A flow management method and system based on signal characteristic analysis

Abstract

The present application discloses a traffic management method and system based on signal feature analysis, which relates to the field of traffic prediction. The method comprises: collecting traffic data corresponding to nodes at different levels in a target area, wherein the nodes at different levels comprise core routers, distribution switches and user equipment terminals; performing signal feature analysis operations on the traffic data of each level node to obtain traffic feature information corresponding to each level node; performing traffic classification according to the traffic feature information corresponding to each level node to determine the traffic type and the traffic scale corresponding to the traffic type; obtaining the traffic prediction type and the traffic prediction scale corresponding to the traffic prediction type through a traffic prediction model according to the traffic type and the traffic scale corresponding to each level node; and adjusting the bandwidth allocation strategy and the hierarchical node structure distribution according to the traffic prediction type and the traffic prediction scale corresponding to each level node.

Description

Technical Field

[0001] The present specification relates to the field of traffic prediction, and more specifically, the present application relates to a traffic management method and system based on signal feature analysis. Background Art

[0002] Traffic management in related technologies collects network traffic data and monitors network bandwidth usage through simple traffic monitoring tools, such as SNMP (Simple Network Management Protocol). Static rules, such as QoS (Quality of Service) policies, are set based on experience or preset thresholds to prioritize or limit certain types of traffic. When traffic reaches the preset threshold, bandwidth restriction measures are taken, such as reducing the priority of non-critical applications. Traffic management policies are usually relatively fixed and lack the ability to dynamically adjust based on real-time data.

[0003] Therefore, it is necessary to propose a traffic management method and system that is dynamically adjusted according to real-time data. Summary of the invention

[0004] A series of simplified concepts are introduced in the Summary of the Invention section, which will be further described in detail in the Detailed Description of the Invention section. The Summary of the Invention section of this application does not mean to attempt to define the key features and essential technical features of the claimed technical solution, nor does it mean to attempt to determine the scope of protection of the claimed technical solution.

[0005] In a first aspect, the present application proposes a traffic management method based on signal feature analysis, the method comprising:

[0006] Collecting traffic data corresponding to nodes at different levels in the target area, wherein the nodes at different levels include core routers, distribution switches, and user equipment terminals;

[0007] Perform signal feature analysis on the traffic data of each level node to obtain the traffic feature information corresponding to each level node;

[0008] Classify the traffic according to the traffic characteristic information corresponding to each level node to determine the traffic type and the traffic scale corresponding to the traffic type;

[0009] According to the above traffic type and the above traffic scale corresponding to each level node, a traffic prediction model is used to obtain a traffic prediction type and a traffic prediction scale corresponding to the traffic prediction type;

[0010] The bandwidth allocation strategy and the hierarchical node structure distribution are adjusted according to the above traffic prediction type and the above traffic prediction scale corresponding to each hierarchical node.

[0011] In a feasible implementation manner, the above-mentioned collection of traffic data corresponding to nodes at different levels in the target area includes:

[0012] Obtain the core router traffic data corresponding to the above core routers in the target area through the NetFlow / SFlow protocol;

[0013] Obtain the distribution switch traffic data corresponding to the above distribution switches in the target area through RSPAN;

[0014] The DHCP server log and Wi-Fi analysis tools are used to obtain the user equipment terminal traffic data corresponding to the above user equipment terminals in the target area.

[0015] In a feasible implementation manner, the signal feature analysis operation is performed on the traffic data of each level node to obtain the traffic feature information corresponding to each level node, including:

[0016] Performing a signal characteristic analysis operation based on the core router traffic data to obtain the period characteristic, dominant frequency characteristic and energy distribution characteristic of the core router;

[0017] Performing a signal feature analysis operation based on the above distribution switch traffic data to obtain the mutation feature and noise feature of the above distribution exchange machine;

[0018] A signal feature analysis operation is performed based on the terminal traffic data of the above-mentioned user equipment to obtain the periodic characteristics, dominant frequency characteristics, noise characteristics and mutation characteristics of the above-mentioned user terminal equipment, wherein the above-mentioned periodic characteristics and dominant frequency characteristics are determined by Fourier transform analysis, the above-mentioned energy distribution is determined by power spectral density analysis, the above-mentioned mutation characteristics are determined by CUSUM algorithm, and the above-mentioned noise characteristics are determined based on a statistical model of normal traffic behavior.

[0019] In a feasible implementation manner, traffic classification is performed according to the traffic characteristic information corresponding to the core router to determine the traffic type and the traffic scale corresponding to the traffic type, including:

[0020] Inputting the periodic characteristics, dominant frequency characteristics and energy distribution characteristics of the core router into the first K-means clustering module to obtain a preliminary classification cluster of the first traffic data;

[0021] All the first traffic data preliminary classification clusters are input into the support vector machine module to obtain the traffic type corresponding to the core router and the traffic scale corresponding to the traffic type.

[0022] In a feasible implementation manner, traffic classification is performed according to the traffic characteristic information corresponding to the distribution switch to determine the traffic type and the traffic scale corresponding to the traffic type, including:

[0023] The mutation characteristics and noise characteristics of the above allocation interaction machine are input into the LSTM module to obtain the time series change characteristics;

[0024] The above time series change characteristics are input into the CNN module to obtain the traffic type corresponding to the above distribution switch and the traffic scale corresponding to the above traffic type.

