best sequence. The peak-to-average power ratio is one of the key indicators for measuring signal quality and can reflect the stability and efficiency of the signal during transmission. By comparing the peak-to-average power ratio of each candidate sequence, it is determined which sequences can effectively control power consumption while maintaining signal strength. From the candidate data transmission sequences, those with the largest peak-to-average power ratio are selected as the target data transmission sequence for each device node. It ensures that during the dynamic routing optimization process, each node can use the data transmission sequence that best suits its current state, thereby optimizing the energy efficiency and performance of the entire network while ensuring the quality of data transmission.
[0032] Step S103, controlling each device node to transmit device signals through a target data transmission sequence, and performing Tethering behavior detection and identification on the initial device network topology structure to obtain a Tethering behavior identification result;
[0033] Specifically, each device node is controlled to transmit device signals through the target data transmission sequence. Ensure that the data transmission of each node is strictly carried out according to the optimized path, so as to improve network efficiency and reduce latency. Collect traffic data for the initial device network topology, including capturing the data traffic size, transmission speed, and detailed transmission path of the data packet in the network. After data cleaning and classification, the collected traffic data is processed to remove noise and non-critical information to obtain the target network traffic data. The target network traffic data is feature extracted through the preset bidirectional long short-term memory network (Bi-LSTM). As an efficient time series analysis model, Bi-LSTM can effectively capture the changing characteristics of network traffic data in the time dimension. These time series features are an important basis for understanding network behavior and predicting future trends. The application of Bi-LSTM can not only reveal the instantaneous characteristics of network traffic, but also capture trends and patterns in a longer time range. The network traffic time series features are input into the preset support vector machine (SVM) model for Tethering behavior detection and identification. As a machine learning classification algorithm, SVM is suitable for processing complex pattern recognition problems. By analyzing the timing characteristics of network traffic, the SVM model can accurately identify Tethering behavior, that is, unauthorized or abnormal network access behavior, thereby providing important support for network security management.
[0034] Step S104: performing routing path interactive topology optimization on each device node according to the Tethering behavior identification result, generating multiple sub-routing paths for each device node, and generating a corresponding first device network topology structure through the multiple sub-routing paths of each device node;
[0035] Specifically, according to the Tethering behavior identification results, node task analysis is performed on each device node to understand the function and role of each node and their behavior patterns in the network. For example, some nodes may undertake the important task of data relay, while other nodes may be more involved in data collection and processing. Based on the node task information, the initial aggregation link is constructed to effectively organize the data flow in the network and ensure that the data can be efficiently and reliably transmitted from the source to the destination. At the same time, the link weight is adjusted to optimize the priority and path selection of data transmission, so that the data transmission is more in line with the current operation status and needs of the network. Based on the optimized target aggregation link, multiple sub-routing paths are generated for each device node. Sub-routing paths are generated based on comprehensive consideration of the current state of the network, the task information of the device, and the weight distribution of the aggregation link. The generated sub-routing paths can provide more options and flexibility for data transmission, thereby improving the overall performance and robustness of the network. The topological path of the initial device network topology is expanded through the sub-routing path of each device node. The newly generated sub-routing path is integrated into the original network structure to form a more optimized and efficient first device network topology. The new topology can not only better adapt to the actual operation needs of the network, but also improve the reliability and security of the network. According to the Tethering behavior identification results, the routing path of each device node is effectively optimized interactively, thereby achieving dynamic optimization and improvement of the entire network structure.
