CN111010312B - Network quality evaluation method and server - Google Patents
Network quality evaluation method and server Download PDFInfo
- Publication number
- CN111010312B CN111010312B CN201911192499.XA CN201911192499A CN111010312B CN 111010312 B CN111010312 B CN 111010312B CN 201911192499 A CN201911192499 A CN 201911192499A CN 111010312 B CN111010312 B CN 111010312B
- Authority
- CN
- China
- Prior art keywords
- quality
- node
- grade
- obtaining
- test
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/16—Threshold monitoring
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The embodiment of the invention relates to the technical field of Internet and discloses a network quality evaluation method and a server. The network quality evaluation method comprises the following steps: acquiring a plurality of detection data obtained by sending detection requests to a plurality of areas by a node to be detected; obtaining test data of a plurality of quality levels according to the corresponding relation between the detection data of each area and a preset quality level and a detection data threshold; and obtaining the network quality evaluation value of the node to be tested according to the corresponding relation between the test data of the plurality of quality grades and the preset quality grade and weight parameter. According to the invention, the actual application environment can be simulated more accurately, so that the acquired network quality evaluation value of the node to be measured is more accurate.
Description
Technical Field
The embodiment of the invention relates to the technical field of internet, in particular to a network quality evaluation method and a server.
Background
A Content Delivery Network (CDN) is an intelligent virtual Network built on the basis of an existing Network, and users can obtain required Content nearby by means of functional modules of load balancing, Content Delivery, scheduling and the like of a central platform by means of edge nodes deployed in various places, so that Network congestion is reduced, and user access response speed and hit rate are improved, wherein the edge nodes are generally deployed in units of machine rooms, and each machine room includes a plurality of edge servers.
For an edge node, the network quality is crucial to the quality of the service provided by the edge node, and therefore, it is generally required to test the network quality of the edge node first, for example, the network quality of the edge node may be tested by an IP polling detection means.
However, the inventors found that at least the following problems exist in the prior art: when testing the network quality of the edge node, a province or a large area is roughly defined as a service area according to the position of the edge node, and the service area used in the test may be different from the actual service area, which results in the inconsistency between the actual service quality of the edge node and the test result.
Disclosure of Invention
The embodiment of the invention aims to provide a network quality evaluation method and a server, which can more accurately simulate an actual application environment, so that the acquired network quality evaluation value of a node to be measured is more accurate.
In order to solve the above technical problem, an embodiment of the present invention provides a network quality evaluation method, including: acquiring a plurality of detection data obtained by sending detection requests to a plurality of areas by a node to be detected; obtaining test data of a plurality of quality levels according to the corresponding relation between the detection data of each area and a preset quality level and a detection data threshold; and obtaining the network quality evaluation value of the node to be tested according to the corresponding relation between the test data of the plurality of quality grades and the preset quality grade and weight parameter.
An embodiment of the present invention further provides a server, including: at least one processor; and a memory communicatively coupled to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the above-mentioned network quality assessment method.
Compared with the prior art, the method and the device for evaluating the network quality of the node to be tested can acquire the detection data from the node to be tested to each area, namely the network quality condition from the node to be tested to each area, can assemble the corresponding relation between the preset quality level and the detection data threshold value to obtain the test data representing the service state of the node to be tested at each quality level, and can acquire the network quality evaluation value of the node to be tested by combining the corresponding relation between the preset quality level and the weight parameter. The corresponding relation between the quality grade and the weight parameter can be set according to the requirement of the node to be detected so as to accurately judge whether the node to be detected meets the requirement; meanwhile, the network quality of the node to be detected is judged according to the actually detected detection data of each area, so that the actual application environment can be simulated more accurately, and the acquired network quality evaluation value of the node to be detected is more accurate.
In addition, the weighting parameter is an expected weighting, and the network quality evaluation value of the node to be tested is obtained according to the corresponding relation between the test data of the multiple quality grades and the preset quality grade and the weighting parameter, and the method comprises the following steps: and obtaining the network quality expected value of the node to be tested according to the corresponding relation between the test data of each quality grade and the expected weight. The embodiment provides a specific implementation manner for obtaining the network quality evaluation value of the node to be tested according to the corresponding relation between the test data of a plurality of quality levels and the preset quality levels and the weight parameters, wherein the weight parameters are selected as the expected weights, so that the calculated network quality expected value can accurately reflect the network quality established by the node to be tested and is used as the reference information for establishing the node.
In addition, the weighting parameter is an actual weighting, and the network quality evaluation value of the node to be tested is obtained according to the corresponding relationship between the test data of the multiple quality grades and the preset quality grade and weighting parameter, and the method includes the following steps: and obtaining the network quality actual value of the node to be tested according to the corresponding relation between the test data of each quality grade and the actual weight. The embodiment provides a specific implementation manner for obtaining the network quality evaluation value of the node to be tested according to the corresponding relation between the test data of a plurality of quality levels and the preset quality levels and the weight parameters, wherein the weight parameters are selected as actual weights, so that the calculated network quality expected value can accurately reflect the current network quality of the node to be tested.