[0025] In a feasible implementation manner, traffic classification is performed according to the traffic characteristic information corresponding to the user equipment terminal to determine the traffic type and the traffic scale corresponding to the traffic type, including:

[0026] Inputting the periodic characteristics, dominant frequency characteristics, noise characteristics and mutation characteristics of the user terminal device into the second K-means clustering module to obtain a preliminary classification cluster of the second traffic data;

[0027] All the second traffic data preliminary classification clusters are input into the random forest module to obtain the traffic type corresponding to the above user equipment terminal and the traffic scale corresponding to the above traffic type.

[0028] In a feasible implementation manner, the specific steps of adjusting the bandwidth allocation strategy according to the traffic prediction type and the traffic prediction scale corresponding to each level node include:

[0029] Determine the mainstream application and the peak time period corresponding to the mainstream application based on the above traffic prediction type and the above traffic prediction scale corresponding to all hierarchical nodes;

[0030] Get the number of nodes under each level node;

[0031] Determine the node weight coefficient corresponding to each level node based on the number of the above nodes under each level node and the level to which each level node belongs;

[0032] Based on the node weight coefficient corresponding to each level node, the mainstream applications, and the peak time periods corresponding to the mainstream applications, the bandwidth allocation strategy of each level node in different time periods is adjusted.

[0033] In a feasible implementation manner, the specific steps of adjusting the hierarchical node structure distribution according to the above-mentioned traffic prediction type and the above-mentioned traffic prediction scale corresponding to each hierarchical node include:

[0034] Analyze the historical performance parameters of all nodes at all levels under different traffic types, where the historical performance parameters include processing power, storage capacity, and transmission rate;

[0035] Perform node performance evaluation according to the traffic prediction type, the traffic prediction scale and the historical performance parameters to obtain a performance evaluation result;

[0036] Node optimization and topology optimization are carried out based on the above performance evaluation results.

[0037] In a feasible implementation, node optimization and topology optimization are performed based on the above performance evaluation results, including:

[0038] Determine the bottleneck node based on the above performance evaluation results;

[0039] Get the upgrade cost information of bottleneck nodes;

[0040] Determine a first optimized path using the shortest path algorithm according to the bottleneck node;

[0041] Determine a second optimization path using a minimum spanning tree algorithm based on the bottleneck node;

[0042] Determine the upgrade cost corresponding to the first optimization path and the upgrade cost corresponding to the second optimization path according to the upgrade cost information;

[0043] Node optimization and topology optimization are performed based on the principle of minimizing upgrade costs.

[0044] In a second aspect, the present application proposes a traffic management system based on signal feature analysis, comprising:

[0045] A collecting unit, used to collect the traffic data corresponding to nodes at different levels in the target area, wherein the nodes at different levels include core routers, distribution switches and user equipment terminals;

[0046] An analysis unit, used to perform a signal characteristic analysis operation on the flow data of each level node to obtain flow characteristic information corresponding to each level node;

[0047] A classification unit, used to classify traffic according to the traffic characteristic information corresponding to each level node to determine the traffic type and the traffic scale corresponding to the traffic type;

[0048] A prediction unit, used for obtaining a traffic prediction type and a traffic prediction scale corresponding to the traffic prediction type through a traffic prediction model according to the traffic type and the traffic scale corresponding to each level node;

[0049] The adjustment unit is used to adjust the bandwidth allocation strategy and the hierarchical node structure distribution according to the above-mentioned traffic prediction type and the above-mentioned traffic prediction scale corresponding to each hierarchical node.

[0050] In summary, the traffic management method based on signal feature analysis of the embodiment of the present application includes: collecting the corresponding traffic data under different levels of nodes in the target area, wherein the above-mentioned different levels of nodes include core routers, distribution switches and user equipment terminals; performing signal feature analysis operations on the traffic data of each level node to obtain the traffic feature information corresponding to each level node; classifying the traffic according to the above-mentioned traffic feature information corresponding to each level node to determine the traffic type and the traffic scale corresponding to the above-mentioned traffic type; obtaining the traffic prediction type and the traffic prediction scale corresponding to the traffic prediction type through the traffic prediction model according to the above-mentioned traffic type and the above-mentioned traffic scale corresponding to each level node; adjusting the bandwidth allocation strategy and the hierarchical node structure distribution according to the above-mentioned traffic prediction type and the above-mentioned traffic prediction scale corresponding to each level node. Traditional traffic management methods rely on static rules and simple thresholds to respond to traffic changes. This method is often not sensitive enough when dealing with sudden or unconventional traffic. By using signal feature analysis and traffic prediction models, the present application can predict traffic trends and peak hours in advance, thereby achieving more dynamic and proactive bandwidth adjustment and network optimization. This predictive management method enables the network to make appropriate resource allocation before the traffic reaches its peak, avoiding potential congestion and service quality degradation. The present application uses complex signal features such as periodic features, mutation features, noise features, dominant frequency features, and energy distribution features to analyze traffic. This in-depth feature analysis provides more sophisticated traffic insights than traditional methods, allowing network administrators to understand the essential characteristics and behavior patterns of traffic. For example, by analyzing the dominant frequency and energy distribution features, important real-time communication or large data transmission tasks can be more accurately identified and prioritized. The traffic classification process implemented in the present application is not only based on the quantity of traffic, but also combines the quality and type of traffic, and further estimates the future traffic scale through the traffic prediction model. This allows the bandwidth allocation strategy to be more refined and personalized, and differentiated management of traffic demands of different types and scales is performed, thereby optimizing resource utilization efficiency and improving user experience. According to the traffic prediction results, the present application can flexibly adjust the bandwidth allocation strategy and hierarchical node structure distribution. This flexibility allows the network to increase bandwidth support or reconfigure the network structure to cope with the predicted traffic growth when necessary, while saving resources and reducing unnecessary cost expenditures when the traffic is low. By implementing the method of the present application, the overall performance and stability of the network are improved, network management is more intelligent and automated, and can effectively cope with the changing needs in modern complex network environments. Compared with traditional technologies, this traffic management strategy based on deep analysis and prediction has significantly improved the efficiency of network resource utilization and the quality of network services, especially in high-demand scenarios such as big data and cloud computing.