[0036] Step S105: performing multi-level data analysis and feature extraction on the first device network topology structure to obtain multi-level fusion features, and dynamically optimizing the routing of the first device network topology structure through the multi-level fusion features to obtain a second device network topology structure;
[0037] Specifically, a multi-level data analysis is performed on the network topology of the first device. This includes basic analysis of the network structure, such as node connectivity, link status, and network traffic, as well as more detailed aspects of network operations, such as data transmission delay, link stability, and node load capacity. Multi-level data analysis aims to fully understand the current state of the network and potential optimization space, laying the foundation for subsequent feature extraction. Feature extraction is performed on the target multi-level data. By applying data processing techniques such as principal component analysis (PCA) and cluster analysis, key features are extracted from complex network data. Multi-level fusion features are an accurate description of the network state, including key nodes of the network topology, data transmission efficiency of links, and distribution patterns of network traffic. The extraction of these features provides a decision-making basis for subsequent dynamic routing optimization. The multi-level fusion features are input into the preset random forest network for dynamic routing optimization analysis. Random forests can process complex data sets and generate accurate prediction models. In this process, random forests generate the first dynamic routing optimization strategy by analyzing multi-level fusion features. This strategy is based on a deep understanding of the current network state and aims to find the best data transmission path to improve the overall efficiency and reliability of the network. According to the first dynamic routing optimization strategy generated by the random forest network, the first device network topology is adjusted and optimized. This includes adjusting the data flow in the network, optimizing the connection method between nodes, and reconfiguring the key path of the network. Through these adjustments, the network topology is optimized to form a more efficient and stable second device network topology.
[0038] Step S106: Perform network status monitoring and performance analysis on the second device network topology structure to obtain network status performance indicators, and globally optimize the second device network topology structure according to the network status performance indicators to generate a target device network topology structure.
[0039] Specifically, the network status of the second device network topology is monitored, and multi-dimensional data in the network is collected, such as data transmission rate, node response time, link stability, and traffic distribution. Based on the collected multi-dimensional network status data, the network performance of the second device network topology is analyzed. The efficiency and reliability of the network path are evaluated to obtain a network path evaluation value. Network performance analysis can be performed through various algorithms, such as queue theory, network flow analysis, or prediction models based on machine learning. Through analysis, the performance of each network path is evaluated, such as transmission delay, data packet loss rate, and bandwidth utilization. These evaluation values provide an important basis for subsequent routing optimization. According to the network path evaluation value, the second dynamic routing optimization strategy for each device node is calculated. The formulation of the strategy needs to take into account the overall performance of the network and the specific conditions of each node, including the load of the node, the stability of the connection, and the priority of data transmission. These dynamic routing optimization strategies are intended to find the best data transmission path for each node to improve the efficiency and stability of the entire network. According to the second dynamic routing optimization strategy, the second device network topology is globally optimized to generate a target device network topology. This global optimization not only involves adjusting the routing strategy of a single node, but also includes the reconfiguration of the entire network topology. This may include redistributing network traffic, optimizing link allocation, or adjusting the connection method between nodes. Through comprehensive optimization, it can ensure that every part of the network operates in the best state, and finally generate the target device network topology.
[0040] In the embodiment of the present application, by conducting in-depth path learning and node information collection on the initial device network topology, combined with linear coding processing and minimum variance unbiased estimator, the state of each device node can be accurately evaluated. This not only improves the accuracy of node state evaluation, but also provides a reliable data basis for subsequent route optimization. By calculating and optimizing multiple sets of candidate data transmission sequences for each device node, the peak-to-average power ratio of data transmission is effectively reduced, and the quality and efficiency of data transmission are significantly improved. This optimization minimizes the distortion of the signal during transmission, thereby ensuring the stability and reliability of data transmission in the network. Not only can the device nodes be controlled to perform effective signal transmission, but also the Tethering behavior in the network can be effectively identified and managed through advanced traffic data collection and analysis technology. Such a management mechanism significantly improves the security and stability of the network and ensures the fair allocation of network resources. According to the Tethering behavior identification results, the device nodes can be targeted for routing path optimization to generate a more reasonable device network topology. This dynamic and flexible optimization strategy fully considers the real-time status and changes in the network, ensuring the efficiency and adaptability of the network structure. By using multi-level data analysis and feature extraction technology, we can understand the network status more comprehensively and deeply, thus providing richer and more accurate decision support for route optimization. Through real-time monitoring of network status and comprehensive performance analysis, we can timely discover potential problems in the network and perform global optimization based on network status performance indicators. Such global optimization not only improves the performance of the network, but also enhances the robustness and reliability of the network, thereby improving the accuracy of dynamic route optimization of the device.