In addition, each region comprises a plurality of test nodes, and the step of acquiring a plurality of detection data from the node to be tested to the plurality of regions comprises the following steps: acquiring detection data returned by each test node when receiving a detection request sent by a node to be detected; and for each area, obtaining the detection data of the area according to the detection data of a plurality of test nodes included in the area. The embodiment provides a specific implementation mode for acquiring a plurality of detection data from a node to be detected to a plurality of areas.
In addition, according to the corresponding relationship between the detection data of each region and the preset quality grade and the detection data threshold, obtaining the test data of a plurality of quality grades, including: acquiring the areas covered by the quality grades according to the corresponding relation between the detection data of each area and the quality grades and the detection data threshold; acquiring a normalized value of the detection data corresponding to each quality grade based on the detection data of the test node covered by each quality grade; the test data includes: the number of areas covered by each quality class, and the normalized value of each quality class. The embodiment provides a specific implementation mode for obtaining test data of a plurality of quality levels according to the detection data of each region and the preset quality level parameters.
In addition, the corresponding relation between the quality grade and the detection data threshold value is obtained by the following method: acquiring a plurality of detection data obtained by sending detection requests to a plurality of areas by each test node; and for each quality grade, acquiring the detection data of the test node covered by the quality grade, and acquiring the detection data threshold corresponding to the quality grade from the detection data covered by the quality grade according to the acquisition mode corresponding to the quality grade. In the present embodiment, the test nodes in the plurality of regions are used to obtain the detection data threshold corresponding to each quality class by applying the dynamic threshold rule, so that the correspondence between the quality class adapted to the network state and the detection data threshold can be obtained.
In addition, the acquiring the probe data of each area includes probe values of a plurality of test nodes included in the area, and the acquiring the normalized value of the probe data corresponding to each quality level based on the probe data of the area covered by each quality level includes: for each quality class, a normalized value of the quality class is calculated according to the following formula: sgWherein, S isgThe normalized values corresponding to the quality levels are represented, Smax represents the maximum value of the detection values of all the test nodes covered by the quality levels, Smin represents the minimum value of the detection values of all the test nodes covered by the quality levels, and Sver represents the average value of the detection values of all the test nodes covered by the quality levels. The embodiment provides a specific calculation formula for obtaining the normalization value of the detection data corresponding to each quality grade based on the detection data of the area covered by each quality grade.
In addition, the corresponding relationship between the quality level and the desired weight is obtained by: acquiring the customer flow of the load of each area in a preset service range; for each quality grade, acquiring customer traffic covered by the quality grade from the customer traffic of each area load according to the traffic range covered by the quality grade; and obtaining the expected weight corresponding to each quality grade according to the customer flow covered by each quality grade and the total customer flow of the service range load. The present embodiment provides a specific implementation manner of obtaining the correspondence between the quality level and the desired weight.
In addition, the corresponding relation between the quality grade and the actual weight is obtained by the following method: acquiring customer flow of each region loaded by a node to be tested; for each quality grade, acquiring customer flow covered by the quality grade from the customer flow loaded by the node to be tested according to the flow range covered by the quality grade; and obtaining the actual weight corresponding to each quality grade according to the client flow covered by each quality grade and the total client flow of the service range load. The present embodiment provides a specific implementation manner of obtaining the correspondence between the quality level and the actual weight.
In addition, the detection data comprises a plurality of performance indexes corresponding to the service mode of the node to be detected; obtaining test data of a plurality of quality levels according to the corresponding relation between the detection data of each area and the preset quality level and the detection data threshold, wherein the test data comprises the following steps: for each performance index, obtaining test data of each quality grade corresponding to the performance index according to the corresponding relation between the performance index and the quality grade of each area and the detection data threshold; obtaining a network quality evaluation value of the node to be tested according to the corresponding relation between the test data of the multiple quality grades and the preset quality grade and the weight parameter, wherein the method comprises the following steps: for each quality grade, obtaining evaluation values of a plurality of performance indexes according to the corresponding relation between the test data of the plurality of performance indexes corresponding to the quality grade and the weight parameter; and obtaining the network quality evaluation value of the node to be tested according to the evaluation values of the plurality of performance indexes and the preset weight distribution of the plurality of performance indexes. In the embodiment, the network quality evaluation value calculated by adopting a plurality of performance indexes is higher in matching degree with the service mode of the machine room to be tested, so that the network quality of the machine room to be tested in the service mode can be more accurately measured; if before the machine room to be tested falls to the ground, the network quality after the machine room falls to the ground can be more accurately reflected according to the network quality evaluation value, and whether the machine room can be built or not is conveniently judged.
In addition, obtaining a network quality evaluation value of the node to be tested according to the corresponding relationship between the test data of the multiple quality grades and the preset quality grade and the weight parameter, including: calculating to obtain the grade of the node to be measured according to the corresponding relation between the quantity of the area covered by each quality grade and the weight parameter; calculating to obtain an index score of the node to be measured according to the corresponding relation between the normalization value of each quality grade and the weight parameter; and obtaining the network quality evaluation value of the node to be tested according to the grade and index scores. The embodiment provides a specific implementation mode for obtaining the network quality evaluation value of the node to be measured according to the corresponding relation between the test data of a plurality of quality grades and the preset quality grade and the weight parameter, and the grade and index of the node to be measured are comprehensively considered, so that the obtained network quality evaluation value is more accurate.
