Multi-dimensional adaptive network evaluation method and device based on software defined network

Abstract

The invention provides a multi-dimensional adaptive network evaluation method and device based on a software defined network, and the method comprises the steps: determining to-be-evaluated network indexes and a network application scene, obtaining a current measurement value and a sensitivity coefficient of each to-be-evaluated network index, and determining each index score based on each current measurement value and the sensitivity coefficient; determining a weight coefficient of each to-be-evaluated network index based on a network application scene, and determining a current network comprehensive score based on each index score and the weight coefficient; determining the network change condition of the current network relative to the network at the previous moment based on the current network comprehensive score; when the network changes, calculating each deterioration degree value based on each index score, calculating each influence factor based on each deterioration degree value and the weight coefficient, and determining the evaluation priority of each to-be-evaluated network index based on the influence factor; and determining the current sampling frequency of the to-be-evaluated network index based on each evaluation priority. According to the invention, network resources occupied by network evaluation are reduced, and the accuracy of network evaluation is improved.

Description

Technical Field

[0001] The present invention relates to the field of network evaluation technology, and in particular to a multi-dimensional adaptive network evaluation method and device based on software-defined networking. Background Art

[0002] In recent years, with the continuous development of network technology, network complexity and diversity have also increased, making the optimal utilization of network resources a pressing issue. To achieve this goal, in-depth network evaluation and analysis are necessary to understand performance, reliability, and security. Network evaluation plays a crucial role in promoting better utilization of network resources. Through systematic network evaluation, we can better understand and control the network's operational status, laying a solid foundation for providing higher-quality services. With the advent of Software-Defined Networking (SDN), its centralized control and programmability have revolutionized network architecture. The core concept of SDN is to separate the control plane from the data plane of network devices, significantly improving network flexibility and dynamic adjustment capabilities. This presents new opportunities for existing network evaluation systems. Its centralized control architecture enables network evaluation to monitor and respond to network status changes in real time, achieving more accurate network metric measurement. Network evaluation can leverage SDN's dynamic measurement capabilities to rapidly adapt to network status changes, ensuring the real-time and accuracy of network evaluation activities.

[0003] From a network assessment perspective, network assessment should serve the network rather than become a burden. On the one hand, network managers hope to reduce the impact of network assessment on network resources when network quality degrades. On the other hand, they hope that reducing network assessment overhead will not significantly affect the accuracy of network assessment. However, current technologies still lack the ability to strike a balance between these two aspects.

[0004] In terms of network data measurement solutions, existing network measurements are categorized as active, passive, and hybrid. The research approach for active network measurement is to infer the overall network performance by constructing, observing, and analyzing the operational status of probe packets on the network. The primary collected data includes the number of requests, throughput, response time, and resource allocation and release times. Passive measurement solutions are limited by switch performance and typically utilize sampling and compression mechanisms, which compromise measurement accuracy. With the advent of SDN, more open control and data plane programmability has reshaped traditional network measurement. With the further development of network technology, network telemetry has become a mainstream technical term, referring to top-down network data collection techniques. However, for network telemetry, network-adaptive network measurement techniques remain a challenge.

[0005] Due to the flexibility of SDN networks in measurement, methods for varying network measurements based on network conditions already exist. However, these methods rely on identifying the size of different flows to adjust the data sampling frequency, implementing elastic network measurement mechanisms through compression sketches, and adjusting measurement intervals through machine learning. While these methods allow for varying network measurements based on network conditions, they fail to prioritize the evaluation of key network metrics when network conditions change, hindering the balance between network performance evaluation and network resource utilization. Therefore, improving the accuracy of network evaluation while reducing network resource usage during changing network conditions is a pressing technical challenge. Summary of the Invention

[0006] In view of this, an embodiment of the present invention provides a multi-dimensional adaptive network evaluation method and apparatus based on software-defined networking to eliminate or improve one or more defects in the prior art.