[0041] In a specific embodiment, the process of executing step S101 may specifically include the following steps:
[0042] (1) Obtaining device connection relationships and device location information of multiple target devices, and constructing an initial device network topology structure based on the device connection relationships and device location information;
[0043] (2) performing path learning on the initial device network topology structure to determine multiple device nodes in the initial device network topology structure;
[0044] (3) Collecting neighbor node information of multiple device nodes respectively to obtain a node status data set of each device node. The node status data set includes: signal strength, energy usage, and data transmission rate;
[0045] (4) Linearly encode the node status data set of each device node using a preset linear encoder to obtain a node linear encoding set of each device node;
[0046] (5) The node linear code set of each device node is used to estimate the state through the preset minimum variance unbiased estimator to obtain the node state evaluation index of each device node.
[0047] Specifically, the connection relationship and location information of multiple target devices are obtained. For example, in a wireless sensor network, the physical location of each sensor node and its communication capability with other nodes are obtained through a network detection tool. Based on this information, an initial network topology diagram is constructed, which reflects the interconnection relationship between nodes and their distribution in space. The initial device network topology structure is path-learned to analyze and determine the communication paths between various device nodes in the network. By applying graph theory algorithms, for example, key nodes and potential communication bottlenecks in the network are identified. For example, the Dijkstra algorithm can help find the shortest path between two nodes, thereby optimizing data transmission efficiency. Neighbor node information is collected for device nodes. The operating status of each node is collected, including their signal strength, energy usage, and data transmission rate. The state data set of each device node is encoded by a preset linear encoder. The role of the linear encoder is to convert the original state data into a formatted and standardized representation, so as to facilitate subsequent processing and analysis. For example, Huffman coding is used to encode the signal strength data to reduce the storage and transmission costs of the data. The state of the encoded data set is estimated by a preset minimum variance unbiased estimator. Extract deeper information from the encoded data, such as the expected life of the equipment, communication quality, and potential failure risk. For example, for a router node in a data center, the minimum variance unbiased estimator can help predict the performance and failure probability under different load conditions.
[0048] In a specific embodiment, the process of executing step S102 may specifically include the following steps:
[0049] (1) Generate multiple phase rotation sequences for each device node according to the node status evaluation index, and obtain the target transmission data of each device node;
[0050] (2) performing product operations on multiple phase rotation sequences and target transmission data respectively to generate multiple sets of candidate data transmission sequences for each device node;
[0051] (3) calculating the peak-to-average power ratio of each group of candidate data transmission sequences respectively, and comparing multiple groups of candidate data transmission sequences according to the peak-to-average power ratio;
[0052] (4) Selecting the candidate data transmission sequence with the largest peak-to-average power ratio from multiple groups of candidate data transmission sequences as the target data transmission sequence for each device node.
[0053] Specifically, multiple phase rotation sequences are generated for each device node according to the node status evaluation index. The phase rotation sequence optimizes the quality and efficiency of signal transmission by changing the phase of the signal. For example, assuming that in a wireless communication network, a device node has a signal strength drop due to a change in physical position, and a specific phase rotation sequence is generated to compensate for the loss of signal strength, thereby maintaining the stability of communication. The target transmission data of each device node is obtained. The target transmission data refers to the actual data that the node needs to send or receive, which includes control instructions, sensor data, or data packets in network communication. The multiple phase rotation sequences and the target transmission data are multiplied to generate multiple sets of candidate data transmission sequences for each device node. The signal shape during the data transmission process is adjusted to optimize the performance of data transmission. Through the product operation, the optimization effect of phase rotation is applied to the actual data transmission, thereby improving the efficiency and reliability of transmission while ensuring data integrity. The generated candidate data transmission sequences are evaluated for performance. The main evaluation index is the peak-to-average power ratio, which is an important index for measuring signal transmission efficiency. The higher the peak-to-average power ratio, the higher the peak power of the signal under a given average power, which generally means that the signal has better penetration and coverage. Therefore, by calculating the peak-to-average power ratio of each set of candidate data transmission sequences, the performance of these sequences can be effectively evaluated and compared. From the candidate data transmission sequences, the sequence with the largest peak-to-average power ratio is selected as the target data transmission sequence for each device node. Such a selection ensures the highest efficiency and the best signal quality during data transmission. For example, selecting the data transmission sequence with the largest peak-to-average power ratio can ensure that the communication reaches the farthest distance and the best communication quality under limited energy consumption.