In addition, according to the grade and the index, a network quality evaluation value of the node to be measured is obtained, which specifically comprises the following steps: calculating the product of the grade and a preset first weight, and taking the sum of the product of the index and a preset second weight as the network quality evaluation value of the node to be tested; the first weight is greater than the second weight. In this embodiment, setting the weight of the rank is greater than the weight of the index, which is equivalent to providing a hierarchical comparison mode, and taking the rank as a primary and the index as a secondary, further improves the accuracy of the obtained network quality assessment value, and better conforms to the actual application environment.
Drawings
One or more embodiments are illustrated by way of example in the accompanying drawings, which correspond to the figures in which like reference numerals refer to similar elements and which are not to scale unless otherwise specified.
Fig. 1 is a detailed flowchart of a network quality evaluation method according to a first embodiment of the present invention;
fig. 2 is a detailed flowchart of a network quality evaluation method according to a first embodiment of the present invention, in which steps 101 to 103 are respectively described in detail;
fig. 3 is a detailed flowchart of a network quality evaluation method according to a second embodiment of the present invention;
fig. 4 is a detailed flowchart of the manner of acquiring the correspondence between the quality level and the desired weight according to the second embodiment of the present invention;
fig. 5 is a detailed flowchart of the manner of acquiring the correspondence between the quality level and the actual weight according to the second embodiment of the present invention;
fig. 6 is a specific flowchart of the manner of acquiring the correspondence between the quality level and the detection data threshold value according to the third embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, it will be appreciated by those of ordinary skill in the art that numerous technical details are set forth in order to provide a better understanding of the present application in various embodiments of the present invention. However, the technical solution claimed in the present application can be implemented without these technical details and various changes and modifications based on the following embodiments.
The first embodiment of the present invention relates to a network quality evaluation method, which is applied to a server, that is, the server is used as a network quality evaluation platform to execute the network quality evaluation method, so that the network quality of an edge node in a CDN network can be evaluated. The CDN is divided into regions according to the positions of the edge nodes, and the divided regions can include cities, provinces, large areas and the like.
Fig. 1 shows a specific flow of the network quality evaluation method according to the present embodiment.
Referring to fig. 2, step 101 includes the following sub-steps:
in sub-step 1011, probe data returned by each test node when receiving a probe request sent by a node to be tested is obtained.
In sub-step 1012, for each region, probe data of the region is obtained according to the probe data of a plurality of test nodes included in the region.
Specifically, the following is specifically described with a machine room as a unit, that is, the node to be tested is the machine room to be tested, each area includes a plurality of test nodes, that is, includes a plurality of test machine rooms, and the machine room to be tested may include a plurality of edge servers. The machine room to be tested sends the same detection requests to the test machine rooms in the plurality of areas respectively and receives the detection data returned by the test machine rooms. For example, the machine room to be tested sends probe requests to the test machine rooms in each area according to the granularity of 5 minutes, 144 probe data returned by each test machine room can be obtained continuously for one day, the network quality evaluation platform obtains 144 probe data corresponding to each test machine room from the machine room to be tested, for each test machine room, noise data exceeding a preset abnormal threshold value can be removed, N probe data (N is an integer less than or equal to 144) are remained, then, from the N probe data, a median or a calculated average value is selected as the probe data of the day granularity of the test machine room, then, probe data of the day granularity of each test machine room can be obtained, for each area, the median or the calculated average value is selected as the probe data of the day granularity of the area from the probe data of the day granularity of the plurality of test machine rooms included in the area, namely the detection data from the machine room to be detected to the area. If the region is province, after the detection data of the granularity of each province day is obtained, the process can be repeated to obtain the detection data of the granularity of a large region day. It should be noted that the detected granularity is only schematically given in the above example, and the detected granularity is set according to the requirement.
In this example, the probe data may include only one performance index, but is not limited thereto, and the probe data may also include a plurality of performance indexes for the traffic pattern, in this case, for each area, probe data of one day granularity is obtained for each performance index, and the probe data of the area includes a plurality of performance index values. For example, when the service mode is an on-demand service, the main service index affecting the service quality is the response speed achievement rate, and the response speed achievement rate is the connection establishment time for the target performance index; for the edge computing service, the main performance indexes affecting the service quality are the packet loss rate and the connection establishing time.
It should be noted that a plurality of regions for testing may be set as required.
and a substep 1021, obtaining the region covered by each quality level according to the corresponding relation between the detection data and the quality level of each region and the detection data threshold.
Specifically, the following takes the number of quality levels as 4 as an example, including first to fourth quality levels, where each quality level corresponds to one detection data threshold, the detection data threshold corresponding to the first quality level is a local threshold, the detection data threshold corresponding to the second quality level is a local large-area threshold, the detection data threshold corresponding to the third quality level is a large-area-crossing threshold, and the detection data threshold corresponding to the fourth quality level may be similarly understood as a large-area-crossing threshold.