[0007] One aspect of the present invention provides a multi-dimensional adaptive network evaluation method based on software-defined networking, the method comprising:

[0008] Determine the network indicator to be evaluated and the network application scenario, obtain the current measurement value of each of the network indicators to be evaluated, determine the sensitivity coefficient of each of the network indicators to be evaluated based on the network application scenario, and determine the indicator score of each of the network indicators to be evaluated based on the current measurement value and sensitivity coefficient of each of the network indicators to be evaluated;

[0009] Determine a weight coefficient of each of the network indicators to be evaluated based on the network application scenario, and determine a current network comprehensive score based on the indicator score and the weight coefficient of each of the network indicators to be evaluated;

[0010] Determining a network change of the current network relative to the network at a previous moment based on the current network comprehensive score;

[0011] When the network changes, calculating the deterioration degree value of each network indicator to be evaluated based on the indicator score of each network indicator to be evaluated, calculating the impact factor of each network indicator to be evaluated based on the deterioration degree value and the weight coefficient of each network indicator to be evaluated, and determining the evaluation priority of each network indicator to be evaluated based on each impact factor;

[0012] The current sampling frequency of each of the network indicators to be evaluated is determined based on the evaluation priority of each of the network indicators to be evaluated, and the network evaluation is performed based on each of the current sampling frequencies.

[0013] In some embodiments of the present invention, the calculation formula for the indicator score of the network indicator to be evaluated is:

[0014]

[0015] Eω represents the index score of the network index ω to be evaluated, Mω represents the current measured value of the network index ω to be evaluated, and Mω ll represents the lower limit of the network indicator ω to be evaluated, δω represents the sensitivity coefficient of the network indicator ω to be evaluated, and Mω sl Indicates the upper limit value of the network indicator ω to be evaluated.

[0016] In some embodiments of the present invention, determining the current network comprehensive score based on the indicator score and weight coefficient of each of the network indicators to be evaluated includes:

[0017] Determining an upper limit filter coefficient and a lower limit filter coefficient of the network indicator to be evaluated based on a current measurement value, an upper limit value of the indicator, and a lower limit value of the indicator;

[0018] The current network comprehensive score is determined based on the indicator score, weight coefficient, upper limit filtering coefficient and lower limit filtering coefficient of each of the network indicators to be evaluated.

[0019] In some embodiments of the present invention, the upper limit filter coefficient is:

[0020]

[0021] The lower limit filter coefficient is:

[0022]

[0023] In some embodiments of the present invention, the calculation formula for the current network comprehensive score is:

[0024]

[0025] Where n represents the total number of network indicators to be evaluated.

[0026] In some embodiments of the present invention, the deterioration degree value of each network indicator to be evaluated is D ω =1-E ω , the impact factor S of each network indicator to be evaluated ω =(1-E ω )·f ω , where S v represents the impact factor of the network indicator v to be evaluated, E v represents the index score of the network index v to be evaluated, f v Represents the weight coefficient of the network indicator v to be evaluated.

[0027] In some embodiments of the present invention, determining the current sampling frequency of each of the network indicators to be evaluated based on the evaluation priority of each of the network indicators to be evaluated includes:

[0028] Determining a baseline frequency for each of the network indicators to be evaluated;

[0029] Determining a reduction coefficient of each of the network indicators to be evaluated based on the evaluation priority of each of the network indicators to be evaluated;

[0030] The current sampling frequency of each of the network indicators to be evaluated is calculated based on the baseline frequency, the reduction coefficient and the deterioration degree value of the network indicator to be evaluated.

[0031] In some embodiments of the present invention, the current sampling frequency is calculated as follows: V dv =V jv ×(1-α ω ·D v ), where V dω Represents the current sampling frequency of the network indicator ω to be evaluated, V jω represents the baseline frequency of the network indicator ω to be evaluated, α ω Denotes the reduction coefficient of the network indicator ω to be evaluated, D ω represents the deterioration value of the network indicator ω to be evaluated; and / or,

[0032] The network indicators to be evaluated include latency indicators, packet loss indicators and bandwidth indicators.

[0033] According to another aspect of the present invention, a multi-dimensional adaptive network evaluation system based on software-defined networking is also provided. The system includes a processor, a memory, and a computer program stored in the memory. The processor is used to execute the computer program. When the computer program is executed, the system implements the steps of the method described in any of the above embodiments.

[0034] According to yet another aspect of the present invention, a computer-readable storage medium is disclosed, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in any of the above embodiments are implemented.