[0054] In a specific embodiment, the process of executing step S103 may specifically include the following steps:
[0055] (1) Control each device node to transmit device signals through the target data transmission sequence;
[0056] (2) Collecting traffic data of the initial device network topology structure to obtain initial network traffic data;
[0057] (3) Clean and classify the initial network traffic data to obtain multiple target network traffic data;
[0058] (4) Using a preset bidirectional long short-term memory network to perform feature weather analysis on multiple target network traffic data, multiple network traffic time series features are obtained;
[0059] (5) Input multiple network traffic time series features into the preset support vector machine model to detect and identify Tethering behavior and obtain Tethering behavior identification results.
[0060] Specifically, each device node is controlled to transmit device signals through the target data transmission sequence. Each node needs to communicate according to the predetermined data transmission scheme to ensure accurate and efficient data transmission. For example, in a network system composed of multiple wireless sensors, each sensor (device node) has specific data transmission requirements and priorities. In order to meet these requirements, the data transmission sequence of each node is adjusted according to its specific situation (such as location, power status, task importance, etc.). This involves changing the size of the data packet, adjusting the transmission frequency, or changing the transmission path. The initial device network topology is used to collect traffic data and collect data transmission records about each node, including the time, size, destination, and transmission path of the data transmission. The initial network traffic data is cleaned and classified. Data cleaning mainly removes those erroneous or irrelevant data to ensure the accuracy of subsequent analysis. Data classification is to classify data into different categories based on the characteristics of the data, such as transmission type, traffic size, or source and destination address. For example, data traffic is classified into regular data transmission, video traffic, or abnormal traffic. This helps to better understand the data flow pattern in the network and identify potential network security threats. The network traffic data is feature extracted through a preset bidirectional long short-term memory network (Bi-LSTM). Bi-LSTM is an efficient machine learning model suitable for processing time series data. It can consider both historical and future information of the data, which helps to understand and predict the dynamic changes of network traffic. For example, by applying Bi-LSTM, the changing trend of network traffic in different time periods can be analyzed to predict the future load of the network or identify abnormal traffic patterns. The network traffic time series features are input into the preset support vector machine (SVM) model to detect and identify Tethering behavior. SVM is a classification algorithm suitable for the analysis of high-dimensional data. The SVM model can identify Tethering behavior, that is, the behavior of unauthorized devices attempting to connect to the network, based on the time series features of network traffic. For example, if in network traffic analysis, SVM identifies that certain data flow patterns are significantly different from normal patterns, this indicates that a device is making an unauthorized network connection.
[0061] In a specific embodiment, the process of executing step S104 may specifically include the following steps:
[0062] (1) Perform node task analysis on each device node based on the Tethering behavior identification results to obtain the node task information of each device node;
[0063] (2) Establish an initial aggregation link based on the node task information, and adjust the weight of the initial aggregation link to obtain the target aggregation link;
[0064] (3) Generate multiple sub-routing paths for each device node based on the target aggregation link;
[0065] (4) The initial device network topology structure is expanded by topology path through multiple sub-routing paths of each device node to generate a corresponding first device network topology structure.