The network quality evaluation platform traverses the detection data from the machine room to be detected to each region, taking the detection data from the machine room to be detected to any region as an example, firstly, whether the detection data is smaller than a local threshold value is judged, if the detection data is smaller than the local threshold value, the region belongs to a first quality grade, namely, the first quality grade covers the region.
If the detection data is larger than the local threshold, whether the detection data is smaller than the local large-area threshold is judged, and if the detection data is smaller than the local large-area threshold, the area belongs to a second quality grade, namely the second quality grade covers the area.
If the detection data is larger than the local large-area threshold, whether the detection data is smaller than the large-area crossing threshold is judged, and if the detection data is smaller than the large-area crossing threshold, the area belongs to a third quality grade, namely the third quality grade covers the area.
If the detection data is larger than the threshold value of the large crossing area, the area belongs to a fourth quality level, namely the fourth quality level covers the area.
After traversing the machine room to be tested to the detection data of each area, the areas covered by each quality grade can be obtained, and the number of the areas covered by each quality grade can be obtained at the same time.
Taking as an example that each probe data includes only one performance index, the probe data of each area includes probe values (i.e., performance index values) of a plurality of test nodes under the area, and for each quality class, a normalized value of the quality class may be calculated according to the following formula:
Sg＝(Smax-Sver)/(Smax-Smin)
wherein S isgThe normalized values corresponding to the quality levels are represented, Smax represents the maximum value of the detection values of all the test nodes covered by the quality levels, Smin represents the minimum value of the detection values of all the test nodes covered by the quality levels, and Sver represents the average value of the detection values of all the test nodes covered by the quality levels.
After the normalization values of the first to fourth quality levels are all calculated, the test data of each quality level can be obtained, and taking the first quality level as an example, the test data of the quality level includes: the number of areas covered by the first quality level, and a normalized value of the first quality level.
In step 102, if the probe data includes a plurality of performance indexes corresponding to the service mode of the node to be tested, for each performance index, obtaining test data corresponding to the performance index of each quality grade according to the performance index and the quality grade parameter of each area; specifically, taking a plurality of performance indexes as a performance index 1 and a performance index 2 as an example, the test data of each quality level includes: the test data of the performance index 1 and the test data of the performance index 2 are calculated in the same manner as described above, and only the detection values of the corresponding performance indexes are used for calculation.
and a substep 1031, calculating to obtain grade levels of the nodes to be measured according to the corresponding relation between the quantity of the regions covered by each quality level and the quality levels and the weight parameters.
In particular, the method of manufacturing a semiconductor device,
where n denotes the number of quality classes, Ai denotes the number of covered regions of the ith quality class, and Bi denotes a weight parameter corresponding to the ith quality class.
And a substep 1032 of calculating to obtain an index score of the node to be measured according to the corresponding relation between the normalization value of each quality grade, the quality grade and the weight parameter.
In particular, the method of manufacturing a semiconductor device,
wherein n represents the number of quality grades, Ci represents the normalized value of the ith quality grade, and Bi represents the weight parameter corresponding to the ith quality grade.
And a substep 1032, obtaining the network quality evaluation value of the node to be measured according to the grade and index scores.
Specifically, the grade and the index have different weights, the grade and the index are multiplied by the corresponding weights, and the products of the two are added to obtain the network quality evaluation value of the node to be measured.
In an example, the first weight may be set to be greater than the second weight, which is equivalent to providing a hierarchical comparison mode, and the level is divided into a main part and an index is divided into an auxiliary part, so that the accuracy of the obtained network quality assessment value is further improved, and the method is more suitable for an actual application environment. For example, the first weight is set to be much larger than the second weight, that is, the rank is weighted and summed with the index score by the absolute weight to obtain the network quality evaluation value of the node to be measured; the first weight and the second weight can be set by combining the latitude of the machine room to be tested.
In this embodiment, if the probe data includes a plurality of performance indexes corresponding to the service mode of the node to be tested, for each quality level, an evaluation value of the plurality of performance indexes is obtained according to the test data of the plurality of performance indexes corresponding to the quality level; and obtaining the network quality evaluation value of the node to be tested according to the evaluation values of the plurality of performance indexes and the preset weight distribution of the plurality of performance indexes. Specifically, for each performance index, an evaluation value can be calculated according to the calculation methods in sub-steps 1031 to 1033, and then a network quality evaluation value of the node to be measured can be obtained according to the evaluation values of the multiple performance indexes, for example, a pareto analysis method is used for weighting and averaging to obtain the network quality evaluation value. Therefore, the network quality evaluation value calculated by adopting a plurality of performance indexes is higher in matching degree with the service mode of the machine room to be tested, and the network quality of the machine room to be tested in the service mode is more accurately measured; if before the machine room to be tested falls to the ground, the network quality after the machine room falls to the ground can be more accurately reflected according to the network quality evaluation value, and whether the machine room can be built or not is conveniently judged.