[0035] The multi-dimensional adaptive network evaluation method based on software-defined networks disclosed in the above embodiment of the present invention first determines the indicator score of each network indicator to be evaluated, then calculates the comprehensive score of the network based on the indicator score of each network indicator to be evaluated, and determines whether the network status has changed based on the comprehensive score of the network. When the network status changes, the deterioration value, impact factor and evaluation priority of each network indicator to be evaluated are calculated, and finally the current sampling frequency of each network indicator to be evaluated is determined based on the evaluation priority. When the network status changes, the method associates and combines the results of the multi-dimensional network evaluation algorithm with the adaptive measurement strategy, that is, according to the evaluation priority of each network indicator, each network indicator is dynamically measured with emphasis, thereby adaptively adjusting the sampling frequency of each network indicator according to the situation of network resources, thereby achieving a balance between network resources and network performance evaluation, that is, while reducing the network resources occupied by the network evaluation, the accuracy of the network evaluation is also improved.

[0036] Additional advantages, objects, and features of the present invention will be set forth in part in the following description and will become apparent to those skilled in the art upon examination of the following or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained by the structures particularly pointed out in the description and drawings.

[0037] Those skilled in the art will understand that the purposes and advantages that can be achieved by the present invention are not limited to the above specific descriptions, and the above and other purposes that can be achieved by the present invention will be more clearly understood based on the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The drawings described herein are intended to provide a further understanding of the present invention, constitute a part of this application, and do not constitute a limitation of the present invention. The components in the drawings are not drawn to scale, but are merely for the purpose of illustrating the principles of the present invention. To facilitate the illustration and description of certain portions of the present invention, corresponding portions in the drawings may be exaggerated, that is, may be larger than other components in an exemplary device actually manufactured according to the present invention. In the drawings:

[0039] Figure 1 Schematic diagram of a flow chart of a multi-dimensional adaptive network evaluation method based on software-defined networking according to an embodiment of the present application.

[0040] Figure 2 Schematic diagram of the architecture of a multi-dimensional adaptive network evaluation system based on software-defined networking according to an embodiment of the present application.

[0041] Figure 3 This is a flowchart of a multi-dimensional adaptive network evaluation method based on software-defined networking according to another embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and the accompanying drawings. Here, the exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0043] It should also be noted that, in order to avoid obscuring the present invention due to unnecessary details, the accompanying drawings only show structures and / or processing steps closely related to the solutions according to the present invention, while other details that are not closely related to the present invention are omitted.

[0044] It should be emphasized that the term "include / comprises" when used herein refers to the existence of features, elements, steps or components, but does not exclude the existence or addition of one or more other features, elements, steps or components.

[0045] It should also be noted here that, unless otherwise specified, the term "connection" in this article can refer not only to a direct connection, but also to an indirect connection with an intermediary, and not only to a wired connection but also to a wireless connection, and the specific connection can be changed based on the actual application scenario.

[0046] Existing network assessment methods are unable to adaptively adjust the emphasis of network indicators based on comprehensive network status assessment results, making it difficult to strike a balance between network performance assessment and network resources. Furthermore, when network resources are limited, the precision and accuracy of network assessment cannot be guaranteed. Therefore, when network resources are limited, resolving the conflict between network assessment accuracy and network load is an urgent issue. To address a series of issues with existing network assessment methods, the present application provides a multi-dimensional adaptive network assessment method and apparatus based on software-defined networking.

[0047] Traditional network monitoring methods usually use a fixed frequency and mode of data collection. This approach is feasible when the network load is low, but when the network load is high or congested, frequent data collection may increase the network burden, affect performance, and even cause service interruptions. The core of this contradiction is that different network conditions require different network indicators to be measured. Measuring all network indicators at the same frequency will result in insufficient measurement accuracy of key and important network indicators, while relatively less important indicators will occupy additional network resources. The multi-dimensional adaptive network evaluation method and device of the present application empower the network indicator measurement strategy with multi-dimensional network evaluation results, thereby focusing on the measurement of network indicators and avoiding the measurement work from aggravating network congestion.

[0048] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the accompanying drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

[0049] Figure 1 FIG. 1 is a flow chart of a multi-dimensional adaptive network evaluation method based on a software-defined network according to an embodiment of the present application. Figure 1 As shown, the multi-dimensional adaptive network evaluation method based on software-defined networking includes at least steps S10 to S50.