[0066] Specifically, node task analysis is performed on each device node based on the Tethering behavior identification results. Tethering behavior, that is, the behavior of devices illegally sharing network connections, can cause network congestion, security risks, or unfair resource allocation. By analyzing the identification results of these behaviors, it is possible to understand which device nodes are involved in improper network use or which nodes are affected. For example, if a sensor node is identified as performing Tethering behavior, the task of this node needs to be re-evaluated to determine whether it can still effectively complete the task of data collection and transmission. Based on the node's task information, an initial aggregation link is established. An aggregation link refers to a network path that transmits data from multiple source nodes to a specific target node (such as a data center or a control station). In this process, the data transmission requirements, network status, and possible optimization space of each node need to be considered. For example, an aggregation link can be established based on the data reporting frequency and importance of each node. After the initial aggregation link is established, weight adjustment is performed to adjust the link usage weight based on network traffic, node importance, or data priority. Through adjustment, it is ensured that important or urgent data can be transmitted first, thereby improving the overall efficiency and responsiveness of the network. Based on the optimized target aggregation link, multiple sub-routing paths are generated for each device node. The generation of sub-routing paths needs to consider the current state of the network, the connection mode of each node, and the data transmission requirements. The initial device network topology is expanded through multiple sub-routing paths for each device node. The newly generated sub-routing paths are integrated into the existing network structure to form a more efficient and flexible first device network topology. For example, by adding or adjusting sub-routing paths, data traffic can be effectively dispersed and congestion on specific links can be reduced, thereby improving the data processing capacity and stability of the entire network.
[0067] In a specific embodiment, the process of executing step S105 may specifically include the following steps:
[0068] (1) Performing multi-level data analysis on the network topology structure of the first device to obtain target multi-level data;
[0069] (2) Extract features from the target multi-level data to obtain multi-level fusion features;
[0070] (3) Perform dynamic routing optimization analysis on the multi-level fusion features through a preset random forest network to obtain the first dynamic routing optimization strategy;
[0071] (4) Dynamically optimize the first device network topology structure according to the first dynamic routing optimization strategy to obtain a second device network topology structure.
[0072] Specifically, a multi-level data analysis is performed on the network topology of the first device. The operation data of the network is collected and analyzed from different levels, including network traffic, connection status between nodes, data transmission rate, and possible network congestion. For example, the data interaction volume, transmission frequency, and corresponding delay time between different sensor nodes are important analysis indicators. Feature extraction is performed on the target multi-level data to identify the most influential information from a large amount of raw data, which can represent the key characteristics of the network status. By using data mining techniques such as principal component analysis (PCA) or cluster analysis, key features such as network traffic patterns and node communication patterns are identified. For example, in a complex data center network, feature extraction can help identify the main drivers of data traffic in the network, such as data requests for specific applications or traffic peaks in a specific time period. Dynamic routing optimization analysis is performed on multi-level fusion features through a preset random forest network. Random forest is a machine learning model that improves the accuracy of prediction by building multiple decision trees and aggregating their results. Random forest can analyze network characteristics and predict the effects of different routing options, thereby providing decision support for routing optimization. For example, random forests can help analyze which routing options can minimize data transmission delays under different network congestion conditions. According to the first dynamic routing optimization strategy provided by the random forest network, the first device network topology is adjusted to obtain an optimized second device network topology. According to the analysis results of the machine learning model, the data transmission path in the network is adjusted to optimize the connection between nodes, including increasing or decreasing the number of nodes in the network. This adjustment is to ensure that the network can operate efficiently in various situations while reducing the risk of possible network failures. For example, in a wireless sensor network, according to the optimization strategy, it is necessary to adjust the data transmission frequency of certain sensor nodes or change their data transmission paths to avoid network congestion and ensure the timely transmission of critical data.
[0073] In a specific embodiment, the process of executing step S106 may specifically include the following steps:
[0074] (1) Performing network status monitoring on the network topology structure of the second device to obtain multi-dimensional network status data;
[0075] (2) performing a network performance analysis on the network topology structure of the second device according to the multi-dimensional network status data to obtain a network path evaluation value;
[0076] (3) calculating a second dynamic routing optimization strategy for each device node according to the network path evaluation value;
[0077] (4) Globally optimize the second device network topology structure according to the second dynamic routing optimization strategy to generate a target device network topology structure.