Compared with the first embodiment, in the present embodiment, first, the detection data from the node to be detected to each region is obtained, which is equivalent to obtaining the network quality status from the node to be detected to each region, then, the preset corresponding relationship between the quality level and the detection data threshold may be aggregated to obtain the test data representing the service state of the node to be detected at each quality level, and then, the preset corresponding relationship between the quality level and the weight parameter may be combined to obtain the network quality evaluation value of the node to be detected. The corresponding relation between the quality grade and the weight parameter can be set according to the requirement of the node to be detected so as to accurately judge whether the node to be detected meets the requirement; meanwhile, the network quality of the node to be detected is judged according to the actually detected detection data of each area, so that the actual application environment can be simulated more accurately, and the acquired network quality evaluation value of the node to be detected is more accurate.
A second embodiment of the present invention relates to a network quality evaluation method, and the present embodiment is mainly different from the first embodiment in that: the weight parameters include desired weights and/or actual weights.
In this embodiment, if the machine room to be tested is a machine room already in use by a construction point, the weight parameter may include an expected weight and/or an actual weight; if the machine room to be tested is the machine room without building points, the weight parameters only comprise expected weights. In this embodiment, a machine room to be tested is taken as a machine room already built and put into use, and the weight parameters include an expected weight and an actual weight. It should be noted that the weight parameter may include only one of the desired weight and the actual weight.
Fig. 3 shows a specific flow of the network quality evaluation method according to the present embodiment.
Step 201 and step 201 are substantially the same as step 101 and step 102, and are not described herein again, the main difference is that step 203 includes:
And a substep 2032 of obtaining an actual network quality value of the node to be tested according to the correspondence between the test data of each quality grade and the actual weight.
Specifically, according to the calculation methods in sub-steps 1031 to 1032 in the first embodiment, the network quality expected value and the network quality actual value of the node to be tested can be calculated according to the correspondence between the quality level and the expected weight and the correspondence between the quality level and the actual weight, so that the maintenance personnel can adjust the scheduling mode for the machine room to be tested to improve the quality of the service provided by the machine room to be tested when the network quality expected value is greater than the network quality actual value.
The following describes how to obtain the correspondence between the quality rank and the desired weight and the correspondence between the quality rank and the actual weight.
In this embodiment, please refer to fig. 4, which is a flowchart illustrating a manner of obtaining a corresponding relationship between a quality level and an expected weight.
Specifically, the preset service range includes a plurality of areas, and the plurality of areas can be areas for providing services for the machine room to be tested, and for all preset domain names with customer traffic, an edge server loading the customer traffic is obtained; for each test machine room, the sum of the customer traffic loaded by each edge server under the test machine room is the customer traffic loaded by the test machine room; similarly, the customer traffic loaded by the area is the sum of the customer traffic loaded by each test machine room in the area.
The customer flow loaded by each test machine room can be divided into the same-operator same-province-local customer flow, the same-operator same-large-area-same-large-area customer flow, the same-operator cross-large-area customer flow and different-operator-cross-operator customer flow according to the position relation between the geographical position of the test machine room and the actual service area. Similarly, the customer traffic of each regional load can be classified in this manner.
It should be noted that, when obtaining the customer traffic of each area load in the preset service range, the customer traffic of one day, one month, and one quarter may be taken according to a preset manner.
And step 302, for each quality level, acquiring the customer traffic covered by the quality level from the customer traffic of each area load according to the traffic range covered by the quality level.
Specifically, following the example in the first embodiment, the multiple quality classes include first to fourth quality classes, a traffic range covered by the first quality class is local customer traffic, a traffic range covered by the second quality class is same-area customer traffic, a traffic range covered by the third quality class is cross-area customer traffic, and a traffic range covered by the fourth quality class is cross-operator customer traffic.
The customer traffic covered by each quality class can be calculated by combining the customer traffic of each area load in step 301 according to the above correspondence.
And 303, obtaining the expected weight corresponding to each quality grade according to the customer flow covered by each quality grade and the total customer flow of the service range load.
Specifically, the client traffic covered by each quality class is divided by the total client traffic of the service area load, so that the desired weight corresponding to each quality class can be obtained.
It should be noted that, in this embodiment, the corresponding relationship between the quality level and the expected weight may be updated at regular intervals according to a preset period, but is not limited to this, and the corresponding relationship between the quality level and the expected weight may also be obtained only before the building point of the computer room.
In this embodiment, please refer to fig. 5, which is a flowchart illustrating a manner of obtaining a correspondence between quality levels and actual weights.
Specifically, the method includes the steps that customer flow of each edge server of a machine room to be tested loading each area in a preset time period is obtained, and according to the position relation between the geographical position of each edge server of the machine room to be tested and an actual service area, the customer flow of the same operator in the same province and local area, the customer flow of the same operator in the same area and the same area, the customer flow of the same operator in the cross-large area and the customer flow of different operators in the cross-operator area can be divided; namely, the customer traffic of each node to be tested is divided into the four types.
And 402, for each quality grade, acquiring the client flow covered by the quality grade from the client flow loaded by the node to be tested according to the flow range covered by the quality grade.