[0050] Step S10: Determine the network indicators to be evaluated and the network application scenario, obtain the current measurement value of each of the network indicators to be evaluated, determine the sensitivity coefficient of each of the network indicators to be evaluated based on the network application scenario, and determine the indicator score of each of the network indicators to be evaluated based on the current measurement value and sensitivity coefficient of each of the network indicators to be evaluated.

[0051] In this embodiment, there are generally multiple network indicators to be evaluated. For example, the network indicators to be evaluated may include latency (td), packet loss (pl) and bandwidth (bd), etc., and exemplary network application scenarios include gaming scenarios and video call scenarios. The current measured value of the network indicator to be evaluated is used to represent the network performance corresponding to the indicator under the current network state, such as the current network latency, whether packet loss is frequent, whether the bandwidth is sufficient, etc. Different network application scenarios focus on different business needs, such as gaming scenarios focusing on latency and video call scenarios focusing on bandwidth. It can be understood that the types and quantities of network indicators to be evaluated listed in this embodiment are only some examples. In other embodiments, the network indicators to be evaluated may also include others; and the specific network application scenarios may also be application scenarios other than gaming scenarios and video call scenarios.

[0052] Because different network metrics to be evaluated have different units, it is impossible to compare the importance of network metrics based on the source measurement values. Therefore, it is necessary to eliminate the unit and range differences between the measurement values. Therefore, it is necessary to map the original measurement values ​​to a relative range. Assume that the measurement values ​​of each network metric to be evaluated are Mω. When the network metrics to be evaluated may include latency, packet loss, and bandwidth, ω is specifically td, pl, or bd.

[0053] In order to eliminate the unit differences and range differences between the measurement values, the current measurement values ​​of the network indicators to be evaluated are first normalized. Among them, Mω ll and Mω sl are the preset lower limit value and upper limit value of the network indicator ω to be evaluated, and the upper limit value and lower limit value are the upper limit and lower limit of the reasonable range of the defined network indicator respectively. ω is the normalized measurement value of the network indicator ω to be evaluated, and the normalized measurement value result x ω∈[0,1], Mv represents the current measured value of the network indicator ω to be evaluated, that is, the actual measured network indicator value at the current moment.

[0054] The sensitivity coefficient of the network indicator to be evaluated can be flexibly adjusted according to business needs. For example, the sensitivity can be increased for scenarios with high real-time requirements (video calls) and reduced for scenarios with high tolerance to fluctuations (file downloads). The sensitivity coefficient of the network indicator ω to be evaluated can be adjusted by δ ω Indicates that the normalized measurement value result is further adjusted according to the parameter (δ ω ) is scaled. The purpose of adjusting the parameter is to control the sensitivity of the index score of each network index to be evaluated - δ ω The larger the value, the more sensitive the indicator score is to changes in the measured value; the smaller the value, the smoother the score change. ω =1.2, the ratio value will be magnified, and the score will react more obviously to small changes; if δ ω =0.8, the score changes more gradually.

[0055] Furthermore, a Gaussian function is used to map the current measurement value to a score between 0 and 1. The characteristic of the Gaussian function is that when the measurement value is close to the middle range, the score is high, indicating a normal state; when the measurement value is close to the upper or lower limit, the score drops rapidly and approaches 0, indicating a serious abnormality. Therefore, the unified indicator scoring formula for different network indicators is: Eω represents the index score of the network index ω to be evaluated, Mω represents the current measured value of the network index ω to be evaluated, and Mω ll represents the lower limit of the network indicator ω to be evaluated, δω represents the sensitivity coefficient of the network indicator ω to be evaluated, and Mω sl Indicates the upper limit value of the network indicator ω to be evaluated.

[0056] Step S20: determining a weight coefficient of each of the network indicators to be evaluated based on the network application scenario, and determining a current network comprehensive score based on the indicator score and the weight coefficient of each of the network indicators to be evaluated.

[0057] In this step, the weight coefficients for each network metric to be evaluated are determined based on the network application scenario's requirements for each network metric. Setting weights for each network metric to be evaluated is intended to reflect the importance of each network metric. Assuming the current measured values ​​of the network metric to be evaluated are Mtd, Mpl, and Mbd, the weight coefficients for the network metric to be evaluated are ftd, fpl, and fbd, respectively, and satisfying ftd + fpl + fbd = 1. In this embodiment, the purpose of setting weight coefficients for each network metric to be evaluated is to adapt to different service requirements by weighting ftd, fpl, and fbd. For example, in gaming scenarios, ftd and fbd are higher, prioritizing latency and bandwidth; in financial scenarios, fpl is higher, prioritizing packet loss rate control.

[0058] Furthermore, the current network comprehensive score is determined based on the indicator score and weight coefficient of each of the network indicators to be evaluated, which may specifically include the following steps: determining the upper limit filter coefficient and the lower limit filter coefficient of the network indicator to be evaluated based on the current measurement value, the indicator upper limit value and the indicator lower limit value of the network indicator to be evaluated; and determining the current network comprehensive score based on the indicator score, the weight coefficient, the upper limit filter coefficient and the lower limit filter coefficient of each of the network indicators to be evaluated.

[0059] In the above embodiment, in order to ensure the validity of the indicator score of the current measurement value, a conditional filter is further set. If the current measurement value exceeds the preset upper limit or is lower than the preset lower limit, the indicator score of the corresponding network indicator is directly reset to zero, that is, to avoid abnormal values ​​from interfering with the overall evaluation. The conditional filter items are Jω_sl and Jω_ll, which are specifically defined as follows: J ω_sl and J ω_ll They represent the upper and lower filter coefficients corresponding to the network indicator ω to be evaluated, M ω 、M ω_ll and M ω_sl The meaning of is as above. In this embodiment, the upper limit filter coefficient and the lower limit filter coefficient are used to determine whether the current measurement value of the network indicator is abnormal. If the indicator exceeds the upper limit value of the indicator or is lower than the lower limit value of the indicator, the indicator score item corresponding to the corresponding network indicator is directly reset to zero to avoid abnormal data interfering with the overall evaluation.

[0060] In step S20, each network indicator is assigned a weight f according to business requirements. ω , reflecting the degree of emphasis on each network indicator in different network scenarios; combining the weight coefficient with the current measurement value to ensure that network resources are tilted towards key network indicators, and then multiplying it with the indicator score of the network indicator to obtain the current network score corresponding to a single network indicator (latency, packet loss rate or bandwidth), such as The meaning of each parameter is as above.

[0061] Furthermore, the current network scores of all network indicators to be evaluated are added together to obtain the current network comprehensive score Qt. The lower the current network comprehensive score Qt, the more serious the network problem. The system will prioritize the measurement frequency of high-weight indicators while reducing the monitoring intensity of secondary indicators, thereby balancing the evaluation accuracy and network load under limited resources. This design can not only accurately capture key anomalies, but also flexibly adapt to the needs of different business scenarios. Specifically, the calculation formula for the current network comprehensive score is: Where n represents the total number of network indicators to be evaluated.

[0062] When the network indicators to be evaluated are latency, packet loss rate and bandwidth, the current network comprehensive score

[0063] Step S30: determining a network change of the current network relative to the network at the previous moment based on the current network comprehensive score.

[0064] In this step, the current network status is determined based on the current network comprehensive score calculated in step S20. That is, when the current network comprehensive score changes relative to the previous network comprehensive score, it can be determined that the network status has changed.

[0065] Step S40: When the network changes, the deterioration degree value of each network indicator to be evaluated is calculated based on the indicator score of each network indicator to be evaluated, the impact factor of each network indicator to be evaluated is calculated based on the deterioration degree value and weight coefficient of each network indicator to be evaluated, and the evaluation priority of each network indicator to be evaluated is determined based on each impact factor.

[0066] When the network changes, this step further calculates the deterioration value of each network indicator to be evaluated. The deterioration value D ω The specific one can be based on D ω =1-E ω This means that the larger the calculated deterioration value is, the more serious the network problem is at that moment.

[0067] Further, the impact factor of each network indicator to be evaluated can be calculated based on the product of the deterioration degree value of each network indicator to be evaluated and the weight coefficient, such as the impact factor S of each network indicator to be evaluated. ω =(1-E ω )·f ω , where S ω represents the impact factor of the network indicator ω to be evaluated, E ω represents the index score of the network index ω to be evaluated, f ω Represents the weight coefficient of the network indicator ω to be evaluated. From the above content, we can know that the weight coefficient f ωReflecting the inherent importance of network indicators ω in network scenarios, the specific ω Adjust the order to ensure that the most serious indicators are processed first when the network deteriorates. For example, if the bandwidth (bd) of E bd Dips (i.e. 1-E bd rise), and its weight f bd Higher, then S bd Significantly increased, triggering an increase in priority.

[0068] After calculating the impact factors of each network indicator to be evaluated, the importance of each indicator can be sorted according to the impact factors of each network indicator to be evaluated to obtain the evaluation priority of the network indicator, that is,

[0069] Step S50: determining the current sampling frequency of each of the network indicators to be evaluated based on the evaluation priority of each of the network indicators to be evaluated, and performing network evaluation based on each current sampling frequency.

[0070] After obtaining the evaluation priorities of the network indicators to be evaluated, the current sampling frequency of each network indicator to be evaluated is further determined according to the evaluation priorities of each network indicator to be evaluated.

[0071] This step can be specifically implemented based on the policy execution module of the multi-dimensional adaptive network evaluation system. That is, when the global link availability (QGt) decreases, the policy execution module can perform focused adaptive measurements on various network indicators based on the results of the network indicator importance ranking algorithm given above.

[0072] In one embodiment, determining the current sampling frequency of each network indicator to be evaluated based on the evaluation priority of each network indicator to be evaluated may specifically include: determining the baseline frequency of each network indicator to be evaluated; determining the reduction coefficient of each network indicator to be evaluated based on the evaluation priority of each network indicator to be evaluated; and calculating the current sampling frequency of each network indicator to be evaluated based on the baseline frequency, reduction coefficient and deterioration degree value of the network indicator to be evaluated.

[0073] First, based on the resource capacity under normal network conditions, a baseline frequency is set for each network metric: assuming latency (td) is measured twice per second, packet loss (pl) is measured once per second, and bandwidth (bd) is measured once every five seconds. This baseline frequency is the optimal frequency for measuring each network metric under normal network conditions. Of course, the baseline frequency can be adjusted based on actual network scenarios.

[0074] Furthermore, the indicator priority is bound to the frequency: High priority indicator (ranked 1): Keep the baseline frequency unchanged to avoid monitoring blind spots caused by frequency reduction. Medium / low priority indicators (ranked 2 / 3): Dynamically reduce the frequency according to priority to release resources. For example, the frequency reduction rule is dynamic frequency reduction. When the current network comprehensive score decreases, the measurement frequency of the medium / low priority indicator is adjusted according to the following formula: V dω =V jω ×(1-α ω ·D ω ), where V dω Represents the current sampling frequency of the network indicator ω to be evaluated, V jω represents the baseline frequency of the network indicator ω to be evaluated, α ω Indicates the reduction coefficient of the network indicator ω to be evaluated (the default is 0.5, which can be adjusted based on actual needs, α ω Used to control the frequency reduction range), D ω In this embodiment, the current sampling frequency of the network indicator to be evaluated can be dynamically adjusted according to the priority, that is, the high priority maintains the baseline frequency, and the low priority reduces the frequency according to the degree of deterioration.

[0075] The multi-dimensional adaptive network evaluation method based on software-defined networking in the above embodiment automatically determines which network indicators are most critical based on the current network status (such as latency, whether packet loss is frequent, and whether bandwidth is sufficient) and the actual needs of different services (for example, games focus on latency, video streaming focuses on bandwidth); Figure 3 As shown, when a problem occurs in the network, it is prioritized to ensure that the monitoring frequency of these important network indicators remains unchanged in order to accurately capture data. Secondly, the monitoring frequency of secondary network indicators is automatically reduced (for example, the delay originally measured twice per second is changed to once per second), reducing unnecessary resource usage and avoiding the evaluation itself from dragging down network performance. In addition, the multi-dimensional adaptive network evaluation method of the present application allows parameters to be adjusted according to specific network scenarios, such as the weight coefficients and frequency reduction amplitudes of different network indicators, which can be adjusted based on actual application scenarios. This method flexibly adapts to various business needs. That is, the present application ensures the accuracy of key data and effectively reduces the network burden by the method of "precise monitoring of key indicators and resource saving of secondary indicators", and achieves a balance between evaluation effect and network load under limited resources.

[0076] Through the above embodiments, it can be found that the multi-dimensional adaptive network evaluation method of the present application achieves focused monitoring of important network indicators while ensuring the accuracy of network performance evaluation and limited network resources. That is, the present application associates and combines the results of the multi-dimensional network evaluation algorithm with the adaptive measurement strategy. First, the network evaluation algorithm obtains the overall network evaluation result and the network evaluation result of each dimension based on the measurement, and obtains the network indicator priority evaluation result based on the above two results. Secondly, when the overall network evaluation result decreases, the adaptive measurement strategy performs focused dynamic measurement of each network indicator according to the priority ranking of the network indicators, thereby ensuring that the measurement is focused when the network condition decreases while reducing the negative impact of the network evaluation process on the network.

[0077] Correspondingly, the present invention also provides a multi-dimensional adaptive network evaluation system based on software-defined networking, the system comprising a processor, a memory, and a computer program stored in the memory, the processor being used to execute the computer program, and when the computer program is executed, the system implements the steps of any of the methods described above.

[0078] Figure 2 This is a schematic diagram of the architecture of a multi-dimensional adaptive network evaluation system based on software-defined networking according to an embodiment of the present application. Figure 2As shown, the system may include an information collection module 001, a policy execution module 003, and an algorithm evaluation module 002. The core of the algorithm evaluation module 002 is a network indicator importance ranking algorithm (also known as a network indicator priority evaluation algorithm), and the core of the policy execution module 003 is an adaptive measurement strategy based on the network indicator importance ranking algorithm. In this system, after collecting the specified network indicators to be evaluated, the information collection module 001 uploads the indicator data to the algorithm evaluation module; the specific implementation of the information collection module 001 is an SDN network controller, which writes the data into a database after collecting relevant indicators; the algorithm evaluation module 002 reads the network indicator data it needs from the database, and evaluates the network as a whole based on the collected multi-dimensional network indicator data, specifically performing a multi-dimensional evaluation of the network from the dimensions of delay, packet loss rate, and bandwidth, and provides a network indicator priority evaluation result obtained based on the multi-dimensional evaluation result, and writes the network indicator priority evaluation result into the database. After the priority assessment is complete, the SDN controller must also implement policy execution logic. This involves periodically reading the network metric priority ranking assessment results from the database and modifying the controller's measurement strategy for network data based on the priority ranking assessment results. For high-priority network metrics, the policy execution module 003 minimizes the SDN controller's measurement frequency to maintain high measurement accuracy. For low-priority metrics, the policy execution module 003 modifies the SDN controller's measurement frequency, appropriately reducing measurement accuracy. When network conditions are good, the normal sampling frequency is used. When network conditions decline, the sampling frequency of low-priority network metrics is reduced based on resource priority.

[0079] The above-described embodiment solves the problem of balancing network assessment overhead and network load. By comprehensively considering multiple key network performance indicators and integrating the network assessment module with the adaptive network data acquisition module, the overall system's network assessment results are more accurate. Furthermore, the multi-dimensional network assessment system rapidly responds to network status changes, enabling network administrators to make timely decisions. Through accurate network performance assessment and timely network optimization, users' network experience is significantly improved, leading to faster, more stable, and more secure network services.

[0080] Embodiments of the present invention further provide a computer-readable storage medium and a computer program product, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in any of the above embodiments are implemented. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the art.

[0081] It should be understood by those skilled in the art that the various exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is specifically 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 the present invention. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0082] It should be understood that the present invention is not limited to the specific configurations and processes described above and illustrated in the figures. For the sake of brevity, a detailed description of known methods is omitted. In the above embodiments, several specific steps are described and illustrated as examples. However, the method of the present invention is not limited to the specific steps described and illustrated. Those skilled in the art may make various changes, modifications, and additions, or change the order of the steps after understanding the spirit of the present invention.

[0083] In the present invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or replace features of other embodiments.