[0078] Specifically, the network status of the second device network topology is monitored to obtain multi-dimensional network status data. Network monitoring tools and technologies are used to collect various data about network operation, including network traffic, communication delay between nodes, packet loss rate, and link utilization. Based on the multi-dimensional network status data, the network performance of the second device network topology is analyzed, and the network path evaluation value is obtained, and data analysis methods and performance evaluation models are used, such as network simulation tools, delay and bandwidth calculation formulas, etc. For example, in a large data center network, the performance of each path in the network is evaluated by analyzing the efficiency of data transmission between different servers, and possible bottlenecks or unstable links are identified. These path evaluation values provide important information about which parts of the network need to be optimized. The second dynamic routing optimization strategy for each device node is calculated based on the network path evaluation value. The current operation status of the network, the performance indicators of each node, and the data traffic pattern are comprehensively considered to formulate the best data transmission path and network configuration. For example, for a wireless sensor network, if it is found that the data transmission between certain sensors is often interfered, it is necessary to adjust the data transmission path of these nodes, or add relay nodes to improve the overall stability and efficiency of the network. According to the second dynamic routing optimization strategy, the second device network topology is globally optimized to generate the target device network topology. The connection mode in the network is reconfigured, the distribution of data traffic is adjusted, or the roles of certain key nodes are changed. For example, by adjusting the connection mode between servers and enhancing the processing power of core routers, the data processing rate of the network can be significantly improved and the delay can be reduced. Global optimization not only improves the performance of the network, but also enhances its anti-interference ability and scalability.
[0079] The above describes the dynamic routing optimization method in the embodiment of the present application. The following describes the dynamic routing optimization device in the embodiment of the present application. Figure 2 In the embodiments of the present application, one embodiment of the dynamic routing optimization device includes:
[0080] The acquisition module 201 is used to perform path learning and neighbor node information acquisition on the preset initial device network topology structure, obtain a node state data set of each device node, and perform linear coding processing and state estimation on the node state data set of each device node to obtain a node state evaluation index of each device node;
[0081] A calculation module 202 is used to calculate multiple groups of candidate data transmission sequences for each device node according to the node status evaluation index, and optimize the multiple groups of candidate data transmission sequences to obtain a target data transmission sequence for each device node;
[0082] The identification module 203 is used to control each device node to perform device signal transmission through the target data transmission sequence, and perform Tethering behavior detection and identification on the initial device network topology structure to obtain a Tethering behavior identification result;
[0083] A generation module 204 is used to perform routing path interaction topology optimization on each device node according to the Tethering behavior identification result, generate multiple sub-routing paths for each device node, and generate a corresponding first device network topology structure through the multiple sub-routing paths of each device node;
[0084] The analysis module 205 is used to perform multi-level data analysis and feature extraction on the first device network topology structure to obtain multi-level fusion features, and dynamically optimize the routing of the first device network topology structure through the multi-level fusion features to obtain a second device network topology structure;
[0085] The optimization module 206 is used to perform network status monitoring and performance analysis on the second device network topology structure to obtain network status performance indicators, and to globally optimize the second device network topology structure according to the network status performance indicators to generate a target device network topology structure.
[0086] Through the synergy of the above components, through in-depth path learning and node information collection of the initial device network topology, combined with linear coding processing and minimum variance unbiased estimator, the state of each device node can be accurately evaluated. This not only improves the accuracy of node state evaluation, but also provides a reliable data basis for subsequent routing optimization. By calculating and optimizing multiple sets of candidate data transmission sequences for each device node, the peak-to-average power ratio of data transmission is effectively reduced, and the quality and efficiency of data transmission are significantly improved. This optimization minimizes the distortion of the signal during transmission, thereby ensuring the stability and reliability of data transmission in the network. Not only can the device nodes be controlled to perform effective signal transmission, but also the Tethering behavior in the network can be effectively identified and managed through advanced traffic data collection and analysis technology. Such a management mechanism significantly improves the security and stability of the network and ensures the fair allocation of network resources. According to the Tethering behavior identification results, the device nodes can be targeted for routing path optimization to generate a more reasonable device network topology. This dynamic and flexible optimization strategy fully considers the real-time status and changes in the network, ensuring the efficiency and adaptability of the network structure. By using multi-level data analysis and feature extraction technology, we can understand the network status more comprehensively and deeply, thus providing richer and more accurate decision support for route optimization. Through real-time monitoring of network status and comprehensive performance analysis, we can timely discover potential problems in the network and perform global optimization based on network status performance indicators. Such global optimization not only improves the performance of the network, but also enhances the robustness and reliability of the network, thereby improving the accuracy of dynamic route optimization of the device.