Specifically, following the example in the first embodiment, the multiple quality classes include first to fourth quality classes, a traffic range covered by the first quality class is local customer traffic, a traffic range covered by the second quality class is same-area customer traffic, a traffic range covered by the third quality class is cross-area customer traffic, and a traffic range covered by the fourth quality class is cross-operator customer traffic.
According to the above correspondence, in combination with the customer traffic of the node load to be measured obtained in step 401, the customer traffic covered by each quality level can be calculated respectively.
And step 403, obtaining the actual weight corresponding to each quality grade according to the customer flow covered by each quality grade and the total customer flow of the service range load.
Specifically, the actual weight corresponding to each quality class can be obtained by dividing the client traffic covered by each quality class by the total client traffic of the service range load.
It should be noted that, in this embodiment, the corresponding relationship between the quality level and the actual weight may be updated at regular intervals according to a preset period, so that the calculated actual value of the network quality is more accurate.
It should be further noted that, when the customer traffic of each area is counted in step 301 and step 401, the customer traffic having a special situation may be eliminated according to the requirement; for example, if a domain name requires a specific machine room in a certain area to provide service, which may cause an anomaly in the machine room and even the customer traffic of the area, the customer traffic of the domain name may be rejected.
Compared with the first embodiment, the present embodiment provides a specific implementation manner for obtaining the network quality assessment value of the node to be tested according to the corresponding relationship between the test data of a plurality of quality levels and the preset quality levels and the weight parameters.
A third embodiment of the present invention relates to a network quality evaluation method, and the present embodiment is mainly different from the first embodiment in that: and applying a dynamic threshold rule to obtain the corresponding relation between the quality grade and the detection data threshold.
A specific flow of the manner of acquiring the quality level and the detection data threshold in this embodiment is shown in fig. 6.
Specifically, each area includes a plurality of test rooms, and for each test room, the test room sends the same probe request to all the test rooms in the CDN network, and receives probe data returned by each test room. For example, the test machine room sends the probe requests to the test machine rooms in the areas according to the granularity of 5 minutes, 144 probe data returned by each test machine room can be obtained continuously for one day, and 144 probe data corresponding to other test machine rooms are obtained from each test machine room; for each test room, noise data exceeding a preset abnormal threshold value can be removed, N detection data (N is an integer less than or equal to 144) remain, then a median or a calculated mean is selected from the N detection data as detection data of the test room day granularity, and then the detection data of each test room day granularity can be obtained, and for each area, the median or the calculated mean is selected as detection data of the area day granularity from the detection data of the multiple test rooms included in the area, namely the detection data from the test room to the area. If the region is province, after the detection data of the granularity of each province day is obtained, the process can be repeated to obtain the detection data of the granularity of a large region day. It should be noted that the detected granularity is only schematically given in the above example, and the detected granularity is set according to the requirement.
Specifically, in continuation to the example in the first embodiment, the plurality of quality levels include first to fourth quality levels; selecting a maximum value from detection data of all machine rooms in the target province to the target province as a local gear threshold value, wherein the detection data of the target province accords with a preset same-province index threshold value, starting from a first quality level, the local gear threshold value is a detection data threshold value corresponding to the first quality level, then acquiring detection data of a test machine room of the same province of the same operator as the target province to the target province, searching a boundary value and a preset percentage grading value from the detection data, selecting a minimum value larger than the local gear threshold value from the boundary value and the grading value as a same-large-area gear threshold value, the same-large-area gear threshold value is a detection data threshold value corresponding to a second quality level, acquiring detection data of the test machine room of the same operator as the target province to the target province, and selecting the maximum value as a large-area gear threshold value from the detection data, the large-span gear threshold is a detection data threshold corresponding to the third quality level, and meanwhile, the large-span gear threshold can also be understood as a detection data threshold (a lower limit value) corresponding to the fourth quality level.
In the present embodiment, the corresponding relationship between the quality class and the probe data threshold value according to the network state can be acquired by acquiring the probe data threshold value corresponding to each quality class by applying the dynamic threshold rule using the test nodes in the plurality of areas, as compared to the first embodiment. The present embodiment can be modified from the second embodiment, and the same technical effects can be achieved.
A fourth embodiment of the present invention relates to a server, which is applied to a server and can be used as a network quality evaluation platform server to evaluate the network quality of an edge node in a CDN network.
In this embodiment, a server includes at least one processor; and a memory communicatively coupled to the at least one processor;
wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the network quality assessment method as in the first to third embodiments.
It will be understood by those of ordinary skill in the art that the foregoing embodiments are specific examples for carrying out the invention, and that various changes in form and details may be made therein without departing from the spirit and scope of the invention in practice.
Claims (13)
1. A method for evaluating network quality, comprising:
acquiring a plurality of detection data obtained by sending detection requests to a plurality of areas by a node to be detected;
obtaining test data of a plurality of quality levels according to the corresponding relation between the detection data of each area and a preset quality level and detection data threshold; the test data includes: the number of the areas covered by each quality grade and the normalization value of the detection data corresponding to each quality grade;
and obtaining the network quality evaluation value of the node to be tested according to the corresponding relation between the test data of the quality grades and the preset quality grade and weight parameter.
