Communication network bandwidth control method

Abstract

The invention discloses a communication network bandwidth control method, and particularly relates to the technical field of network bandwidth control. Flow monitoring equipment is arranged in a communication network, flow is monitored in real time, flow data of different users and applications are collected, and on the basis of historical flow data and a real-time change trend, a burst flow prediction model is constructed, and flow peak fluctuation and sudden increase characteristics are analyzed, so that the real-time response capability of a bandwidth allocation strategy is evaluated in advance, and the real-time response capability of the bandwidth allocation strategy is evaluated in advance. When the response capability is insufficient, bandwidth allocation is dynamically adjusted according to a priority strategy, bandwidth requirements of high-transmission-priority applications are preferentially guaranteed, bandwidth occupation of low-priority applications is limited so as to ensure the continuity of key services, and after bandwidth adjustment, the allocation strategy is further refined through real-time flow feedback, so that the bandwidth allocation efficiency is improved. The efficient utilization of resources and the stability of network service quality are guaranteed, so that the problem of bandwidth allocation lag in emergencies is solved, and flexible resource optimization management is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of network bandwidth control, and in particular to a communication network bandwidth control method. Background Art

[0002] In modern communications networks, with the diversification of internet applications and the rapid growth of data demands, the limited availability of bandwidth has become a significant factor constraining network service quality. High-traffic applications (such as high-definition video, real-time audio streaming, and online gaming) consume significant amounts of network bandwidth, placing significant strain on communications networks. To address bandwidth shortages, bandwidth control technology has emerged. It rationally allocates and manages network traffic to ensure the stability of critical services, optimize user experience, and prevent excessive resource usage by a single user or application. Bandwidth control methods primarily rely on traffic monitoring, bandwidth allocation strategies, and prioritization mechanisms to ensure fair and efficient service delivery within the constraints of limited bandwidth resources.

[0003] Traditional bandwidth control technologies typically implement bandwidth caps and priority control. However, with the increasing complexity of networks and application diversity, these approaches are struggling to meet the dynamic and precise requirements of modern networks. For example, traditional static bandwidth limits often fail to adapt to dynamic changes in network load, resulting in some services being unable to access sufficient bandwidth during peak hours or facing overly strict limits during off-peak hours, impacting resource utilization efficiency.

[0004] The existing technology has the following shortcomings: In dynamic network environments, user traffic demands can surge, especially during unexpected network events such as breaking news or live events. Existing bandwidth control methods often rely on static bandwidth allocation strategies, which are unable to adapt quickly to sudden traffic surges. This can lead to temporary network congestion, interruptions to critical services, and even a decrease in overall network service quality. Furthermore, after a sudden traffic event subsides, the lag inherent in static bandwidth allocation strategies can lead to uneven bandwidth allocation and reduced resource utilization. For example, after a sudden demand event subsides, the network management system may fail to restore bandwidth allocation in a timely manner, resulting in inefficient utilization of previously limited resources and waste. Summary of the Invention

[0005] The purpose of the present invention is to provide a communication network bandwidth control method to solve the shortcomings of the background technology.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a communication network bandwidth control method, comprising the following steps: S1: Real-time traffic is monitored by traffic monitoring equipment installed in the communication network, and traffic data of different users and various applications is continuously collected. The traffic data includes traffic rate, traffic peak and transmission priority; S2: Based on collected historical traffic data and real-time traffic trends, a burst traffic prediction model is constructed. This model analyzes traffic peak fluctuation characteristics and sudden growth characteristics to evaluate the real-time response capability of static bandwidth allocation strategies to burst traffic. S3: When the static bandwidth allocation policy has a low real-time response capability to sudden traffic, bandwidth resources are dynamically allocated according to the preset priority policy, ensuring that high-priority applications receive guaranteed bandwidth. Bandwidth allocation for low-priority applications is adjusted to reduce their bandwidth usage and free up resources for high-priority applications. S4: Analyze the effect of bandwidth adjustment and further refine bandwidth allocation based on real-time traffic feedback to ensure network service quality and resource utilization efficiency.

[0007] Preferably, in S2, a flow peak fluctuation index is generated by analyzing the peak fluctuation characteristics of the flow, and the method for obtaining the flow peak fluctuation index is: Assume that the traffic peak value y follows the probability distribution ,in It is a parameter that describes the peak fluctuation characteristics. θ is a random variable with a prior distribution p(θ). It is assumed that θ obeys the normal distribution. ;in, and is the prior mean and variance; using the collected historical traffic peak data , x is the total number of historical traffic peak data, and the likelihood function is defined as , define each The observations of are independent and normally distributed, then: ; is the expectation of peak fluctuation, is the variance of traffic fluctuation; calculate the posterior distribution , making ; By minimizing q(θ) and the true posterior The KL divergence between , variational inference finds the optimal q(θ), which is expressed as: ; Equivalently, the maximum variational lower bound ELBO is defined as: ; After optimization, the optimal parameter mean of q(θ) is obtained and variance , use q(θ) to calculate the flow peak fluctuation index, the expression is: ; Where SQ is the flow peak fluctuation index.

[0008] Preferably, in S2, a traffic sudden increase anomaly index is generated by analyzing the sudden increase characteristics of the traffic, and the method for obtaining the traffic sudden increase anomaly index is: Collect traffic data within the M time period to form a time series ,in represents the flow value at time j, N is the total number of data points, and the time domain flow data is converted to the frequency domain using fast Fourier transform. The expression is: ;in is the kth frequency component in the frequency domain, Represents different frequency components, is the kernel function of Fourier transform, which calculates the spectrum amplitude of each frequency component. Represents the intensity of the signal at each frequency component, and the calculation expression is: ;in, and They are the real and imaginary parts of the frequency domain signal, respectively, and set the high frequency threshold , extract the sum of the amplitudes in the high frequency range , the expression is: ; Set the low frequency range threshold , calculate the sum of the amplitudes in the low-frequency range : The traffic surge anomaly index is defined using the ratio of high-frequency amplitude to low-frequency amplitude. The expression is: ; ED is the traffic sudden increase abnormality index.

[0009] Preferably, in S2, the traffic peak fluctuation index and the traffic surge abnormality index are converted into a first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the real-time response capability value label of the static bandwidth allocation strategy to burst traffic as the prediction target, and takes minimizing the sum of the prediction errors of the real-time response capability value labels of all static bandwidth allocation strategies to burst traffic as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The real-time response capability value of the static bandwidth allocation strategy to burst traffic is determined according to the model output result, wherein the machine learning model is a polynomial regression model.

[0010] Preferably, in S2, the obtained real-time response capability value of the static bandwidth allocation strategy to burst traffic is compared with the real-time response capability benchmark value of the static bandwidth allocation strategy to burst traffic under the standard state in the historical data. If the real-time response capability value of the static bandwidth allocation strategy to burst traffic is greater than the real-time response capability benchmark value of the static bandwidth allocation strategy to burst traffic, it means that the real-time response capability of the static bandwidth allocation strategy to burst traffic is high, and a high real-time response capability signal is generated at this time; if the real-time response capability value of the static bandwidth allocation strategy to burst traffic is less than or equal to the real-time response capability benchmark value of the static bandwidth allocation strategy to burst traffic, it means that the real-time response capability of the static bandwidth allocation strategy to burst traffic is low, and a low real-time response capability signal is generated at this time.

[0011] Preferably, in S3, when the static bandwidth allocation strategy has low real-time response capability to sudden traffic, bandwidth resources are dynamically allocated according to a preset priority strategy, specifically: Assume that the total bandwidth requirement of the current network is , driven by the needs of high-priority applications and the needs of low-priority applications constitute, then: Under burst traffic conditions, priority is given to bandwidth requirements of high-priority applications. A dynamic bandwidth allocation factor α is defined to dynamically adjust bandwidth allocation for high- and low-priority applications based on their priorities. The bandwidth allocation formula is: ; is the adjusted high-priority application bandwidth; ; is the adjusted bandwidth for low-priority applications.

[0012] Preferably, if the bandwidth requirements of high priority applications Greater than or equal to the allocated guaranteed bandwidth , then it is necessary to release bandwidth from low-priority applications, and the amount of released bandwidth ΔB is defined as: ; Reduce the released bandwidth ΔB from the low-priority application bandwidth to ensure that the needs of high-priority applications are met; the bandwidth of low-priority applications after adjustment is: Where, It is the bandwidth that is ultimately allocated to low-priority applications, and the remaining resources are allocated while ensuring the needs of high-priority applications.

[0013] Preferably, in S4, the effect of bandwidth adjustment is analyzed, and bandwidth allocation is further refined based on real-time traffic feedback, specifically: After bandwidth adjustment, the adjusted traffic data is collected in real time through the monitoring system to form a time series ,in Represents the traffic usage value at the rth time point. Set the size of the sliding window W. For each time point r, calculate the standard deviation within the window W. As new traffic data arrives, the sliding window advances and the standard deviation within the new window is repeatedly calculated; Set a standard deviation threshold to detect abnormal traffic fluctuations. If the real-time standard deviation is greater than the standard deviation threshold, bandwidth allocation needs to be adjusted. In this case, bandwidth for high-priority applications is increased to prioritize their bandwidth needs. Bandwidth for low-priority applications is restricted to free up more bandwidth for high-priority applications. If the real-time standard deviation is less than or equal to the standard deviation threshold, traffic fluctuations are within a controllable range and the current bandwidth allocation remains unchanged.

[0014] In the above technical solution, the technical effects and advantages provided by the present invention are: 1. By incorporating traffic monitoring equipment, this solution can collect real-time traffic data from various users and build a burst traffic prediction model based on historical data to evaluate the responsiveness of bandwidth allocation strategies. If insufficient bandwidth allocation response is detected, the system dynamically adjusts bandwidth allocation based on priority, prioritizing high-priority applications and reducing bandwidth usage by low-priority applications, ensuring the stable operation of critical services under burst traffic.

[0015] 2. This invention leverages real-time traffic feedback and analyzes bandwidth adjustment effects using sliding window standard deviation. Based on this feedback, it further optimizes bandwidth allocation strategies to achieve refined responses and dynamic management of traffic fluctuations. Compared to traditional methods, this invention not only improves the adaptability of bandwidth allocation and resource utilization efficiency, avoiding resource waste after sudden demand subsides, but also maintains the stability and flexibility of network service quality. It also significantly enhances the network system's ability to respond to sudden traffic events and optimizes resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0019] For examples, see Figure 1 As shown, the communication network bandwidth control method described in this embodiment includes the following steps: S1: Real-time traffic is monitored by traffic monitoring equipment installed in the communication network, and traffic data of different users and various applications is continuously collected. The traffic data includes traffic rate, traffic peak and transmission priority; S2: Based on collected historical traffic data and real-time traffic trends, a burst traffic prediction model is constructed. This model analyzes traffic peak fluctuation characteristics and sudden growth characteristics to evaluate the real-time response capability of static bandwidth allocation strategies to burst traffic. S3: When the static bandwidth allocation policy has a low real-time response capability to sudden traffic, bandwidth resources are dynamically allocated according to the preset priority policy, ensuring that high-priority applications receive guaranteed bandwidth. Bandwidth allocation for low-priority applications is adjusted to reduce their bandwidth usage and free up resources for high-priority applications. S4: Analyze the effect of bandwidth adjustment and further refine bandwidth allocation based on real-time traffic feedback to ensure network service quality and resource utilization efficiency.

[0020] In S1, real-time traffic is monitored by traffic monitoring equipment installed in the communication network, and traffic data of different users and various applications are continuously collected. The traffic data includes traffic rate, traffic peak and transmission priority, specifically: Deploy traffic monitoring devices, including deep packet inspection (DPI) and network traffic analyzers, at network entry points or key nodes (such as routers and switches). These devices capture traffic packets passing through these nodes and perform fine-grained traffic analysis. Configure the monitoring devices in real-time capture and analysis mode. The traffic monitoring devices record each passing packet in real time and extract essential traffic information (such as source IP address, destination IP address, port, and protocol type) to classify traffic. Due to the large volume of data, pre-processing steps filter out unnecessary or noisy data, retaining only critical packets. Based on pre-defined rules (such as application type, user group, and specific IP address segments), only relevant traffic types are analyzed, reducing the burden on the devices. Data packets are categorized by user and application, and real-time statistics are compiled for traffic rate, peak value, and average traffic flow. Categorization can be based on application protocol (such as HTTP, HTTPS, FTP), user IP address, and traffic type. This statistical information is used for subsequent bandwidth allocation and burst traffic prediction.

[0021] A user and application identification table is established within the system, assigning each user and application a unique identifier (such as an IP address or MAC address). The table also records each application's protocol type and user identity (e.g., regular user or VIP user) to facilitate differentiation during data collection. Using the real-time data collection capabilities of traffic monitoring devices, specific fields are extracted from each packet, including source / destination IP addresses, port numbers, application protocol type, and packet size. Traffic rate and peak traffic volume can be calculated by analyzing packet size and transmission time over a period of time. Collected packets are aggregated at regular intervals, combining multiple packets into traffic statistics, such as per-second traffic rate and per-minute peak traffic volume. Aggregated data is stored in a database to support subsequent bandwidth analysis and adjustment. Traffic transmission priority is determined based on application protocol type (e.g., video streaming, file transfer, web browsing) and user identity (e.g., VIP user, regular user).

[0022] Priority policies can be categorized as high priority or low priority. High priority is typically used for applications with high bandwidth and latency requirements, such as video conferencing, real-time communications, and financial transactions, as well as for specific users (such as VIP users). Low priority is used for applications with low bandwidth sensitivity or high fault tolerance, such as file downloads and data backups. Specifically:

[0023] Prioritize applications based on protocol type and configure rules, using protocol type as one of the criteria for prioritization: High-priority protocols include real-time video streaming (such as RTP), financial transaction protocols (such as FIX), and voice communications (such as VoIP). Low-priority protocols include file transfer protocols (such as FTP), email transfer protocols (such as SMTP), and HTTP traffic (normal web browsing).

[0024] Prioritize traffic based on user identity and configure priority policies. Another criterion for prioritizing traffic based on user identity is: High-priority users: Traffic from VIP or enterprise users can be automatically prioritized to ensure they meet bandwidth requirements in bursty traffic environments. Low-priority users: Traffic from ordinary users is prioritized to a low priority by default to reduce overall bandwidth usage.

[0025] S2: Based on the collected historical traffic data and real-time traffic change trends, a burst traffic prediction model is constructed. The model analyzes the peak fluctuation characteristics and sudden growth characteristics of traffic to evaluate the real-time response capability of the static bandwidth allocation strategy to burst traffic.

[0026] The purpose of building a traffic burst prediction model is to identify potential traffic bursts in advance by analyzing peak fluctuations and sudden growth characteristics in historical traffic data. This allows us to evaluate the effectiveness and adaptability of existing static bandwidth allocation strategies in dealing with traffic bursts. This provides real-time early warning signals to network systems, enabling them to adjust bandwidth allocation strategies before traffic surges occur. This ensures that bandwidth requirements for high-priority applications and critical services are prioritized, improving service quality and resource utilization efficiency across the entire network.

[0027] The peak flow fluctuation index is generated by analyzing the peak flow fluctuation characteristics. The method for obtaining the peak flow fluctuation index is as follows: Assume that the traffic peak value y follows the probability distribution ,in It is a parameter that describes the peak fluctuation characteristics (such as mean and standard deviation). θ is a random variable with its own prior distribution p(θ). It is assumed that θ obeys the normal distribution. ;in, and is the prior mean and variance; using the collected historical traffic peak data , x is the total number of historical traffic peak data, and the likelihood function is defined as To reflect the possibility of observation data under given parameters, define each The observations of are independent and normally distributed, then: ; is the expectation of peak fluctuation, is the variance of traffic fluctuations; Bayesian inference requires calculating the posterior distribution However, since direct solution is often not feasible, variational inference is used to obtain an approximate distribution that is easy to handle. , making ; By minimizing q(θ) and the true posterior The Kullback-Leibler (KL) divergence between , variational inference finds the optimal q(θ), which is expressed as: ; Equivalently, maximize the variational lower bound ELBO, defined as: ; After optimization, the optimal parameter mean of q(θ) is obtained and variance Using q(θ) to calculate the peak flow fluctuation index, the fluctuation intensity can usually be expressed based on the variance of θ, as follows: ; Where SQ is the flow peak fluctuation index.

[0028] A high traffic peak fluctuation index indicates high network traffic volatility, with frequent or significant peak fluctuations. In this case, static bandwidth allocation policies have limited real-time responsiveness, as fixed bandwidth allocations cannot quickly adapt to sudden surges in traffic. This can easily lead to insufficient bandwidth during peak hours, causing network congestion or latency, and impacting the normal operation of critical services. In these situations, the system requires more flexible bandwidth management mechanisms, such as dynamic bandwidth adjustment or increased traffic forecasting, to address the shortcomings of static policies in highly volatile environments.

[0029] Conversely, a low traffic peak fluctuation index indicates low network traffic volatility and relatively stable traffic peaks. In this case, static bandwidth allocation policies offer better real-time responsiveness, as fixed bandwidth allocations are sufficient to handle stable traffic fluctuations, minimizing bandwidth shortages and resource waste. Because traffic demand is relatively stable, static policies can effectively meet demand in low-volatility environments without requiring frequent adjustments, thus reducing management complexity and resource costs.

[0030] The traffic sudden increase anomaly index is generated by analyzing the sudden increase characteristics of traffic. The method for obtaining the traffic sudden increase anomaly index is as follows: Collect traffic data within the M time period to form a time series ,in represents the flow value at time j, N is the total number of data points, and the time domain flow data is converted to the frequency domain using fast Fourier transform. The expression is: ;in is the kth frequency component in the frequency domain, Represents different frequency components, is the kernel function of Fourier transform, which calculates the spectrum amplitude of each frequency component. Represents the intensity of the signal at each frequency component, and the calculation expression is: ;in, and The real and imaginary parts of the frequency domain signal respectively. A sudden increase in traffic is usually reflected as a significant increase in high-frequency components in the spectrum. Set the high frequency threshold (e.g. half of the total frequency), extract the sum of the amplitudes in the high frequency range , the expression is: ; Regular fluctuations in flow are usually in the low frequency range, so set the low frequency range threshold (e.g. the first 25% of the total frequency), calculate the sum of the amplitudes in the low frequency range : The traffic surge anomaly index is defined using the ratio of high-frequency amplitude to low-frequency amplitude. The expression is: ; ED is the traffic sudden increase abnormality index.

[0031] A high traffic surge anomaly index indicates that high-frequency components account for a higher proportion of traffic fluctuations, indicating significant bursty traffic growth. This situation indicates rapid and unpredictable fluctuations in traffic demand. Static bandwidth allocation policies have limited real-time responsiveness, as fixed bandwidth allocations cannot adapt to sudden traffic surges. This can lead to network congestion, increased latency, and even critical service disruptions. A high burst anomaly index typically requires the introduction of more flexible bandwidth management policies to dynamically allocate resources, divert, or prioritize critical service traffic to improve adaptability to sudden traffic surges.

[0032] Conversely, when the traffic surge anomaly index is low, it indicates that the high-frequency components of traffic fluctuations are low, network traffic is relatively stable, and sudden increases are rare. In this low-surge environment, static bandwidth allocation policies offer better real-time responsiveness, as fixed bandwidth can adequately handle relatively stable traffic demand, making resource shortages and network congestion less likely to occur. In this scenario, static policies not only meet traffic demand but also reduce system complexity, eliminating the need for frequent adjustments and ensuring stable bandwidth allocation and efficient resource utilization.

[0033] The traffic peak fluctuation index and the traffic sudden increase abnormality index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the real-time response capability value label of the static bandwidth allocation strategy to burst traffic as the prediction target, and takes minimizing the sum of the prediction errors of the real-time response capability value labels of all static bandwidth allocation strategies to burst traffic as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The real-time response capability value of the static bandwidth allocation strategy to burst traffic is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0034] The method for obtaining the real-time response capability value of the static bandwidth allocation strategy to burst traffic is to obtain the corresponding function expression from the first eigenvector training data of the trained machine learning model: Where, is the output function of the model, is the peak flow fluctuation index, is the traffic surge abnormality index, The real-time response capability of the static bandwidth allocation policy to burst traffic.

[0035] The obtained real-time response capability value of the static bandwidth allocation policy to burst traffic is compared with the real-time response capability benchmark value of the static bandwidth allocation policy to burst traffic under the standard state in the historical data. If the real-time response capability value of the static bandwidth allocation policy to burst traffic is greater than the real-time response capability benchmark value of the static bandwidth allocation policy to burst traffic, it indicates that the real-time response capability of the static bandwidth allocation policy to burst traffic is high, and a high real-time response capability signal is generated. If the real-time response capability value of the static bandwidth allocation policy to burst traffic is less than or equal to the real-time response capability benchmark value of the static bandwidth allocation policy to burst traffic, it indicates that the real-time response capability of the static bandwidth allocation policy to burst traffic is low, and a low real-time response capability signal is generated.

[0036] S3: When the static bandwidth allocation strategy has low real-time response capability to sudden traffic, bandwidth resources are dynamically allocated according to the preset priority strategy so that high-priority applications can obtain guaranteed bandwidth. The bandwidth allocation of low-priority applications is adjusted to reduce their bandwidth usage and free up resources for high-priority applications.

[0037] Assume that the total bandwidth requirement of the current network is , driven by the needs of high-priority applications and the needs of low-priority applications constitute, then: Under burst traffic conditions, priority is given to bandwidth requirements of high-priority applications. A dynamic bandwidth allocation factor α (0<α≤1) is defined to dynamically adjust bandwidth allocation for high- and low-priority applications based on their priorities. The bandwidth allocation formula is: ; is the adjusted high-priority application bandwidth; ; is the adjusted low-priority application bandwidth; If the bandwidth requirements of high priority applications Greater than or equal to the allocated guaranteed bandwidth , then it is necessary to release bandwidth from low-priority applications, and the amount of released bandwidth ΔB is defined as: ; Reduce the released bandwidth ΔB from the low-priority application bandwidth to ensure that the needs of high-priority applications are met. The bandwidth of low-priority applications after adjustment is: Where, It is the bandwidth that is ultimately allocated to low-priority applications, and the remaining resources are allocated while ensuring the needs of high-priority applications.

[0038] S4: Analyze the effect of bandwidth adjustment and further refine bandwidth allocation based on real-time traffic feedback to ensure network service quality and resource utilization efficiency.

[0039] After bandwidth adjustment, the adjusted traffic data is collected in real time through the monitoring system to form a time series ,in Denote the traffic usage value at time point r. A sliding window of size W is set (for example, 10 minutes or a fixed number of data points). The data within this window is used to calculate real-time traffic fluctuations. For each time point r, the standard deviation within window W is calculated. As new traffic data arrives, the sliding window advances, removing the oldest data from the window and adding the latest data. The standard deviation calculation for the new window is repeated.

[0040] Set a standard deviation threshold to detect abnormal traffic fluctuations. Typically, this threshold is based on the standard deviation of the stable period in historical data. If the real-time standard deviation is greater than the standard deviation threshold, it indicates significant burst traffic and bandwidth allocation may need to be adjusted. In this case, increase bandwidth for high-priority applications to prioritize their bandwidth needs and prevent burst traffic from impacting critical services. Appropriately limit the bandwidth of low-priority applications to free up more bandwidth for high-priority applications. If the real-time standard deviation is less than or equal to the standard deviation threshold, traffic fluctuations are within a controllable range. Maintain the current bandwidth allocation to ensure resource efficiency.

[0041] Continuously monitor the effectiveness of the adjusted bandwidth, re-apply the adjusted traffic data to the sliding window, and update the real-time standard deviation. If, after multiple adjustments, the real-time standard deviation remains greater than the standard deviation threshold, the existing adjustment strategy is failing to mitigate fluctuations. Further increase the high-priority bandwidth share or reconfigure the priority strategy until traffic fluctuations return to a stable state.

[0042] In this embodiment, real-time traffic is monitored by setting up traffic monitoring equipment in the communication network, and traffic data of different users and applications, including traffic rate, traffic peak and transmission priority, are continuously collected. Based on historical traffic data and real-time traffic change trends, a burst traffic prediction model is constructed to analyze the peak fluctuation and sudden growth characteristics of traffic, so as to evaluate the real-time response capability of the static bandwidth allocation strategy to burst traffic. When the response capability is insufficient, the bandwidth allocation is dynamically adjusted according to the preset priority strategy, giving priority to guaranteeing the bandwidth requirements of high transmission priority applications, and appropriately reducing the bandwidth occupancy of low-priority applications to release resources. By analyzing the effect of bandwidth adjustment and combining real-time traffic feedback, bandwidth allocation is further optimized to ensure network service quality and resource utilization efficiency.

[0043] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0044] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the process or function described in 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 device. 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 one website, computer, server or data center to another website, computer, server or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. 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. The semiconductor medium can be a solid-state drive.

[0045] It should be understood that the term "and / or" herein is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist at the same time, and B exists alone, where A and B may be singular or plural. In addition, the character " / " herein generally indicates that the objects associated with each other are in an "or" relationship, but it may also indicate an "and / or" relationship, which can be understood by referring to the context. A person of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0046] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A communication network bandwidth control method, characterized by: The following steps are included: S1: Real-time traffic is monitored by traffic monitoring equipment installed in the communication network, and traffic data of different users and various applications is continuously collected. The traffic data includes traffic rate, traffic peak and transmission priority; S2: Based on collected historical traffic data and real-time traffic trends, a burst traffic prediction model is constructed. This model analyzes traffic peak fluctuation characteristics and sudden growth characteristics to evaluate the real-time response capability of static bandwidth allocation strategies to burst traffic. S3: When the static bandwidth allocation policy has low real-time response capabilities to burst traffic, bandwidth resources are dynamically allocated according to the preset priority policy, so that high-priority applications can obtain guaranteed bandwidth; Adjust bandwidth allocation for low-priority applications to reduce their bandwidth usage and free up resources for high-priority applications. S4: Analyze the effect of bandwidth adjustment and further refine bandwidth allocation based on real-time traffic feedback to ensure network service quality and resource utilization efficiency.

2. A communication network bandwidth control method according to claim 1, characterized in that: In S2, the peak fluctuation characteristics of the traffic are analyzed to generate a traffic peak fluctuation index. The method for obtaining the traffic peak fluctuation index is: Assume that the traffic peak value y follows the probability distribution ,in It is a parameter that describes the peak fluctuation characteristics. θ is a random variable with a prior distribution p(θ). It is assumed that θ obeys the normal distribution. ;in, and is the prior mean and variance; using the collected historical traffic peak data , x is the total number of historical traffic peak data, and the likelihood function is defined as , define each The observations of are independent and normally distributed, then: ; is the expectation of peak fluctuation, is the variance of traffic fluctuation; calculate the posterior distribution , making ; By minimizing q(θ) and the true posterior The KL divergence between , variational inference finds the optimal q(θ), which is expressed as: ; Equivalently, the maximum variational lower bound ELBO is defined as: ; After optimization, the optimal parameter mean of q(θ) is obtained and variance , use q(θ) to calculate the flow peak fluctuation index, the expression is: ; Where SQ is the flow peak fluctuation index.

3. A communication network bandwidth control method according to claim 2, characterized in that: In S2, a traffic surge anomaly index is generated by analyzing the sudden growth characteristics of traffic. The method for obtaining the traffic surge anomaly index is as follows: Collect traffic data within the M time period to form a time series ,in represents the flow value at time j, N is the total number of data points, and the time domain flow data is converted to the frequency domain using fast Fourier transform. The expression is: ;in is the kth frequency component in the frequency domain, Represents different frequency components, is the kernel function of Fourier transform, which calculates the spectrum amplitude of each frequency component. Represents the intensity of the signal at each frequency component, and the calculation expression is: ;in, and They are the real and imaginary parts of the frequency domain signal, respectively, and set the high frequency threshold , extract the sum of the amplitudes in the high frequency range , the expression is: ; Set the low frequency range threshold , calculate the sum of the amplitudes in the low-frequency range : The traffic surge anomaly index is defined using the ratio of high-frequency amplitude to low-frequency amplitude. The expression is: ; ED is the traffic sudden increase abnormality index.

4. A communication network bandwidth control method according to claim 3, characterized in that: In S2, the traffic peak fluctuation index and the traffic surge abnormality index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the real-time response capability value label of the static bandwidth allocation strategy to burst traffic as the prediction target, and takes minimizing the sum of the prediction errors of the real-time response capability value labels of all static bandwidth allocation strategies to burst traffic as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The real-time response capability value of the static bandwidth allocation strategy to burst traffic is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

5. A communication network bandwidth control method according to claim 4, characterized in that: In S2, the obtained real-time response capability value of the static bandwidth allocation strategy to burst traffic is compared with the real-time response capability benchmark value of the static bandwidth allocation strategy to burst traffic under the standard state in the historical data. If the real-time response capability value of the static bandwidth allocation strategy to burst traffic is greater than the real-time response capability benchmark value of the static bandwidth allocation strategy to burst traffic, it indicates that the real-time response capability of the static bandwidth allocation strategy to burst traffic is high, and a high real-time response capability signal is generated at this time; if the real-time response capability value of the static bandwidth allocation strategy to burst traffic is less than or equal to the real-time response capability benchmark value of the static bandwidth allocation strategy to burst traffic, it indicates that the real-time response capability of the static bandwidth allocation strategy to burst traffic is low, and a low real-time response capability signal is generated at this time.

6. A communication network bandwidth control method according to claim 1, characterized in that: In S3, when the static bandwidth allocation strategy has low real-time response capability to sudden traffic, bandwidth resources are dynamically allocated according to the preset priority strategy, specifically: Assume that the total bandwidth requirement of the current network is , driven by the needs of high-priority applications and the needs of low-priority applications constitute, then: Under burst traffic conditions, priority is given to bandwidth requirements of high-priority applications. A dynamic bandwidth allocation factor α is defined to dynamically adjust bandwidth allocation for high- and low-priority applications based on their priorities. The bandwidth allocation formula is: ; is the adjusted high-priority application bandwidth; ; is the adjusted bandwidth for low-priority applications.

7. A communication network bandwidth control method according to claim 6, characterized in that: If the bandwidth requirements of high priority applications Greater than or equal to the allocated guaranteed bandwidth , then it is necessary to release bandwidth from low-priority applications, and the amount of released bandwidth ΔB is defined as: ; Reduce the released bandwidth ΔB from the low-priority application bandwidth to ensure that the needs of high-priority applications are met; the bandwidth of low-priority applications after adjustment is: Where, It is the bandwidth that is ultimately allocated to low-priority applications, and the remaining resources are allocated while ensuring the needs of high-priority applications.

8. A communication network bandwidth control method according to claim 7, characterized in that: In S4, the effect of bandwidth adjustment is analyzed, and bandwidth allocation is further refined based on real-time traffic feedback. Specifically: After bandwidth adjustment, the adjusted traffic data is collected in real time through the monitoring system to form a time series ,in Represents the traffic usage value at the rth time point. Set the size of the sliding window W. For each time point r, calculate the standard deviation within the window W. As new traffic data arrives, the sliding window advances and the standard deviation within the new window is repeatedly calculated; Set a standard deviation threshold to detect abnormal traffic fluctuations. If the real-time standard deviation is greater than the standard deviation threshold, bandwidth allocation needs to be adjusted. In this case, bandwidth for high-priority applications is increased to prioritize their bandwidth needs. Bandwidth for low-priority applications is restricted to free up more bandwidth for high-priority applications. If the real-time standard deviation is less than or equal to the standard deviation threshold, traffic fluctuations are within a controllable range and the current bandwidth allocation remains unchanged.

Images