[0087] The present application also provides a dynamic routing optimization device, which includes a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the dynamic routing optimization method in the above-mentioned embodiments.
[0088] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the dynamic routing optimization method.
[0089] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0090] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.
[0091] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims
1. A dynamic routing optimization method, characterized in that: The dynamic routing optimization method comprises: Perform path learning and neighbor node information collection on the preset initial device network topology structure to obtain a node state data set of each device node, and perform linear encoding processing and state estimation on the node state data set of each device node to obtain a node state evaluation index for each device node; Calculate multiple groups of candidate data transmission sequences for each device node according to the node status evaluation index, and optimize the multiple groups of candidate data transmission sequences to obtain a target data transmission sequence for each device node; Controlling each device node to transmit device signals through the target data transmission sequence, and performing Tethering behavior detection and identification on the initial device network topology structure to obtain a Tethering behavior identification result; According to the Tethering behavior identification result, each device node is respectively optimized for routing path interaction topology, multiple sub-routing paths are generated for each device node, and a corresponding first device network topology structure is generated through the multiple sub-routing paths of each device node; Performing multi-level data analysis and feature extraction on the first device network topology structure to obtain multi-level fusion features, and dynamically optimizing routing of the first device network topology structure by using the multi-level fusion features to obtain a second device network topology structure; The network status of the second device network topology structure is monitored and the performance is analyzed to obtain a network status performance index, and the second device network topology structure is globally optimized according to the network status performance index to generate a target device network topology structure.
2. The dynamic routing optimization method according to claim 1, characterized in that: The preset initial device network topology structure is subjected to path learning and neighbor node information collection to obtain a node state data set of each device node, and the node state data set of each device node is subjected to linear coding processing and state estimation to obtain a node state evaluation index of each device node, including: Acquire device connection relationships and device location information of multiple target devices, and construct an initial device network topology structure according to the device connection relationships and the device location information; Performing path learning on the initial device network topology structure to determine a plurality of device nodes in the initial device network topology structure; Collecting neighbor node information of the plurality of device nodes respectively to obtain a node status data set of each device node, wherein the node status data set includes: signal strength, energy usage, and data transmission rate; Performing linear encoding processing on the node status data set of each device node respectively through a preset linear encoder to obtain a node linear encoding set of each device node; The node linear code set of each device node is respectively used to perform state estimation through a preset minimum variance unbiased estimator to obtain a node state evaluation index for each device node.
3. The dynamic routing optimization method according to claim 1, characterized in that: The step of calculating multiple groups of candidate data transmission sequences for each device node according to the node status evaluation index, and optimizing the multiple groups of candidate data transmission sequences to obtain a target data transmission sequence for each device node includes: Generate multiple phase rotation sequences for each device node according to the node status evaluation index, and obtain target transmission data for each device node; Performing product operations on the multiple phase rotation sequences and the target transmission data respectively to generate multiple groups of candidate data transmission sequences for each device node; Calculating a peak-to-average power ratio of each group of candidate data transmission sequences respectively, and comparing the multiple groups of candidate data transmission sequences according to the peak-to-average power ratio; A candidate data transmission sequence with the largest peak-to-average power ratio is selected from the multiple groups of candidate data transmission sequences as the target data transmission sequence for each device node.
4. The dynamic routing optimization method according to claim 1, characterized in that: The controlling each device node to transmit a device signal through the target data transmission sequence, and performing Tethering behavior detection and identification on the initial device network topology structure to obtain a Tethering behavior identification result includes: Controlling each device node to perform device signal transmission through the target data transmission sequence; Collecting flow data of the initial device network topology structure to obtain initial network flow data; Performing data cleaning and data classification on the initial network flow data to obtain a plurality of target network flow data; Extracting features of the plurality of target network traffic data through a preset bidirectional long short-term memory network to obtain a plurality of network traffic time series features; The multiple network traffic time series features are input into a preset support vector machine model to perform Tethering behavior detection and identification to obtain a Tethering behavior identification result.
5. The dynamic routing optimization method according to claim 1, characterized in that: The method of performing routing path interactive topology optimization on each device node according to the Tethering behavior identification result, generating multiple sub-routing paths for each device node, and generating a corresponding first device network topology structure through the multiple sub-routing paths of each device node includes: Perform node task analysis on each device node according to the Tethering behavior identification result to obtain node task information of each device node; Establishing an initial aggregation link according to the node task information, and adjusting the weight of the initial aggregation link to obtain a target aggregation link; Generating a plurality of sub-routing paths for each device node according to the target aggregation link; The initial device network topology structure is topologically expanded through multiple sub-routing paths of each device node to generate a corresponding first device network topology structure.
6. The dynamic routing optimization method according to claim 1, characterized in that: The performing multi-level data analysis and feature extraction on the first device network topology structure to obtain multi-level fusion features, and performing dynamic routing optimization on the first device network topology structure by using the multi-level fusion features to obtain a second device network topology structure, includes: Performing multi-level data analysis on the network topology structure of the first device to obtain target multi-level data; Extracting features from the target multi-level data to obtain multi-level fusion features; Performing dynamic routing optimization analysis on the multi-level fusion features through a preset random forest network to obtain a first dynamic routing optimization strategy; Dynamically optimize the first device network topology structure according to the first dynamic routing optimization strategy to obtain a second device network topology structure.
7. The dynamic routing optimization method according to claim 6, characterized in that: The performing of network status monitoring and performance analysis on the second device network topology structure to obtain a network status performance indicator, and performing global optimization on the second device network topology structure according to the network status performance indicator to generate a target device network topology structure, includes: Performing network status monitoring on the second device network topology structure to obtain multi-dimensional network status data; Performing a network performance analysis on the second device network topology structure according to the multi-dimensional network status data to obtain a network path evaluation value; Calculate a second dynamic routing optimization strategy for each device node according to the network path evaluation value; The second device network topology structure is globally optimized according to the second dynamic routing optimization strategy to generate a target device network topology structure.
8. A dynamic routing optimization device, characterized in that: The dynamic routing optimization device comprises: The acquisition module is used to perform path learning and neighbor node information acquisition on the preset initial device network topology structure, obtain the node state data set of each device node, and perform linear coding processing and state estimation on the node state data set of each device node to obtain the node state evaluation index of each device node; A calculation module, used to calculate multiple groups of candidate data transmission sequences for each device node according to the node status evaluation index, and optimize the multiple groups of candidate data transmission sequences to obtain a target data transmission sequence for each device node; an identification module, used to control each device node to perform device signal transmission through the target data transmission sequence, and perform Tethering behavior detection and identification on the initial device network topology structure to obtain a Tethering behavior identification result; A generation module, configured to perform routing path interaction topology optimization on each device node according to the Tethering behavior identification result, generate multiple sub-routing paths for each device node, and generate a corresponding first device network topology structure through the multiple sub-routing paths of each device node; an analysis module, configured to perform multi-level data analysis and feature extraction on the first device network topology structure to obtain multi-level fusion features, and dynamically optimize routing of the first device network topology structure by using the multi-level fusion features to obtain a second device network topology structure; The optimization module is used to perform network status monitoring and performance analysis on the second device network topology structure to obtain network status performance indicators, and to globally optimize the second device network topology structure based on the network status performance indicators to generate a target device network topology structure.
9. A dynamic routing optimization device, characterized in that: The dynamic routing optimization device comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the dynamic routing optimization device to execute the dynamic routing optimization method according to any one of claims 1 to 7.
10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the dynamic routing optimization method according to any one of claims 1 to 7 is implemented.
Citation Information
Patent Citations
Communication device, communication method, and communication system
CN116648957A
Soft bus networking connection method and device, equipment and storage medium
CN117527479A
Selection of network routing topology
US20200228409A1
Method for evaluating network quality, electronic device and storage medium
US20230188426A1