2. The method according to claim 1, wherein the weighting parameter is an expected weighting, and obtaining the network quality assessment value of the node to be tested according to the correspondence between the test data of the quality levels and the preset quality levels and the weighting parameter comprises:
and obtaining the network quality expected value of the node to be tested according to the corresponding relation between the test data of each quality grade and the expected weight.
3. The method according to claim 1, wherein the weighting parameter is an actual weighting, and obtaining the network quality assessment value of the node to be tested according to the correspondence between the test data of the quality levels and the preset quality levels and the weighting parameter comprises:
and obtaining the network quality actual value of the node to be tested according to the corresponding relation between the test data of each quality grade and the actual weight.
4. The method according to claim 1, wherein each of the regions includes a plurality of test nodes, and the obtaining a plurality of probe data of the node under test to the plurality of regions includes:
acquiring detection data returned by each test node when receiving a detection request sent by a node to be detected;
and for each region, obtaining the detection data of the region according to the detection data of the plurality of test nodes included in the region.
5. The method according to claim 4, wherein obtaining test data of a plurality of quality levels according to the correspondence between the probe data of each of the regions and the preset quality level and the probe data threshold comprises:
acquiring the areas covered by the quality grades according to the corresponding relation between the detection data of the areas and the quality grades and detection data thresholds;
and obtaining a normalized value of the detection data corresponding to each quality grade based on the detection data of the test node covered by each quality grade.
6. The method according to claim 5, wherein the corresponding relationship between the quality level and the threshold of the probing data is obtained by:
acquiring a plurality of detection data obtained by each test node sending a detection request to a plurality of regions;
and for each quality grade, acquiring the detection data of the test node covered by the quality grade, and acquiring the detection data threshold corresponding to the quality grade from the detection data covered by the quality grade according to the acquisition mode corresponding to the quality grade.
7. The method according to claim 5, wherein the obtaining the probe data corresponding to each quality class based on the probe data of the area covered by each quality class comprises:
for each of the quality classes, calculating a normalized value for the quality class according to the following formula:
Sg＝(Smax-Sver)/(Smax-Smin)
wherein S isgAnd the normalized values corresponding to the quality levels are represented, Smax represents the maximum value of the detection values of all the test nodes covered by the quality levels, Smin represents the minimum value of the detection values of all the test nodes covered by the quality levels, and Sver represents the average value of the detection values of all the test nodes covered by the quality levels.
8. The method according to claim 2, wherein the correspondence between the quality level and the desired weight is obtained by:
acquiring the customer flow of each regional load in a preset service range;
for each quality grade, acquiring customer traffic covered by the quality grade from the customer traffic of the loads in each area according to the traffic range covered by the quality grade;
and obtaining the expected weight corresponding to each quality grade according to the customer flow covered by each quality grade and the total customer flow of the service range load.
9. The method according to claim 3, wherein the correspondence between the quality level and the actual weight is obtained by:
acquiring customer flow of each region loaded by the node to be tested;
for each quality grade, acquiring customer traffic covered by the quality grade from the customer traffic loaded by the node to be tested according to the traffic range covered by the quality grade;
and obtaining the actual weight corresponding to each quality grade according to the customer flow covered by each quality grade and the total customer flow of the preset service range load.
10. The method according to claim 1, wherein the probe data includes a plurality of performance indicators corresponding to the service mode of the node to be tested; the obtaining of the test data of a plurality of quality levels according to the corresponding relationship between the detection data of each of the regions and the preset quality level and detection data threshold includes:
for each performance index, obtaining test data of each quality grade corresponding to the performance index according to the corresponding relation between the performance index of each region and the quality grade and a detection data threshold;
the obtaining of the network quality evaluation value of the node to be tested according to the corresponding relationship between the test data of the multiple quality levels and the preset quality levels and the weight parameters includes:
for each quality grade, obtaining evaluation values of a plurality of performance indexes according to the corresponding relation between the test data of the performance indexes corresponding to the quality grade and the weight parameter;
and obtaining the network quality evaluation value of the node to be tested according to the evaluation values of the plurality of performance indexes and the weight distribution of the plurality of preset performance indexes.
11. The method according to claim 1, wherein obtaining the network quality assessment value of the node to be tested according to the correspondence between the test data of the multiple quality levels and preset quality levels and weight parameters comprises:
calculating to obtain the grade of the node to be measured according to the corresponding relation between the quantity of the area covered by each quality grade and the weight parameter;
calculating to obtain an index score of the node to be measured according to the corresponding relation between the normalization value of each quality grade and the weight parameter;
and obtaining the network quality evaluation value of the node to be tested according to the grade and the index.
12. The method for evaluating network quality according to claim 11, wherein the obtaining of the network quality evaluation value of the node to be tested according to the grade and the index score specifically includes:
calculating the product of the grade and a preset first weight, and taking the sum of the product of the index and a preset second weight as the network quality evaluation value of the node to be tested; the first weight is greater than the second weight.