[0084] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations to the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A multi-dimensional adaptive network evaluation method based on software-defined networking, characterized in that: The network evaluation method includes: Determine the network indicator to be evaluated and the network application scenario, obtain the current measurement value of each of the network indicators to be evaluated, determine the sensitivity coefficient of each of the network indicators to be evaluated based on the network application scenario, and determine the indicator score of each of the network indicators to be evaluated based on the current measurement value and sensitivity coefficient of each of the network indicators to be evaluated; Determine a weight coefficient of each of the network indicators to be evaluated based on the network application scenario, and determine a current network comprehensive score based on the indicator score and the weight coefficient of each of the network indicators to be evaluated; Determining a network change of the current network relative to the network at a previous moment based on the current network comprehensive score; When the network changes, calculating the deterioration degree value of each network indicator to be evaluated based on the indicator score of each network indicator to be evaluated, calculating the impact factor of each network indicator to be evaluated based on the deterioration degree value and the weight coefficient of each network indicator to be evaluated, and determining the evaluation priority of each network indicator to be evaluated based on each impact factor; The current sampling frequency of each of the network indicators to be evaluated is determined based on the evaluation priority of each of the network indicators to be evaluated, and the network evaluation is performed based on each of the current sampling frequencies.

2. The multi-dimensional adaptive network evaluation method based on software-defined networking according to claim 1 is characterized in that: The calculation formula for the indicator score of the network indicator to be evaluated is: Eω represents the index score of the network index ω to be evaluated, Mω represents the current measured value of the network index ω to be evaluated, and Mω ll represents the lower limit of the network indicator ω to be evaluated, δω represents the sensitivity coefficient of the network indicator ω to be evaluated, and Mω sl Indicates the upper limit value of the network indicator ω to be evaluated.

3. The multi-dimensional adaptive network evaluation method based on software-defined networking according to claim 2 is characterized in that: Determine the current network comprehensive score based on the indicator scores and weight coefficients of each of the network indicators to be evaluated, including: Determining an upper limit filter coefficient and a lower limit filter coefficient of the network indicator to be evaluated based on a current measurement value, an upper limit value of the indicator, and a lower limit value of the indicator; The current network comprehensive score is determined based on the indicator score, weight coefficient, upper limit filtering coefficient and lower limit filtering coefficient of each of the network indicators to be evaluated.

4. The multi-dimensional adaptive network evaluation method based on software-defined networking according to claim 3 is characterized in that: The upper limit filter coefficient is: The lower limit filter coefficient is:

5. The multi-dimensional adaptive network evaluation method based on software-defined networking according to claim 4 is characterized in that: The calculation formula of the current network comprehensive score is: Where n represents the total number of network indicators to be evaluated.

6. The multi-dimensional adaptive network evaluation method based on software-defined networking according to claim 1, characterized in that: The deterioration value of each network indicator to be evaluated is D ω =1-E ω , the impact factor S of each network indicator to be evaluated ω =(1-E ω )·f ω , where S ω represents the impact factor of the network indicator ω to be evaluated, E ω represents the index score of the network index ω to be evaluated, f ω Represents the weight coefficient of the network indicator ω to be evaluated.

7. The multi-dimensional adaptive network evaluation method based on software defined network according to claim 1, characterized in that: Determining a current sampling frequency of each of the network indicators to be evaluated based on the evaluation priority of each of the network indicators to be evaluated includes: Determining a baseline frequency for each of the network indicators to be evaluated; Determining a reduction coefficient of each of the network indicators to be evaluated based on the evaluation priority of each of the network indicators to be evaluated; The current sampling frequency of each of the network indicators to be evaluated is calculated based on the baseline frequency, the reduction coefficient and the deterioration degree value of the network indicator to be evaluated.

8. The multi-dimensional adaptive network evaluation method based on software-defined networking according to claim 7, characterized in that: The calculation formula for the current sampling frequency is: V dω =V jω ×(1-α ω ·D ω ), where V dω Represents the current sampling frequency of the network indicator ω to be evaluated, V jω represents the baseline frequency of the network indicator ω to be evaluated, α ω Denotes the reduction coefficient of the network indicator ω to be evaluated, D ω represents the deterioration value of the network indicator ω to be evaluated; and / or, The network indicators to be evaluated include latency indicators, packet loss indicators and bandwidth indicators.

9. A multi-dimensional adaptive network evaluation system based on software-defined networking, the system comprising a processor, a memory, and a computer program stored in the memory, characterized in that: The processor is configured to execute the computer program. When the computer program is executed, the system implements the steps of the method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images