[0051] The traffic management method based on signal feature analysis proposed in this application, and other advantages, objectives and features of this application will be reflected in part through the following description, and in part will also be understood by technical personnel in this field through research and practice of this application. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present specification. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0053] Figure 1 A flow chart of a traffic management method based on signal feature analysis provided in an embodiment of the present application;

[0054] Figure 2 A schematic diagram of the structure of a traffic management system based on signal feature analysis provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] 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" and "having" and any of their variations 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. The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments.

[0056] See also Figure 1 , which is a flow diagram of a traffic management method based on signal feature analysis provided in an embodiment of the present application, which may specifically include:

[0057] S110, collecting traffic data corresponding to nodes at different levels in the target area, wherein the nodes at different levels include core routers, distribution switches, and user equipment terminals;

[0058] For example, traffic data of nodes at all levels in the target area are first collected. These nodes include core routers, distribution switches, and user equipment terminals. The collected data will include but is not limited to the data flow size of each node, the timestamp of the flow, and other information. These data are the basis for subsequent analysis and can reflect the operation status of the network at different levels.

[0059] S120, performing a signal feature analysis operation on the traffic data of each level node to obtain traffic feature information corresponding to each level node;

[0060] Exemplarily, the flow data collected in step S110 is subjected to signal feature analysis. The purpose of the analysis is to obtain flow feature information, which includes periodic features, mutation features, noise features, dominant frequency features, and energy distribution features. These features provide a deeper understanding of the flow data, such as periodic changes in data flow, frequency distribution, and abnormal changes in flow data.

[0061] S130, classifying traffic according to the traffic feature information corresponding to each level node to determine the traffic type and the traffic scale corresponding to the traffic type;

[0062] Exemplarily, the system uses the traffic feature information obtained from step S120 to classify the traffic. This classification process helps determine the type of traffic and the corresponding traffic scale. For example, video traffic and audio traffic can be distinguished based on the dominant frequency characteristics in the data stream, and possible data emergencies can be identified based on mutation characteristics.

[0063] S140, obtaining a traffic prediction type and a traffic prediction scale corresponding to the traffic prediction type through a traffic prediction model according to the traffic type and the traffic scale corresponding to each level node;

[0064] Exemplarily, a traffic prediction model is used to predict future traffic types and traffic scales. These predictions are based on the traffic types and traffic scales determined in step S130. The traffic prediction model can be a statistical model based on historical data or a more complex machine learning model, which is used to predict the traffic conditions that nodes at each level may encounter within a specific time.

[0065] S150. Adjust the bandwidth allocation strategy and the hierarchical node structure distribution according to the traffic prediction type and the traffic prediction scale corresponding to each hierarchical node.

[0066] Exemplarily, the bandwidth allocation strategy and hierarchical node structure distribution are adjusted according to the traffic prediction type and traffic prediction scale obtained from step S140. This includes but is not limited to measures such as reconfiguring network bandwidth, optimizing routing strategies, and increasing or decreasing node resources. These adjustments are intended to optimize network performance, ensure that nodes at all levels can operate efficiently during peak traffic periods, and save resources when traffic is low.

[0067] In summary, traditional traffic management methods mostly rely on static rules and simple thresholds to respond to traffic changes, which is often not responsive enough when dealing with sudden or unconventional traffic. This application uses signal feature analysis and traffic prediction models to predict traffic trends and peak hours in advance, thereby achieving more dynamic and proactive bandwidth adjustment and network optimization. This predictive management method enables the network to make appropriate resource allocation before the traffic reaches its peak, avoiding potential congestion and service quality degradation. This application uses complex signal features such as periodic features, mutation features, noise features, dominant frequency features, and energy distribution features to analyze traffic. This in-depth feature analysis provides more sophisticated traffic insights than traditional methods, allowing network administrators to understand the essential characteristics and behavior patterns of traffic. For example, by analyzing the dominant frequency and energy distribution features, important real-time communication or large data transmission tasks can be more accurately identified and prioritized. The traffic classification process implemented in this application is not only based on the quantity of traffic, but also combines the quality and type of traffic, and further estimates the future traffic scale through the traffic prediction model. This allows the bandwidth allocation strategy to be more refined and personalized, and differentiated management of traffic demands of different types and scales is performed, thereby optimizing resource utilization efficiency and improving user experience. According to the traffic prediction results, the present application can flexibly adjust the bandwidth allocation strategy and hierarchical node structure distribution. This flexibility allows the network to increase bandwidth support or reconfigure the network structure when necessary to cope with the predicted traffic growth, while saving resources and reducing unnecessary cost expenditures when the traffic is low. By implementing the method of the present application, the overall performance and stability of the network are improved, and network management is more intelligent and automated, which can effectively cope with the changing demands in modern complex network environments. Compared with traditional technologies, this traffic management strategy based on deep analysis and prediction significantly improves the efficiency of network resource utilization and the quality of network services, especially in high-demand scenarios such as big data and cloud computing.

[0068] In some examples, the above-mentioned collection of traffic data corresponding to nodes at different levels in the target area includes:

[0069] Obtain the core router traffic data corresponding to the above core routers in the target area through the NetFlow / SFlow protocol;

[0070] Obtain the distribution switch traffic data corresponding to the above distribution switches in the target area through RSPAN;

[0071] The DHCP server log and Wi-Fi analysis tools are used to obtain the user equipment terminal traffic data corresponding to the above user equipment terminals in the target area.

[0072] For example, for core routers, NetFlow or SFlow protocols are used to obtain traffic data. These protocols can provide important information required by network administrators, such as traffic source, destination, transmission protocol, and various traffic statistics. NetFlow / SFlow is a commonly used technology in network analysis, which can help network administrators understand the data flow in the network, such as the size, direction, and change trend of traffic.

[0073] At the distribution switch level, RSPAN (Remote Switched Port Analyzer) technology is used to monitor the traffic of switch ports. RSPAN allows network administrators to remotely monitor the traffic of multiple switch ports from anywhere in the network, which is essential for analyzing and optimizing data flows. With RSPAN, the nature of the traffic passing through a specific switch can be understood in detail, including the type and quantity of various data packets, as well as the timing information of the traffic.

[0074] For user device terminals, traffic data collection relies on DHCP server logs and Wi-Fi analysis tools. DHCP server logs provide detailed information about device IP address allocation and leases, which is very useful for tracking device activities and traffic patterns within the network. At the same time, Wi-Fi analysis tools can be used to monitor wireless network usage, including signal strength, connection quality, and the amount of data transmitted over the Wi-Fi network.

[0075] In some examples, the signal feature analysis operation is performed on the traffic data of each level node to obtain traffic feature information corresponding to each level node, including:

[0076] Performing a signal characteristic analysis operation based on the core router traffic data to obtain the period characteristic, dominant frequency characteristic and energy distribution characteristic of the core router;

[0077] Performing a signal feature analysis operation based on the above distribution switch traffic data to obtain the mutation feature and noise feature of the above distribution exchange machine;

[0078] A signal feature analysis operation is performed based on the terminal traffic data of the above-mentioned user equipment to obtain the periodic characteristics, dominant frequency characteristics, noise characteristics and mutation characteristics of the above-mentioned user terminal equipment, wherein the above-mentioned periodic characteristics and dominant frequency characteristics are determined by Fourier transform analysis, the above-mentioned energy distribution is determined by power spectral density analysis, the above-mentioned mutation characteristics are determined by CUSUM algorithm, and the above-mentioned noise characteristics are determined based on a statistical model of normal traffic behavior.

[0079] Exemplary, traffic data analysis of core routers: The traffic data of core routers is subjected to signal feature analysis operations through Fourier transform analysis to determine the periodic changes of data streams. The periodic feature reflects the repetitive pattern of traffic data, such as traffic fluctuations on daily or weekend days. The traffic data of core routers is also subjected to signal feature analysis operations through Fourier transform analysis to determine the dominant frequency features of the data stream. The dominant frequency features reveal the most significant frequency components in the traffic data, which helps to identify the main activities in the traffic or the most common types of data transmission. The signal feature analysis operation is performed through power spectral density analysis to determine the energy distribution of the traffic data. This feature describes the energy distribution of the traffic data at each frequency, which helps to evaluate the intensity and dispersion of the traffic.

[0080] Traffic data analysis of distribution switches: Use the CUSUM algorithm (Cumulative Sum Control Chart algorithm) to perform signal feature analysis operations to obtain mutation features in the traffic data of distribution switches. Mutation features can indicate abnormal changes in traffic, such as a sharp increase or decrease in traffic, which is critical for timely detection of network problems. Perform signal feature analysis operations based on statistical models of normal traffic behavior to determine noise features. This feature helps identify random fluctuations and unexpected changes in traffic data, providing more information for traffic management.

[0081] Traffic data analysis of user equipment terminals: Also use the Fourier transform analysis method to perform signal feature analysis operations to obtain periodic features in the traffic data of user equipment terminals to help identify regular patterns in device use. Use the Fourier transform analysis method to perform signal feature analysis operations to analyze the dominant frequency characteristics of user equipment terminal traffic. Use a statistical model based on normal traffic behavior to perform signal feature analysis operations to identify noise characteristics. Use the CUSUM algorithm to perform signal feature analysis operations to analyze the traffic data of user equipment terminals and detect mutation features.

[0082] Through the above analysis methods, the system can comprehensively consider the traffic characteristics of nodes at each level and effectively classify and manage traffic. These analysis results provide scientific basis and support for subsequent traffic prediction, bandwidth allocation strategy adjustment and optimization of hierarchical node structure distribution.

[0083] In some examples, traffic classification is performed according to the traffic characteristic information corresponding to the core router to determine the traffic type and the traffic scale corresponding to the traffic type, including:

[0084] Inputting the periodic characteristics, dominant frequency characteristics and energy distribution characteristics of the core router into the first K-means clustering module to obtain a preliminary classification cluster of the first traffic data;

[0085] All the first traffic data preliminary classification clusters are input into the support vector machine module to obtain the traffic type corresponding to the core router and the traffic scale corresponding to the traffic type.

[0086] Exemplarily, the periodic features, dominant frequency features and energy distribution features extracted from the traffic data of the core router are used as inputs. These features represent the key signal attributes of the traffic data and can reflect the behavior and pattern of the traffic.

[0087] The first K-means clustering module is used to process these input features. K-means clustering is an unsupervised learning method that groups data points into a predetermined number of clusters so that points within the same cluster have high similarity to each other and low similarity to points in other clusters. Through the K-means clustering algorithm, the data is divided into several clusters, each cluster representing a group of flow data with similar periodic characteristics, dominant frequency, and energy distribution. These clusters form a preliminary classification of the flow data and provide a basis for further analysis.

[0088] All preliminary classification clusters obtained from the first K-means clustering module are passed as input data to the support vector machine (SVM) module. The SVM module analyzes the input classification clusters and performs a final classification based on the characteristics of the traffic, determining each type of traffic and its corresponding traffic scale. In this way, different types of traffic handled by the core router, such as video streaming, data downloading, or real-time communication, can be accurately distinguished and identified, and the scale of these types of traffic can be estimated.

[0089] Based on the signal feature analysis of the traffic data of the core router, traffic classification is performed through advanced data processing modules (K-means clustering and support vector machine), which not only improves the efficiency and accuracy of traffic management, but also enables network managers to better understand and predict network load, thereby optimizing the configuration and use of network resources. The application of this method helps to improve network performance and user experience while reducing operating costs.

[0090] In some examples, traffic classification is performed according to the traffic characteristic information corresponding to the distribution switch to determine the traffic type and the traffic scale corresponding to the traffic type, including:

[0091] The mutation characteristics and noise characteristics of the above allocation interaction machine are input into the LSTM module to obtain the time series change characteristics;

[0092] The above time series change characteristics are input into the CNN module to obtain the traffic type corresponding to the above distribution switch and the traffic scale corresponding to the above traffic type.

[0093] Exemplarily, the mutation features and noise features extracted from the traffic data of the distribution switch are input into the LSTM (Long Short-Term Memory Network) module. These features reflect the dynamic changes and irregular fluctuations of traffic data in time series, and are suitable for capturing long-term dependencies and short-term fluctuations of time series data. The LSTM module analyzes the mutation and noise features of the input, and captures the dependencies and patterns in the time series through its network structure, thereby effectively obtaining the time series change characteristics. LSTM makes predictions by memorizing past information, which is very suitable for time series data analysis and can handle and identify long-term dependencies and complex time series prediction tasks.

[0094] The time series change features obtained after processing by the LSTM module are input into the CNN (Convolutional Neural Network) module. The powerful spatial hierarchical feature extraction capability of CNN enables it to further analyze and identify complex patterns in traffic data. The CNN module extracts deep features of time series data through multi-layer convolution and pooling operations, and maps these features to the final classification results through the fully connected layer. This step allows the system to identify and classify the traffic types corresponding to the distribution switches and estimate the scale of each type of traffic. In this way, the main components of network traffic and their scale can be accurately determined, providing a basis for the management and optimization of network resources.

[0095] By combining LSTM and CNN, two advanced deep learning technologies, the traffic data of the distribution switch is deeply analyzed and accurately classified. This method can not only effectively utilize the advantages of LSTM in processing time series data, but also take advantage of the powerful ability of CNN in feature extraction, thereby improving the accuracy of traffic classification and the efficiency of network management.

[0096] In some examples, traffic classification is performed according to the traffic characteristic information corresponding to the user equipment terminal to determine the traffic type and the traffic scale corresponding to the traffic type, including:

[0097] Inputting the periodic characteristics, dominant frequency characteristics, noise characteristics and mutation characteristics of the user terminal device into the second K-means clustering module to obtain a preliminary classification cluster of the second traffic data;

[0098] All the second traffic data preliminary classification clusters are input into the random forest module to obtain the traffic type corresponding to the above user equipment terminal and the traffic scale corresponding to the above traffic type.

[0099] Exemplarily, the periodic features, dominant frequency features, noise features and mutation features of the user equipment terminal are collected. These features are key signal attributes of time series data, describing the traffic behavior and pattern of the user equipment terminal. These traffic features are input into the second K-means clustering module. Through the clustering algorithm, the traffic data of the user equipment terminal is grouped into different clusters according to its features. Each cluster represents a group of traffic data with similar features, forming a preliminary classification cluster.

[0100] All preliminary classification clusters obtained from the second K-means clustering module are used as data input to the random forest module. The random forest module analyzes these classification clusters and uses a collection of decision trees to predict the type of traffic and estimate its size. This process enables the system to determine the type of traffic corresponding to the user device terminal, such as video streaming, social media interaction, file downloading, etc., as well as the size of each type of traffic with high accuracy.

[0101] This method can effectively classify the traffic data of user equipment terminals and accurately determine the type and scale of traffic. First, the second K-means clustering module is used for preliminary classification, and then the random forest module is used to further refine the classification results. This hierarchical processing method not only improves the accuracy of classification, but also enhances the efficiency of traffic management, making the allocation and optimization of network resources more targeted and effective.

[0102] In some examples, the specific steps of adjusting the bandwidth allocation strategy according to the traffic prediction type and the traffic prediction scale corresponding to each level node include:

[0103] Determine the mainstream application and the peak time period corresponding to the mainstream application based on the above traffic prediction type and the above traffic prediction scale corresponding to all hierarchical nodes;

[0104] Get the number of nodes under each level node;

[0105] Determine the node weight coefficient corresponding to each level node based on the number of the above nodes under each level node and the level to which each level node belongs;

[0106] Based on the node weight coefficient corresponding to each level node, the mainstream applications, and the peak time periods corresponding to the mainstream applications, the bandwidth allocation strategy of each level node in different time periods is adjusted.

[0107] For example, based on the traffic prediction types and traffic prediction scales corresponding to all level nodes, the system first analyzes and identifies the mainstream applications in the network. These are the applications that dominate the network traffic, such as video streaming, large data transmission, online games, etc. At the same time, based on the traffic prediction data, the peak hours of these mainstream applications are determined, that is, the time interval with the largest traffic consumption.

[0108] Get the number of nodes under each level node, including all subordinate nodes at the core router, distribution switch, and user equipment terminal levels.

[0109] Based on the number of subordinate nodes and the level to which the hierarchical node belongs, the system calculates a weight coefficient for each node. This weight coefficient reflects the importance of the node in the network and its responsibility for traffic processing. The calculation of the node weight coefficient can be based on multiple factors, such as the node's traffic load, the number of connected devices, and the number of users served.

[0110] Based on the above node weight coefficients, mainstream applications, and peak hours corresponding to mainstream applications, the bandwidth allocation strategy is adjusted. This strategy includes providing more bandwidth resources to key nodes during peak hours to support the smooth operation of mainstream applications, while appropriately reducing the bandwidth of these nodes during non-peak hours to optimize resource utilization and cost-effectiveness.

[0111] Through these steps, the network can dynamically adjust its resource allocation based on actual traffic demand and forecast models, thereby ensuring the performance of key applications while maximizing the overall efficiency of resources and the stability of the network. This approach not only improves user experience, but also improves the intelligence and adaptability of network management.

[0112] In some examples, the specific steps of adjusting the hierarchical node structure distribution according to the above traffic prediction type and the above traffic prediction scale corresponding to each hierarchical node include:

[0113] Analyze the historical performance parameters of all nodes at all levels under different traffic types, where the historical performance parameters include processing power, storage capacity, and transmission rate;

[0114] Perform node performance evaluation according to the traffic prediction type, the traffic prediction scale and the historical performance parameters to obtain a performance evaluation result;

[0115] Node optimization and topology optimization are carried out based on the above performance evaluation results.

[0116] Exemplarily, the performance parameters of all hierarchical nodes (such as core routers, distribution switches, and user equipment terminals) in the past when processing different types of traffic are collected and analyzed. These historical performance parameters include processing power, storage capacity, and transmission rate. This analysis helps identify the efficiency and ability of each node in processing specific traffic types.

[0117] Based on the traffic forecast type and traffic forecast scale, combined with historical performance parameters, the performance of nodes at each level is evaluated. This evaluation considers the degree of match between future traffic demand and the current and historical performance of the node to determine whether there are potential performance bottlenecks or resource surpluses.

[0118] Based on the performance evaluation results, the nodes are optimized as necessary. This may include measures such as enhancing processing power, expanding storage capacity, or increasing transmission rates to ensure that the nodes can effectively handle the predicted traffic volume. The physical structure or logical topology of the network is further optimized based on the performance evaluation results. This may involve reconfiguring network connections, adjusting routing strategies, or restructuring the network architecture to improve the efficiency and performance of the entire network.

[0119] Suppose historical data shows that the core router reaches its transmission rate limit when handling a large amount of video traffic during peak hours on weekdays. The traffic forecast model predicts that this will happen more frequently in the future. The assessment may show that the core router's processing capacity during peak hours is insufficient to handle future traffic growth. The assessment will consider processing power, storage capacity, and current transmission rates. Possible optimization measures include upgrading the router's hardware, adding more storage resources, or introducing more efficient data processing algorithms. In the case of frequent congestion in the core router, it may be necessary to add new routers or redesign the network architecture, such as introducing more distribution switches to spread the traffic load.

[0120] Through these steps, network managers can ensure that each part of the network can effectively cope with the predicted traffic demand, while improving the overall performance and reliability of the network. This optimization strategy based on detailed performance analysis and system helps to achieve the optimal configuration and utilization of resources.

[0121] In some examples, node optimization and topology optimization are performed based on the above performance evaluation results, including:

[0122] Determine the bottleneck node based on the above performance evaluation results;

[0123] Get the upgrade cost information of bottleneck nodes;

[0124] Determine a first optimized path using the shortest path algorithm according to the bottleneck node;

[0125] Determine a second optimization path using a minimum spanning tree algorithm based on the bottleneck node;

[0126] Determine the upgrade cost corresponding to the first optimization path and the upgrade cost corresponding to the second optimization path according to the upgrade cost information;

[0127] Node optimization and topology optimization are performed based on the principle of minimizing upgrade costs.

[0128] For example, the performance evaluation results of nodes at each level are analyzed to identify those nodes that cannot meet the scheduled traffic demand or whose performance is significantly degraded during peak hours. These nodes are identified as bottleneck nodes and are the primary targets for optimization.

[0129] Gather cost information for any upgrades needed for each identified bottleneck node. This includes the cost of hardware upgrades, software upgrades, or any necessary architectural adjustments.

[0130] Determine the first optimized path using the shortest path algorithm. Apply a shortest path algorithm (such as Dijkstra's algorithm) to identify the most efficient path to optimize these bottleneck nodes. This approach aims to find the shortest path from the source point to the destination point (bottleneck node) to minimize delay and cost.

[0131] The second optimal path is determined using a minimum spanning tree algorithm, which is used to identify a cost network structure covering all nodes.

[0132] Determine the upgrade costs corresponding to the first optimization path and the second optimization path according to the upgrade cost information: compare the costs of the optimization solutions obtained using the shortest path algorithm and the minimum spanning tree algorithm, and evaluate the economic benefits of the two.

[0133] Node optimization and topology optimization based on the principle of minimizing upgrade costs: Finally, based on the cost-benefit analysis, the optimization solution with the lowest cost is selected for implementation. This may involve selecting the path provided by the shortest path algorithm, or adopting the suggestions of the minimum spanning tree algorithm, or some combination of the two, to ensure that bottleneck problems are solved while minimizing costs, and to optimize the performance and stability of the entire network.

[0134] like Figure 2 As shown, the present application proposes a traffic management system based on signal feature analysis, including:

[0135] A collecting unit 21 is used to collect the corresponding traffic data of nodes at different levels in the target area, wherein the nodes at different levels include core routers, distribution switches and user equipment terminals;

[0136] The analysis unit 22 is used to perform a signal characteristic analysis operation on the flow data of each level node to obtain flow characteristic information corresponding to each level node;

[0137] A classification unit 23, configured to classify traffic according to the traffic feature information corresponding to each level node to determine the traffic type and the traffic scale corresponding to the traffic type;

[0138] A prediction unit 24, configured to obtain a traffic prediction type and a traffic prediction scale corresponding to the traffic prediction type through a traffic prediction model according to the traffic type and the traffic scale corresponding to each level node;

[0139] The adjusting unit 25 is used to adjust the bandwidth allocation strategy and the hierarchical node structure distribution according to the above-mentioned traffic prediction type and the above-mentioned traffic prediction scale corresponding to each hierarchical node.

[0140] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0141] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0142] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0143] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0144] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0145] An embodiment of the present application also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device executes the process of predicting the decomposition rate in alumina production in the corresponding embodiment.

[0146] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can store or a data storage device such as a server or a data center that includes one or more available media integration. The available medium can be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state drive (SSD)), etc.

[0147] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0148] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0149] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0150] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0151] 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, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.

[0152] 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 traffic management method based on signal feature analysis, comprising: Collecting traffic data corresponding to nodes at different levels in the target area, wherein the nodes at different levels include core routers, distribution switches, and user equipment terminals; Perform signal feature analysis on the traffic data of each level node to obtain the traffic feature information corresponding to each level node; Classify the traffic according to the traffic characteristic information corresponding to each level node to determine the traffic type and the traffic scale corresponding to the traffic type; According to the traffic type and the traffic scale corresponding to each level node, a traffic prediction model is used to obtain a traffic prediction type and a traffic prediction scale corresponding to the traffic prediction type; Adjust the bandwidth allocation strategy and the hierarchical node structure distribution according to the traffic prediction type and the traffic prediction scale corresponding to each hierarchical node; The signal feature analysis operation is performed on the traffic data of each level node to obtain the traffic feature information corresponding to each level node, including: Performing a signal characteristic analysis operation based on the core router traffic data to obtain a period characteristic, a dominant frequency characteristic, and an energy distribution characteristic of the core router; Performing a signal feature analysis operation based on the distribution switch traffic data to obtain the mutation feature and noise feature of the distribution exchange machine; A signal feature analysis operation is performed based on the terminal traffic data of the user equipment to obtain the periodic features, dominant frequency features, noise features and mutation features of the user terminal equipment, wherein the periodic features and dominant frequency features are determined by Fourier transform analysis, the energy distribution is determined by power spectral density analysis, the mutation features are determined by CUSUM algorithm, and the noise features are determined based on a statistical model of normal traffic behavior.

2. The traffic management method based on signal feature analysis according to claim 1 is characterized in that: The collection of traffic data corresponding to nodes at different levels in the target area includes: Obtain the core router traffic data corresponding to the core router in the target area through the NetFlow / SFlow protocol; Acquire the distribution switch flow data corresponding to the distribution switch in the target area through RSPAN; The user equipment terminal traffic data corresponding to the user equipment terminal in the target area is obtained through the DHCP server log and the Wi-Fi analysis tool.

3. The traffic management method based on signal feature analysis according to claim 1 is characterized in that: Classifying traffic according to the traffic feature information corresponding to the core router to determine the traffic type and the traffic scale corresponding to the traffic type includes: Inputting the periodic characteristics, dominant frequency characteristics and energy distribution characteristics of the core router into a first K-means clustering module to obtain a preliminary classification cluster of the first traffic data; All the first traffic data preliminary classification clusters are input into the support vector machine module to obtain the traffic type corresponding to the core router and the traffic scale corresponding to the traffic type.

4. The traffic management method based on signal feature analysis according to claim 1 is characterized in that: Classifying traffic according to the traffic characteristic information corresponding to the distribution switch to determine the traffic type and the traffic scale corresponding to the traffic type includes: Inputting the mutation characteristics and noise characteristics of the allocation interaction machine into the LSTM module to obtain the time series change characteristics; The time series variation characteristics are input into a CNN module to obtain the traffic type corresponding to the distribution switch and the traffic scale corresponding to the traffic type.

5. The traffic management method based on signal feature analysis according to claim 1 is characterized in that: Classifying traffic according to the traffic characteristic information corresponding to the user equipment terminal to determine a traffic type and a traffic scale corresponding to the traffic type includes: Inputting the periodic characteristics, dominant frequency characteristics, noise characteristics and mutation characteristics of the user terminal device into a second K-means clustering module to obtain a preliminary classification cluster of the second traffic data; All the second traffic data preliminary classification clusters are input into the random forest module to obtain the traffic type corresponding to the user equipment terminal and the traffic scale corresponding to the traffic type.

6. The traffic management method based on signal feature analysis according to claim 1 is characterized in that: The specific steps of adjusting the bandwidth allocation strategy according to the traffic prediction type and the traffic prediction scale corresponding to each level node include: Determine the mainstream application and the peak time period corresponding to the mainstream application based on the traffic prediction type and the traffic prediction scale corresponding to all hierarchical nodes; Get the number of nodes under each level node; Determine a node weight coefficient corresponding to each level node based on the number of nodes subordinate to each level node and the level to which each level node belongs; The bandwidth allocation strategy of each level node in different time periods is adjusted based on the node weight coefficient corresponding to each level node, the mainstream application, and the peak time period corresponding to the mainstream application.

7. The traffic management method based on signal feature analysis according to claim 1 is characterized in that: The specific steps of adjusting the hierarchical node structure distribution according to the traffic prediction type and the traffic prediction scale corresponding to each hierarchical node include: Analyze the historical performance parameters of all nodes at all levels under different traffic types, where the historical performance parameters include processing power, storage capacity, and transmission rate; Perform node performance evaluation according to the traffic prediction type, the traffic prediction scale and the historical performance parameters to obtain a performance evaluation result; Node optimization and topology optimization are performed based on the performance evaluation results.

8. The traffic management method based on signal feature analysis according to claim 7 is characterized in that: Node optimization and topology optimization are performed based on the performance evaluation results, including: Determine the bottleneck node based on the performance evaluation result; Get the upgrade cost information of bottleneck nodes; Determine a first optimized path using a shortest path algorithm according to the bottleneck node; Determine a second optimization path using a minimum spanning tree algorithm according to the bottleneck node; Determine, according to the upgrade cost information, an upgrade cost corresponding to the first optimization path and an upgrade cost corresponding to the second optimization path; Node optimization and topology optimization are performed based on the principle of minimizing upgrade costs.

9. A traffic management system based on signal feature analysis, characterized in that: include: A collecting unit, configured to collect traffic data corresponding to nodes at different levels in a target area, wherein the nodes at different levels include core routers, distribution switches, and user equipment terminals; An analysis unit, used to perform a signal characteristic analysis operation on the flow data of each level node to obtain flow characteristic information corresponding to each level node; A classification unit, used to classify traffic according to the traffic feature information corresponding to each level node to determine the traffic type and the traffic scale corresponding to the traffic type; A prediction unit, configured to obtain a traffic prediction type and a traffic prediction scale corresponding to the traffic prediction type through a traffic prediction model according to the traffic type and the traffic scale corresponding to each level node; An adjustment unit, configured to adjust a bandwidth allocation strategy and a hierarchical node structure distribution according to the traffic prediction type and the traffic prediction scale corresponding to each hierarchical node; The signal feature analysis operation is performed on the traffic data of each level node to obtain the traffic feature information corresponding to each level node, including: Performing a signal characteristic analysis operation based on the core router traffic data to obtain a period characteristic, a dominant frequency characteristic, and an energy distribution characteristic of the core router; Performing a signal feature analysis operation based on the distribution switch traffic data to obtain the mutation feature and noise feature of the distribution exchange machine; A signal feature analysis operation is performed based on the terminal traffic data of the user equipment to obtain the periodic features, dominant frequency features, noise features and mutation features of the user terminal equipment, wherein the periodic features and dominant frequency features are determined by Fourier transform analysis, the energy distribution is determined by power spectral density analysis, the mutation features are determined by CUSUM algorithm, and the noise features are determined based on a statistical model of normal traffic behavior.

Images