13. A server, comprising: at least one processor; and a memory communicatively coupled to the at least one processor;
wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of assessing network quality of any one of claims 1 to 12.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911192499.XA CN111010312B (en) | 2019-11-28 | 2019-11-28 | Network quality evaluation method and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911192499.XA CN111010312B (en) | 2019-11-28 | 2019-11-28 | Network quality evaluation method and server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111010312A CN111010312A (en) | 2020-04-14 |
| CN111010312B true CN111010312B (en) | 2021-11-19 |
Family
ID=70112234
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911192499.XA Active CN111010312B (en) | 2019-11-28 | 2019-11-28 | Network quality evaluation method and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111010312B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112953766B (en) * | 2021-02-02 | 2022-03-15 | 成都卓源网络科技有限公司 | Network comprehensive safety assessment method for active and passive acquisition of distributed probes |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101330709A (en) * | 2008-06-12 | 2008-12-24 | ***通信集团设计院有限公司 | Method and device for determining data business service quality |
| CN108154300A (en) * | 2017-12-25 | 2018-06-12 | 东软集团股份有限公司 | Point of interest site selecting method, device and computer equipment |
| CN108306769A (en) * | 2018-01-31 | 2018-07-20 | 北京奇艺世纪科技有限公司 | A kind of the node deployment control method and system of CDN |
| CN110418350A (en) * | 2018-04-27 | 2019-11-05 | 中兴网鲲信息科技（上海）有限公司 | A kind of shared site appraisal procedure and system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101984624A (en) * | 2010-11-03 | 2011-03-09 | 蓝汛网络科技（北京）有限公司 | Method and device for distributing network flow |
| CN104683182B (en) * | 2015-02-15 | 2018-04-10 | 上海帝联信息科技股份有限公司 | IDC network of computer room service quality detection method and device |
| US10432486B2 (en) * | 2017-01-06 | 2019-10-01 | Mz Ip Holdings, Llc | System and method for updating application clients using a plurality of content delivery networks |
| CN107707378B (en) * | 2017-06-29 | 2018-11-13 | 贵州白山云科技有限公司 | A kind of CDN covering scheme generation method and device |
-
2019
- 2019-11-28 CN CN201911192499.XA patent/CN111010312B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101330709A (en) * | 2008-06-12 | 2008-12-24 | ***通信集团设计院有限公司 | Method and device for determining data business service quality |
| CN108154300A (en) * | 2017-12-25 | 2018-06-12 | 东软集团股份有限公司 | Point of interest site selecting method, device and computer equipment |
| CN108306769A (en) * | 2018-01-31 | 2018-07-20 | 北京奇艺世纪科技有限公司 | A kind of the node deployment control method and system of CDN |
| CN110418350A (en) * | 2018-04-27 | 2019-11-05 | 中兴网鲲信息科技（上海）有限公司 | A kind of shared site appraisal procedure and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111010312A (en) | 2020-04-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022111327A1 (en) | Risk level data processing method and apparatus, and storage medium and electronic device | |
| CN110149237B (en) | Hadoop platform computing node load prediction method | |
| Hu et al. | Web service recommendation based on time series forecasting and collaborative filtering | |
| CN107026750B (en) | User Internet QoE evaluation method and device | |
| WO2018161447A1 (en) | Protection method and system for cdn client source station | |
| CN106535204B (en) | A kind of service coverage method for evaluating quality and device | |
| CN104683182B (en) | IDC network of computer room service quality detection method and device | |
| CN107547269B (en) | Method for constructing intelligent substation communication flow threshold model based on FARIMA | |
| CN109241343A (en) | A kind of brush amount user identifying system, method and device | |
| CN109583926A (en) | A kind of used car pricing method apparatus and system | |
| Tigkas et al. | Comparative study of evolutionary algorithms for the automatic calibration of the Medbasin-D conceptual hydrological model | |
| CN105227689A (en) | Based on the Target IP location algorithm of local time delay distribution similarity tolerance | |
| CN108768942A (en) | A kind of ddos attack detection method and detection device based on adaptive threshold | |
| CN108390775A (en) | A kind of user experience quality evaluation method and system based on SPICE | |
| CN111010312B (en) | Network quality evaluation method and server | |
| CN113988441A (en) | Power wireless network link quality prediction and model training method and device | |
| Hagen-Zanker et al. | Map comparison methods for comprehensive assessment of geosimulation models | |
| Paul et al. | Characterizing internet access and quality inequities in california m-lab measurements | |
| CN113987503A (en) | Regional level network survivability difference evaluation method and device based on topology analysis | |
| CN109525455A (en) | Hydrological real-time monitoring network state comprehensive evaluation method | |
| Willis | Load forecasting for distribution planning-error and impact on design | |
| Bektemyssova et al. | Time series forecasting by the arima method | |
| CN108521435B (en) | Method and system for user network behavior portrayal | |
| CN116777202A (en) | Power distribution network line risk assessment method, device, equipment and readable storage medium | |
| CN111242425A (en) | Power data processing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |