WO2019144560A1 - 一种检测网络质量的方法和*** - Google Patents

一种检测网络质量的方法和*** Download PDF

Info

Publication number
WO2019144560A1
WO2019144560A1 PCT/CN2018/091204 CN2018091204W WO2019144560A1 WO 2019144560 A1 WO2019144560 A1 WO 2019144560A1 CN 2018091204 W CN2018091204 W CN 2018091204W WO 2019144560 A1 WO2019144560 A1 WO 2019144560A1
Authority
WO
WIPO (PCT)
Prior art keywords
node
network
network speed
data processing
processing device
Prior art date
Application number
PCT/CN2018/091204
Other languages
English (en)
French (fr)
Inventor
黄梅红
郑文丽
Original Assignee
网宿科技股份有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 网宿科技股份有限公司 filed Critical 网宿科技股份有限公司
Priority to US16/463,212 priority Critical patent/US11108676B2/en
Priority to EP18893320.4A priority patent/EP3550768B1/en
Publication of WO2019144560A1 publication Critical patent/WO2019144560A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/142—Network analysis or design using statistical or mathematical methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5003—Managing SLA; Interaction between SLA and QoS
    • H04L41/5009—Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
    • H04L41/5012—Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF] determining service availability, e.g. which services are available at a certain point in time
    • H04L41/5016—Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF] determining service availability, e.g. which services are available at a certain point in time based on statistics of service availability, e.g. in percentage or over a given time
    • 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
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0876—Network utilisation, e.g. volume of load or congestion level
    • H04L43/0894—Packet rate
    • 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
    • H04L43/14—Arrangements for monitoring or testing data switching networks using software, i.e. software packages
    • 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
    • H04L43/50—Testing arrangements
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/06—Protocols specially adapted for file transfer, e.g. file transfer protocol [FTP]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/104—Peer-to-peer [P2P] networks
    • H04L67/1074—Peer-to-peer [P2P] networks for supporting data block transmission mechanisms
    • H04L67/1078—Resource delivery mechanisms

Definitions

  • the present invention relates to the field of data transmission technologies, and in particular, to a method and system for detecting network quality.
  • CDN Content Delivery Network
  • the terminal may send a resource acquisition request to a certain network portal of the CDN service cluster through the terminal.
  • the CDN service cluster may detect the network quality of the multiple transmission links between the network portal and the node by using a network detection tool preset in the network portal. The CDN service cluster can then select a transmission link with the best network quality to transmit data between the terminal and the node.
  • embodiments of the present invention provide a method and system for detecting network quality.
  • the technical solution is as follows:
  • a method of detecting network quality comprising:
  • the data processing device acquires a file download rate of the first node in the target period in the preset number of historical statistical periods for the file on the second node;
  • the data processing device Determining, by the data processing device, a plurality of network quality detection parameters from the first node to the second node in the target time period according to the file download rate, where the multiple network quality detection parameters include a conventional network Speed, network speed change range and network speed lower limit;
  • the central scheduling device detects a network quality from the first node to the second node in the target time period based on the network fluctuation model.
  • the data processing device acquires a file downloading rate of the first node for the second node in the target time period in the preset number of historical statistical periods, including:
  • the central scheduling device randomly selects at least one device in a device group of each node, and sends a preset test file to at least one device of each node for storage;
  • At least one device of the second node periodically sends a download trigger request of the test file to at least one device of the first node, such that at least one device of the first node periodically periodically from at least one device of the second node Download the test file;
  • the data processing device acquires, by the at least one device of the first node in the target time period in the preset number of historical statistical periods, the file download rate of the test file is downloaded from the at least one device of the second node.
  • the data processing device determines, according to the file downloading rate, a plurality of network quality detection parameters from the first node to the second node in the target time period, including:
  • the data processing device determines, as the median, quartile, and minimum values of the file download rates of the files on the second node by the first node in the target time period in the preset number of historical statistical periods, respectively The normal network speed, the network speed change range, and the network speed lower limit from the first node to the second node in the target period.
  • the data processing device creates a network fluctuation model from the first node to the second node in the target time period according to the multiple network quality detection parameters and respective preset weights, including:
  • the data processing device determines a maximum normal network speed, a maximum network speed change range, and a maximum network speed lower limit from any node to the second node in the target period;
  • the data processing device normalizes the plurality of network quality detection parameters based on the maximum normal network speed, the maximum network speed change amplitude, and the highest network speed lower limit;
  • the data processing device according to the normalized plurality of network quality detection parameters and respective preset weights, and the file download success rate of any node in the target period in the preset number of historical statistical periods for the second node Establishing a network fluctuation model from the first node to the second node in the target period.
  • the data processing device creates a network fluctuation model from the first node to the second node in the target time period according to the multiple network quality detection parameters and respective preset weights, including:
  • the data processing device determines a maximum normal network speed, a maximum network speed change range, and a maximum network speed lower limit from the first node to the arbitrary node in the target period;
  • the data processing device normalizes the plurality of network quality detection parameters based on the maximum normal network speed, the maximum network speed change amplitude, and the highest network speed lower limit;
  • the data processing device according to the normalized multiple network quality detection parameters and respective preset weights, and the file download success rate of the first node for any node in the target time period in a preset number of historical statistical periods Establishing a network fluctuation model from the first node to the second node in the target period.
  • a system for detecting network quality comprising a data processing device, a central scheduling device, and a plurality of nodes including a first node and a second node, wherein:
  • the data processing device is configured to acquire a file download rate of the file sent by the first node to the second node in a target time period in a preset number of historical statistical periods, and determine the target time period according to the file download rate.
  • a plurality of network quality detection parameters from the first node to the second node, where the plurality of network quality detection parameters include a conventional network speed, a network speed change amplitude, and a network speed lower limit, according to the multiple networks Generating, by the quality detection parameter and the respective preset weights, a network fluctuation model from the first node to the second node in the target time period, and providing the network fluctuation model to the central scheduling device;
  • the central scheduling device is configured to detect, according to the network fluctuation model, a network quality from the first node to the second node in the target time period.
  • the central scheduling device is further configured to randomly select at least one device in a device group of each node, and send a preset test file to at least one device of each node for storage;
  • At least one device of the second node periodically sends a download trigger request of the test file to at least one device of the first node, so that at least one device of the first node periodically periodically from at least one of the second nodes Download the test file at the device;
  • the data processing device is configured to acquire, by the at least one device of the first node in the target time period in the preset number of historical statistical periods, the file download rate of the test file from the at least one device of the second node.
  • the data processing device is specifically configured to:
  • Determining, in the target period, the median, quartile, and minimum values of all file download rates of the file on the second node by the first node in the target time period in the preset number of historical statistical periods The normal network speed, the network speed change range, and the network speed lower limit from the first node to the second node.
  • the data processing device is specifically configured to:
  • the data processing device is specifically configured to:
  • the data processing device acquires a file downloading rate of the first node in the target period in the preset number of historical statistical periods for the file on the second node; and the data processing device determines the target period from the first node according to the file downloading rate.
  • the data processing device according to the plurality of network quality detection parameters and respective preset weights Creating a network fluctuation model from the first node to the second node in the target period, and providing the network fluctuation model to the central scheduling device; the central scheduling device detecting the network from the first node to the second node in the target period based on the network fluctuation model quality.
  • the network fluctuation model can be created by the file download between nodes, and then the network quality model can be used to detect the network quality between nodes.
  • No network detection tool can be set, which can reduce the detection cost of network quality and relatively quantitative analysis. Measuring network quality provides an intuitive and quantitative data foundation for historical analysis, forecasting, and scheduling optimization of network quality.
  • FIG. 1 is a schematic diagram of a network framework for detecting network quality according to an embodiment of the present invention
  • FIG. 2 is a flowchart of a method for detecting network quality according to an embodiment of the present invention.
  • the embodiment of the invention provides a method for detecting network quality, which can be implemented by a data processing device in a CDN service cluster and a central scheduling device, and is implemented by a node in a CDN service cluster.
  • the specific network framework is shown in FIG. 1 . Shown.
  • the data processing device can be used to collect service execution data between nodes in the CDN service cluster, and analyze and organize the collected service execution data, and the central scheduling device can process the service execution data based on the data processing device.
  • the CDN service cluster is uniformly scheduled and managed.
  • the foregoing network device may include a processor, a memory, and a transceiver.
  • the processor may be used to perform processing for detecting network quality in the following process, and the memory may be used to store data and data generated in the following processing.
  • the transceiver can be used to receive and transmit relevant data in the following processing. It can be understood that the functions of the above data processing device and the central scheduling device can also be implemented by multiple components in the same network device. In this embodiment, the data processing device and the central scheduling device are described as independent network devices, and other situations are similar, and will not be further described.
  • Step 201 The data processing device acquires a file download rate of the file on the second node by the first node in the target time period in a preset number of historical statistical periods.
  • the statistical period may be manually set, and the period of the statistical file download rate may be 1 day. Each statistical period may be divided into multiple time periods, and each time period may be 1 hour. Both the first node and the second node may be any one of the CDN service clusters.
  • the data processing device can monitor the file transmission between the nodes in the CDN service cluster in real time. If a node downloads the file from another node, the data processing device can calculate the corresponding file download rate and file download rate. r can be equal to the file size divided by the file download time.
  • the data processing device may acquire a preset number of historical statistical periods, and each node downloads files from other nodes in the target time period. File download rate. The subsequent description takes the statistical period as 1 day and the target period as 1 hour as an example. Other situations are similar.
  • the data processing device acquires a preset number of days (such as k days).
  • a dedicated test file is set on each node, and the data processing device can only obtain the file download rate of the node for the test file.
  • the processing of step 201 can be as follows: the central scheduling device is in the device group of each node.
  • At least one device randomly sending a preset test file to at least one device of each node for storage; at least one device of the second node periodically sending a download trigger of the test file to at least one device of the first node Requesting, so that at least one device of the first node periodically downloads a test file from at least one device of the second node; the data processing device acquires at least one of the first nodes in the target time period within a preset number of historical statistical periods The device downloads a file download rate of the test file from at least one device of the second node.
  • each node in the CDN service cluster may be a device group composed of multiple devices, and the central scheduling device may randomly select at least one device in each device group of the node, and then send the preset test file to In the above at least one device, therefore, an equal amount of devices can be stored at each node to store the preset test file.
  • the file size of the above test file can be about 2M, which can avoid the impact of node performance when the test file is downloaded because the file is too large, or the file download rate is inaccurate because the file is too small.
  • each device storing the test file can periodically send a download trigger request for the test file to each of the selected devices of the other node to cause the devices to periodically download the test file from the device.
  • a device selected at the j-node can periodically send (for example, every 10 minutes) a download trigger request of the test file to all selected devices at the i-node, so that the device of the i-node Periodically download test files from the device.
  • the data processing device can acquire the file download rate of the test file downloaded from the device of the j-node by the device of the i-node in the h-th hour of k days.
  • Step 202 The data processing device determines, according to the file download rate, a plurality of network quality detection parameters from the first node to the second node in the target time period.
  • the plurality of network quality detection parameters include a conventional network speed, a network speed change range, and a network speed lower limit.
  • the data processing device can determine the i-node to the j-node in the h-th hour of each cycle according to the file download rate.
  • the conventional network speed x (h, i, j) is the network speed when the network between the two nodes is normal;
  • the network speed variation amplitude y (h, i, j) reflects the network stability between the two nodes;
  • the lower limit z (h, i, j) is the network speed when the network quality between the two nodes is extremely poor.
  • the process for determining the network quality detection parameter may be specifically as follows: the data processing device will preset the median value of the file download rate of the first node to the file on the second node in the target period, The division distance and the minimum value are respectively determined as the normal network speed, the network speed change range, and the network speed lower limit from the first node to the second node in the target period.
  • the file downloading rate can be summarized and sorted, and then the median value of all file download rates can be obtained.
  • the interquartile range and the minimum value are respectively determined as the normal network speed x (h, i, j) from the first node to the second node in the target period, the network speed variation range y (h, i, j) and the network The lower limit z (h, i, j) .
  • r (n, h, i, j) represents the file download rate of the file on the j node for the nth day and the hth hour
  • the k-day data is processed by the i node:
  • n (N-k+1, N-k+2,...,N).
  • Step 203 The data processing device creates a network fluctuation model from the first node to the second node in the target time period according to the plurality of network quality detection parameters and respective preset weights, and provides the network fluctuation model to the central scheduling device.
  • the technician can set respective preset weights for multiple network quality detection parameters according to the network quality evaluation criteria of the CDN service cluster, and the sum of all preset weights is a fixed value.
  • the network quality evaluation standard is that the normal network speed x (h, i, j) preset weight is the better the network speed is better, the network can have certain volatility, and the network speed is not considered when the network is extremely poor.
  • a is higher, the preset weight b of the network speed change range y (h, i, j) is lower, and the preset weight c of the network speed lower limit z (h, i, j) is extremely low.
  • the data processing device can create the hth according to the plurality of network quality detection parameters and the respective preset weights.
  • the network fluctuation model f(h, i, j) g(ax, by, cz) from the i-node to the j-node in the hour.
  • the network quality detection parameters may be normalized from the i dimension, and then the network fluctuation model is created.
  • the processing of step 203 may be specifically as follows: the data processing device determines the target node from any node. Maximum conventional network speed to the second node, maximum network speed change range and maximum network speed lower limit; data processing equipment normalizes multiple network quality detection parameters based on maximum normal network speed, maximum network speed change range, and maximum network speed lower limit The data processing device successfully downloads the file of the file on the second node according to the normalized multiple network quality detection parameters and the respective preset weights, and the target period in the preset number of historical statistical periods Rate, creating a network fluctuation model from the first node to the second node in the target period.
  • the data processing device may first obtain a file download rate of any node in the h hour in the k-day for the file on the j node, and then determine a normal network speed from each node to the j node in the target time period based on the processing of step 202. , the network speed change range and the network speed lower limit, and then determine the maximum conventional network speed in all the conventional network speed, the network speed change range and the network speed lower limit Maximum speed of change And the maximum speed limit Further, the data processing device may normalize the plurality of network quality detection parameters from the i-node to the j-node in the h-th period based on the maximum normal network speed, the maximum network speed change amplitude, and the highest network speed lower limit:
  • the data processing device can detect the parameters X (h, i, j) , Y (h, i, j) , Z (h, i, j) and respective presets according to the normalized multiple network quality.
  • the weights a, b, c, and the above file download success rate create a network fluctuation model from the i node to the j node in the hth period:
  • the network quality detection parameters may be normalized from the j dimension, and then the network fluctuation model is created.
  • the processing of step 203 may be specifically as follows: the data processing device determines the target time period from the first Maximum normal network speed, maximum network speed change range and maximum network speed lower limit from one node to any node; data processing equipment categorizes multiple network quality detection parameters based on maximum normal network speed, maximum network speed change range and maximum network speed lower limit a processing process; the data processing device according to the normalized multiple network quality detection parameters and respective preset weights, and the file download success rate of the first node for any node in the target time period in a preset number of historical statistical periods , creating a network fluctuation model from the first node to the second node in the target period.
  • the two network fluctuation models created based on different i-dimensions and j-dimension normalization processes can be used to detect the network quality from the first node to the second node, and the specific application scenarios may be different.
  • the normalization process based on the i dimension can obtain the network quality of the file covering the other nodes on the j node; the normalization process from the j dimension can obtain the network quality of the file downloaded by the i node on other nodes.
  • the network fluctuation model created after the i-dimension normalization process can be used to detect the network quality between the nodes to determine from which node to download the file;
  • the service scope of the nodes is based on the network fluctuation model created after the j-dimension normalization process, the network quality between the nodes is detected to determine which nodes the node can provide the file download service.
  • Step 204 The central scheduling device detects the network quality from the first node to the second node in the target time period based on the network fluctuation model.
  • the network quality detection processing between any two nodes in the CDN service cluster in any period of time can be realized, and after the processing of steps 201 to 204 is performed multiple times, it can be effective.
  • the network quality between all nodes of the CDN service cluster in all time periods is detected.
  • the central scheduling device may also perform multiple applications based on the network fluctuation model between the nodes, which may be as follows:
  • the central scheduling device can also detect the service cost of the third node and the fourth node, select the node with low service cost, or use the node with lower service cost as the standard. Nodes with higher service costs are configured to reduce the total service cost of the CDN service cluster.
  • the central scheduling device may determine the network quality from other nodes to the node in each period of a statistical period based on the network fluctuation model between the nodes, and then adjust the node in each period according to the situation.
  • the scope of services that is, which nodes are provided with file download services at different time periods.
  • the central scheduling device can determine the network quality from the node to other nodes in each period of a statistical period based on the network fluctuation model between the nodes, and then can adjust the node in each period according to the target.
  • the file download range which determines which nodes to download files from at each time.
  • the nodes can be classified according to preset criteria, such as according to regional classification, or according to network operator classification, and then the network fluctuation model between nodes can be used to judge the network quality between the classifications as a whole, such as detecting nodes and Network quality between regions to detect network quality between different network operators.
  • the network state between the two nodes in the previous period of the current statistical period ie, the value of f(h, i, j)
  • the network status is used to predict the network status between the two nodes at the current time of the current cycle.
  • the data processing device obtains a file download rate of the file on the second node in the target time period in the preset number of historical statistical periods; and the data processing device determines the target time period from the first according to the file download rate.
  • a plurality of network quality detection parameters from the node to the second node, wherein the plurality of network quality detection parameters include a conventional network speed, a network speed change amplitude, and a network speed lower limit; and the data processing device detects the plurality of network quality detection parameters and respective presets Weighting, creating a network fluctuation model from the first node to the second node in the target period, and providing the network fluctuation model to the central scheduling device; the central scheduling device detecting the target period from the first node to the second node based on the network fluctuation model Network quality.
  • the network fluctuation model can be created by the file download between nodes, and then the network quality model can be used to detect the network quality between nodes.
  • No network detection tool can be set, which can reduce the detection cost of network quality and relatively quantitative analysis. Measuring network quality provides an intuitive and quantitative data foundation for historical analysis, forecasting, and scheduling optimization of network quality.
  • an embodiment of the present invention further provides a system for detecting network quality, where the system includes a data processing device, a central scheduling device, and a plurality of nodes including a first node and a second node, where:
  • the data processing device is configured to acquire a file download rate of the file sent by the first node to the second node in a target time period in a preset number of historical statistical periods, and determine the target time period according to the file download rate.
  • a plurality of network quality detection parameters from the first node to the second node, where the plurality of network quality detection parameters include a conventional network speed, a network speed change amplitude, and a network speed lower limit, according to the multiple networks Generating, by the quality detection parameter and the respective preset weights, a network fluctuation model from the first node to the second node in the target time period, and providing the network fluctuation model to the central scheduling device;
  • the central scheduling device is configured to detect, according to the network fluctuation model, a network quality from the first node to the second node in the target time period.
  • the central scheduling device is further configured to randomly select at least one device in a device group of each node, and send a preset test file to at least one device of each node for storage;
  • At least one device of the second node periodically sends a download trigger request of the test file to at least one device of the first node, so that at least one device of the first node periodically periodically from at least one of the second nodes Download the test file at the device;
  • the data processing device is configured to acquire, by the at least one device of the first node in the target time period in the preset number of historical statistical periods, the file download rate of the test file from the at least one device of the second node.
  • the data processing device is specifically configured to:
  • Determining, in the target period, the median, quartile, and minimum values of all file download rates of the file on the second node by the first node in the target time period in the preset number of historical statistical periods The normal network speed, the network speed change range, and the network speed lower limit from the first node to the second node.
  • the data processing device is specifically configured to:
  • the data processing device is specifically configured to:
  • the data processing device acquires a file downloading rate of the first node in the target period in the preset number of historical statistical periods for the file on the second node; and the data processing device determines the target period from the first node according to the file downloading rate.
  • the data processing device according to the plurality of network quality detection parameters and respective preset weights Creating a network fluctuation model from the first node to the second node in the target period, and providing the network fluctuation model to the central scheduling device; the central scheduling device detecting the network from the first node to the second node in the target period based on the network fluctuation model quality.
  • the network fluctuation model can be created by the file download between nodes, and then the network quality model can be used to detect the network quality between nodes.
  • No network detection tool can be set, which can reduce the detection cost of network quality and relatively quantitative analysis. Measuring network quality provides an intuitive and quantitative data foundation for historical analysis, forecasting, and scheduling optimization of network quality.
  • a person skilled in the art may understand that all or part of the steps of implementing the above embodiments may be completed by hardware, or may be instructed by a program to execute related hardware, and the program may be stored in a computer readable storage medium.
  • the storage medium mentioned may be a read only memory, a magnetic disk or an optical disk or the like.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Environmental & Geological Engineering (AREA)
  • Probability & Statistics with Applications (AREA)
  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Algebra (AREA)
  • Computer Security & Cryptography (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Small-Scale Networks (AREA)

Abstract

本发明公开了一种检测网络质量的方法和***，属于数据传输技术领域。所述方法包括：数据处理设备获取预设数目个历史统计周期内目标时段中第一节点对于第二节点上文件的文件下载速率；数据处理设备根据文件下载速率确定目标时段中从第一节点到第二节点的多个网络质量检测参数，其中，多个网络质量检测参数包括常规网速、网速变化幅度和网速下限；数据处理设备根据多个网络质量检测参数和各自的预设权重，创建目标时段中从第一节点到第二节点的网络波动模型，并将网络波动模型提供给中心调度设备；中心调度设备基于网络波动模型检测目标时段中从第一节点到第二节点的网络质量。采用本发明，可以降低检测网络质量的成本。

Description

一种检测网络质量的方法和*** 技术领域
本发明涉及数据传输技术领域，特别涉及一种检测网络质量的方法和***。
背景技术
随着互联网技术的不断进步，CDN(内容分发网络，Content Delivery Network)服务也随之快速发展。CDN服务集群中的节点可以存储有大量的数据资源，用户可以以较短的时间，就近获取其中存储的数据资源。
当用户需要通过终端从某个节点处获取数据资源时，可以通过终端向CDN服务集群的某个网络入口发送资源获取请求。CDN服务集群在接收到资源获取请求后，可以通过预先设置在该网络入口的网络检测工具，检测该网络入口至上述节点间多条传输链路的网络质量。然后CDN服务集群可以选择网络质量最佳的一条传输链路来传输终端和节点之间数据。
在实现本发明的过程中，发明人发现现有技术至少存在以下问题：
CDN服务集群的规模持续扩大，节点的分布越来越广泛，网络入口的数量也大幅增加，为了保证传输链路的网络质量的有效检测，技术人员需要在所有的网络入口处均设置网络检测工具，这样，网络质量的检测成本较高。
发明内容
为了解决现有技术的问题，本发明实施例提供了一种检测网络质量的方法和***。所述技术方案如下：
第一方面，提供了一种检测网络质量的方法，所述方法包括：
数据处理设备获取预设数目个历史统计周期内目标时段中第一节点对于第二节点上文件的文件下载速率；
所述数据处理设备根据所述文件下载速率确定所述目标时段中从所述第一节点到所述第二节点的多个网络质量检测参数，其中，所述多个网络质量检测参数包括常规网速、网速变化幅度和网速下限；
所述数据处理设备根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，并将所述网络波动模型提供给中心调度设备；
所述中心调度设备基于所述网络波动模型检测所述目标时段中从所述第一节点到所述第二节点的网络质量。
可选的，所述数据处理设备获取预设数目个历史统计周期内目标时段中第一节点对于第二节点的文件下载速率，包括：
所述中心调度设备在每个节点的设备组中随机选择至少一个设备，将预设的测试文件发送至所述每个节点的至少一个设备中进行存储；
第二节点的至少一个设备周期性向第一节点的至少一个设备发送所述测试文件的下载触发请求，以使所述第一节点的至少一个设备周期性从所述第二节点的至少一个设备处下载所述测试文件；
所述数据处理设备获取预设数目个历史统计周期内目标时段中所述第一节点的至少一个设备从所述第二节点的至少一个设备处下载所述测试文件的文件下载速率。
可选的，所述数据处理设备根据所述文件下载速率确定所述目标时段中从所述第一节点到所述第二节点的多个网络质量检测参数，包括：
所述数据处理设备将预设数目个历史统计周期内目标时段中所述第一节点对于所述第二节点上文件的所有文件下载速率的中值、四分位距和最小值，分别确定为所述目标时段中从所述第一节点到所述第二节点的常规网速、网速变化幅度和网速下限。
可选的，所述数据处理设备根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，包括：
所述数据处理设备确定所述目标时段中从任意节点到所述第二节点的最大常规网速、最大网速变化幅度和最高网速下限；
所述数据处理设备基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
所述数据处理设备根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中任意节点对于所述第二节 点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
可选的，所述数据处理设备根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，包括：
所述数据处理设备确定所述目标时段中从第一节点到所述任意节点的最大常规网速、最大网速变化幅度和最高网速下限；
所述数据处理设备基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
所述数据处理设备根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中所述第一节点对于任意节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
第二方面，提供了一种检测网络质量的***，所述***包括数据处理设备、中心调度设备和包含第一节点与第二节点在内的多个节点，其中：
所述数据处理设备，用于获取预设数目个历史统计周期内目标时段中所述第一节点对于所述第二节点上文件的文件下载速率，根据所述文件下载速率确定所述目标时段中从所述第一节点到所述第二节点的多个网络质量检测参数，其中，所述多个网络质量检测参数包括常规网速、网速变化幅度和网速下限，根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，并将所述网络波动模型提供给所述中心调度设备；
所述中心调度设备，用于基于所述网络波动模型检测所述目标时段中从所述第一节点到所述第二节点的网络质量。
可选的，所述中心调度设备，还用于在每个节点的设备组中随机选择至少一个设备，将预设的测试文件发送至所述每个节点的至少一个设备中进行存储；
所述第二节点的至少一个设备周期性向第一节点的至少一个设备发送所述测试文件的下载触发请求，以使所述第一节点的至少一个设备周期性从所述第二节点的至少一个设备处下载所述测试文件；
所述数据处理设备，用于获取预设数目个历史统计周期内目标时段中所述第一节点的至少一个设备从所述第二节点的至少一个设备处下载所述测试文件的文件下载速率。
可选的，所述数据处理设备具体用于：
将预设数目个历史统计周期内目标时段中所述第一节点对于所述第二节点上文件的所有文件下载速率的中值、四分位距和最小值，分别确定为所述目标时段中从所述第一节点到所述第二节点的常规网速、网速变化幅度和网速下限。
可选的，所述数据处理设备具体用于：
确定所述目标时段中从任意节点到所述第二节点的最大常规网速、最大网速变化幅度和最高网速下限；
基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中任意节点对于所述第二节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
可选的，所述数据处理设备具体用于：
确定所述目标时段中从所述第一节点到所述任意节点的最大常规网速、最大网速变化幅度和最高网速下限；
基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中所述第一节点对于任意节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
本发明实施例提供的技术方案带来的有益效果是：
本发明实施例中，数据处理设备获取预设数目个历史统计周期内目标时段中第一节点对于第二节点上文件的文件下载速率；数据处理设备根据文件下载速率确定目标时段中从第一节点到第二节点的多个网络质量检测参数，其中，多个网络质量检测参数包括常规网速、网速变化幅度和网速下限；数据处理设备根据多个网络质量检测参数和各自的预设权重，创建目标时段中从第一节点到第二节点的网络波动模型，并将网络波动模型提供给中心调度设备；中心调 度设备基于网络波动模型检测目标时段中从第一节点到第二节点的网络质量。这样，可以通过节点间的文件下载情况创建网络波动模型，然后通过网络波动模型来检测节点间的网络质量，无需设置任何网络检测工具，既可以降低网络质量的检测成本，又能够相对定量的分析衡量网络质量，为网络质量的历史分析、预测和调度优化提供直观定量的数据基础。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的一种检测网络质量的网络框架示意图；
图2是本发明实施例提供的一种检测网络质量的方法流程图。
具体实施方式
为使本发明的目的、技术方案和优点更加清楚，下面将结合附图对本发明实施方式作进一步地详细描述。
本发明实施例提供了一种检测网络质量的方法，该方法可以由CDN服务集群中的数据处理设备和中心调度设备共同实现，并由CDN服务集群中的节点辅助实现，具体网络框架如图1所示。其中，数据处理设备可以用于采集CDN服务集群中各节点间的业务执行数据，并对采集到的业务执行数据进行分析、整理等处理，中心调度设备可以基于数据处理设备对业务执行数据的处理结果，对CDN服务集群进行统一调度和管理。上述网络设备中均可以包括处理器、存储器、收发器，处理器可以用于进行下述流程中的检测网络质量的处理，存储器可以用于存储下述处理过程中需要的数据以及产生的数据，收发器可以用于接收和发送下述处理过程中的相关数据。可以理解，上述数据处理设备和中心调度设备的功能也可以由同一网络设备中的多个部件来实现。本实施例以数据处理设备和中心调度设备为独立的网络设备进行说明，其它情况与之类似，不再一一赘述。
下面将结合具体实施方式，对图2所示的处理流程进行详细的说明，内容 可以如下：
步骤201，数据处理设备获取预设数目个历史统计周期内，目标时段中第一节点对于第二节点上文件的文件下载速率。
其中，统计周期可以是人为设定的，统计文件下载速率的周期，可以是1天，每个统计周期可以分为多个时段，每个时段可以为1小时。第一节点和第二节点均可以是CDN服务集群中的任意一个节点。
在实施中，数据处理设备可以实时监控CDN服务集群中各个节点间的文件传输情况，如果某个节点从另外一个节点处下载了文件，数据处理设备则可以计算相应的文件下载速率，文件下载速率r可以等于文件大小除以文件下载时间。当需要对指定的目标时段内CDN服务集群中各节点间的网络质量进行检测时，数据处理设备可以获取预设数目个历史统计周期内，各节点在目标时段中每次从其它节点下载文件的文件下载速率。后续以统计周期为1天，目标时段为1小时为例进行说明，其它情况与之类似。以第一节点(可称为i节点)和第二节点(可称为j节点)为例，需要检测第h小时内两节点的网络质量时，数据处理设备获取预设数目天(如k天)内第h小时中i节点对于j节点上文件的文件下载速率。
可选的，各个节点上设置有专用的测试文件，数据处理设备可以只获取节点对于测试文件的文件下载速率，相应的，步骤201的处理可以如下：中心调度设备在每个节点的设备组中随机选择至少一个设备，将预设的测试文件发送至每个节点的至少一个设备中进行存储；第二节点的至少一个设备，周期性地向第一节点的至少一个设备发送测试文件的下载触发请求，以使第一节点的至少一个设备，周期性地从第二节点的至少一个设备处下载测试文件；数据处理设备获取预设数目个历史统计周期内，目标时段中第一节点的至少一个设备从第二节点的至少一个设备处下载测试文件的文件下载速率。
在实施中，CDN服务集群中的每个节点均可以为多个设备组成的设备组，中心调度设备可以在每个节点的设备组中随机选择至少一个设备，然后将预设的测试文件发送至上述至少一个设备中，因此，每个节点处均可以存在等量设备存储有上述预设的测试文件。上述测试文件的文件大小可以取2M左右，从而可以避免下载测试文件时因为文件过大而影响节点性能，或者因为文件过小而导致文件下载速率不准。之后，每个存储有测试文件的设备可以周期性地向其 它节点的每个选择出的设备发送测试文件的下载触发请求，以使这些设备周期性从本设备下载测试文件。以i节点和j节点为例，j节点处被选中的一个设备可以周期性(如每隔10分钟)向i节点处所有被选中的设备发送测试文件的下载触发请求，以使i节点的设备周期性从该设备处下载测试文件。这样，需要检测第h小时从i节点到j节点的网络质量时，数据处理设备可以获取k天内第h小时中i节点的设备从j节点的设备处下载测试文件的文件下载速率。
步骤202，数据处理设备根据文件下载速率确定目标时段中从第一节点到第二节点的多个网络质量检测参数。
其中，多个网络质量检测参数包括常规网速、网速变化幅度和网速下限。
在实施中，数据处理设备在获取到k天内第h小时中i节点对于j节点上文件的文件下载速率后，可以根据文件下载速率确定每个周期内第h小时中从i节点到j节点的常规网速x (h,i,j)、网速变化幅度y (h,i,j)和网速下限z (h,i,j)。此处，常规网速x (h,i,j)为两节点间的网络正常时的网速；网速变化幅度y (h,i,j)体现了两节点间的网络稳定性；网速下限z (h,i,j)为两节点间的网络质量极差时的网速。
可选的，确定网络质量检测参数的处理具体可以如下：数据处理设备将预设数目个历史统计周期内，目标时段中第一节点对于第二节点上文件的所有文件下载速率的中值、四分位距和最小值，分别确定为目标时段中从第一节点到第二节点的常规网速、网速变化幅度和网速下限。
在实施中，数据处理设备在获取到k天内第h小时中i节点对于j节点上文件的文件下载速率后，可以对所有文件下载速率进行汇总和整理，然后可以将所有文件下载速率的中值、四分位距和最小值，分别确定为目标时段中从第一节点到第二节点的常规网速x (h,i,j)、网速变化幅度y (h,i,j)和网速下限z (h,i,j)。具体的，r (n,h,i,j)表示第n天，第h小时，i节点对于j节点上文件的文件下载速率，处理k天的数据，则有：
Figure PCTCN2018091204-appb-000001
其中，h＝(0,1,…,23)，n＝(N-k+1,N-k+2,…,N)。
步骤203，数据处理设备根据多个网络质量检测参数和各自的预设权重，创建目标时段中从第一节点到第二节点的网络波动模型，并将网络波动模型提供 给中心调度设备。
在实施中，技术人员可以按照CDN服务集群的网络质量评判标准，对多个网络质量检测参数设置各自的预设权重，且所有预设权重之和为固定值。例如，网络质量评判标准为正常情况下网速越大越好、网络可以存在一定的波动性、无需考虑网络极差时的网速，则常规网速x (h,i,j)的预设权重a较高，网速变化幅度y (h,i,j)的预设权重b较低，网速下限z (h,i,j)的预设权重c极低。这样，数据处理设备在确定了每个周期内第h小时中从i节点到j节点的多个网络质量检测参数之后，可以根据多个网络质量检测参数和各自的预设权重，来创建第h小时中从i节点到j节点的网络波动模型f(h,i,j)＝g(ax,by,cz)。之后，数据处理设备可以将网络波动模型f(h,i,j)＝g(ax,by,cz)提供给中心调度设备。
可选的，可以先从i维度对多个网络质量检测参数进行归一化处理，然后再创建网络波动模型，相应的，步骤203的处理可以具体如下：数据处理设备确定目标时段中从任意节点到第二节点的最大常规网速、最大网速变化幅度和最高网速下限；数据处理设备基于最大常规网速、最大网速变化幅度和最高网速下限对多个网络质量检测参数进行归一化处理；数据处理设备根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中任意节点对于第二节点上文件的文件下载成功率，创建目标时段中从第一节点到第二节点的网络波动模型。
在实施中，数据处理设备可以先获取k天内第h小时中任意节点对于j节点上文件的文件下载速率，再基于步骤202的处理确定出目标时段中从每个节点到j节点的常规网速、网速变化幅度和网速下限，然后在所有的常规网速、网速变化幅度和网速下限中确定最大常规网速
Figure PCTCN2018091204-appb-000002
最大网速变化幅度
Figure PCTCN2018091204-appb-000003
和最高网速下限
Figure PCTCN2018091204-appb-000004
进一步的，数据处理设备可以基于最大常规网速、最大网速变化幅度和最高网速下限对第h时段中从i节点到j节点的多个网络质量检测参数进行归一化处理：
Figure PCTCN2018091204-appb-000005
同时，数据处理设备还可以获取k天内第h时段中任意节点对于j节点的文件下载成功率s (h,j)＝(总下载次数-下载失败次数)/总下载次数。之后，数据处理设备可以根据归一化处理后的多个网络质量检测参数X (h,i,j)、Y (h,i,j)、Z (h,i,j)和各自的预设权重a、b、c，以及上述文件下载成功率，创建第h时段中从i节点到j节点的网络波动模型：
Figure PCTCN2018091204-appb-000006
可选的，还可以先从j维度对多个网络质量检测参数进行归一化处理，然后再创建网络波动模型，相应的，步骤203的处理可以具体如下：数据处理设备确定目标时段中从第一节点到任意节点的最大常规网速、最大网速变化幅度和最高网速下限；数据处理设备基于最大常规网速、最大网速变化幅度和最高网速下限对多个网络质量检测参数进行归一化处理；数据处理设备根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中第一节点对于任意节点的文件下载成功率，创建目标时段中从第一节点到第二节点的网络波动模型。
值得一提的是，基于不同i维度和j维度归一化处理创建的两种网络波动模型均可以用于检测从第一节点到第二节点的网络质量，其具体的应用场景可以存在不同，具体的，基于i维度进行归一化处理，可以获取j节点上文件覆盖其他节点的网络质量情况；从j维度归一化处理，可以获取i节点下载其他节点上文件的网络质量情况。故而，当某个节点需要下载某个文件的时候，可以基于i维度归一化处理后创建的网络波动模型来检测节点间的网络质量，以判断从哪个节点处下载文件；而当需要设置某个节点的服务范围时，基于j维度归一化处理后创建的网络波动模型来检测节点间的网络质量，以决定该节点可以向哪些节点提供文件下载服务。
步骤204，中心调度设备基于网络波动模型检测目标时段中从第一节点到第二节点的网络质量。
在实施中，中心调度设备获取到数据处理设备提供的网络波动模型f(h,i,j)＝g(ax,by,cz)后，可以基于该网络波动模型f(h,i,j)＝g(ax,by,cz)来检测第h小时中从第一节点到第二节点的网络质量。
可以理解，基于上述步骤201至步骤204的处理，可以实现任意时段内CDN服务集群中任意两个节点间的网络质量的检测处理，而在多次进行步骤201至 步骤204的处理后，可以有效检测出所有时段内CDN服务集群的所有节点间的网络质量。
可选的，中心调度设备还可以基于节点间的网络波动模型进行多种应用，具体可以如下：
应用一：对于第三节点和第四节点，如果在任意时刻从第三节点到其它任意节点的网络波动模型，与从第四节点到相应节点的网络波动模型相似，则说明第三节点和第四节点所提供的网络服务的质量基本相同，故而中心调度设备还可以检测第三节点和第四节点的服务成本，选择服务成本低的节点，也可以以服务成本较低的节点为标准，对服务成本较高的节点进行配置，以降低CDN服务集群的总服务成本。
应用二，对于任意一个节点，中心调度设备可以基于节点间的网络波动模型，确定一个统计周期中各个时段内从其他节点到该节点的网络质量，然后可以针对性地调整该节点在每个时段的服务范围，即决定各个时段分别向哪些节点提供文件下载服务。
应用三，对于任意一个节点，中心调度设备可以基于节点间的网络波动模型，确定一个统计周期中各个时段内从该节点到其他节点的网络质量，然后可以针对性地调整该节点在每个时段的文件下载范围，即决定各个时段从哪些节点处下载文件。
应用四，可以对节点按照预设标准进行分类，如按照区域分类，或按照网络运营商分类，然后可以通过节点间的网络波动模型，整体上判断各分类间的网络质量，如可以检测节点和区域间的网络质量，检测不同网络运营商间的网络质量。
应用五，对于任意两个节点，可以结合当前统计周期的前一时段两节点间的网络状态(即f(h,i,j)的值)以及多个历史统计周期的当前时段两节点间的网络状态，来预测当前周期的当前时刻两节点间的网络状态。
本发明实施例中，数据处理设备获取预设数目个历史统计周期内，目标时段中第一节点对于第二节点上文件的文件下载速率；数据处理设备根据文件下载速率确定目标时段中从第一节点到第二节点的多个网络质量检测参数，其中，多个网络质量检测参数包括常规网速、网速变化幅度和网速下限；数据处理设备根据多个网络质量检测参数和各自的预设权重，创建目标时段中从第一节点 到第二节点的网络波动模型，并将网络波动模型提供给中心调度设备；中心调度设备基于网络波动模型检测目标时段中从第一节点到第二节点的网络质量。这样，可以通过节点间的文件下载情况创建网络波动模型，然后通过网络波动模型来检测节点间的网络质量，无需设置任何网络检测工具，既可以降低网络质量的检测成本，又能够相对定量的分析衡量网络质量，为网络质量的历史分析、预测和调度优化提供直观定量的数据基础。
基于相同的技术构思，本发明实施例还提供了一种检测网络质量的***，所述***包括数据处理设备、中心调度设备和包含第一节点与第二节点在内的多个节点，其中：
所述数据处理设备，用于获取预设数目个历史统计周期内目标时段中所述第一节点对于所述第二节点上文件的文件下载速率，根据所述文件下载速率确定所述目标时段中从所述第一节点到所述第二节点的多个网络质量检测参数，其中，所述多个网络质量检测参数包括常规网速、网速变化幅度和网速下限，根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，并将所述网络波动模型提供给所述中心调度设备；
所述中心调度设备，用于基于所述网络波动模型检测所述目标时段中从所述第一节点到所述第二节点的网络质量。
可选的，所述中心调度设备，还用于在每个节点的设备组中随机选择至少一个设备，将预设的测试文件发送至所述每个节点的至少一个设备中进行存储；
所述第二节点的至少一个设备周期性向第一节点的至少一个设备发送所述测试文件的下载触发请求，以使所述第一节点的至少一个设备周期性从所述第二节点的至少一个设备处下载所述测试文件；
所述数据处理设备，用于获取预设数目个历史统计周期内目标时段中所述第一节点的至少一个设备从所述第二节点的至少一个设备处下载所述测试文件的文件下载速率。
可选的，所述数据处理设备具体用于：
将预设数目个历史统计周期内目标时段中所述第一节点对于所述第二节点上文件的所有文件下载速率的中值、四分位距和最小值，分别确定为所述目标 时段中从所述第一节点到所述第二节点的常规网速、网速变化幅度和网速下限。
可选的，所述数据处理设备具体用于：
确定所述目标时段中从任意节点到所述第二节点的最大常规网速、最大网速变化幅度和最高网速下限；
基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中任意节点对于所述第二节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
可选的，所述数据处理设备具体用于：
确定所述目标时段中从所述第一节点到所述任意节点的最大常规网速、最大网速变化幅度和最高网速下限；
基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中所述第一节点对于任意节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
本发明实施例中，数据处理设备获取预设数目个历史统计周期内目标时段中第一节点对于第二节点上文件的文件下载速率；数据处理设备根据文件下载速率确定目标时段中从第一节点到第二节点的多个网络质量检测参数，其中，多个网络质量检测参数包括常规网速、网速变化幅度和网速下限；数据处理设备根据多个网络质量检测参数和各自的预设权重，创建目标时段中从第一节点到第二节点的网络波动模型，并将网络波动模型提供给中心调度设备；中心调度设备基于网络波动模型检测目标时段中从第一节点到第二节点的网络质量。这样，可以通过节点间的文件下载情况创建网络波动模型，然后通过网络波动模型来检测节点间的网络质量，无需设置任何网络检测工具，既可以降低网络质量的检测成本，又能够相对定量的分析衡量网络质量，为网络质量的历史分析、预测和调度优化提供直观定量的数据基础。
本领域普通技术人员可以理解实现上述实施例的全部或部分步骤可以通过 硬件来完成，也可以通过程序来指令相关的硬件完成，所述的程序可以存储于一种计算机可读存储介质中，上述提到的存储介质可以是只读存储器，磁盘或光盘等。
以上所述仅为本发明的较佳实施例，并不用以限制本发明，凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (10)

  1. 一种检测网络质量的方法，其特征在于，所述方法包括：
    数据处理设备获取预设数目个历史统计周期内，目标时段中第一节点对于第二节点上文件的文件下载速率；
    所述数据处理设备根据所述文件下载速率确定所述目标时段中从所述第一节点到所述第二节点的多个网络质量检测参数，其中，所述多个网络质量检测参数包括常规网速、网速变化幅度和网速下限；
    所述数据处理设备根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，并将所述网络波动模型提供给中心调度设备；
    所述中心调度设备基于所述网络波动模型检测所述目标时段中从所述第一节点到所述第二节点的网络质量。
  2. 根据权利要求1所述的方法，其特征在于，所述数据处理设备获取预设数目个历史统计周期内目标时段中第一节点对于第二节点的文件下载速率，包括：
    所述中心调度设备在每个节点的设备组中随机选择至少一个设备，将预设的测试文件发送至所述每个节点的至少一个设备中进行存储；
    第二节点的至少一个设备周期性向第一节点的至少一个设备发送所述测试文件的下载触发请求，以使所述第一节点的至少一个设备周期性从所述第二节点的至少一个设备处下载所述测试文件；
    所述数据处理设备获取预设数目个历史统计周期内目标时段中所述第一节点的至少一个设备从所述第二节点的至少一个设备处下载所述测试文件的文件下载速率。
  3. 根据权利要求1所述的方法，其特征在于，所述数据处理设备根据所述文件下载速率确定所述目标时段中从所述第一节点到所述第二节点的多个网络质量检测参数，包括：
    所述数据处理设备将预设数目个历史统计周期内目标时段中所述第一节点 对于所述第二节点上文件的所有文件下载速率的中值、四分位距和最小值，分别确定为所述目标时段中从所述第一节点到所述第二节点的常规网速、网速变化幅度和网速下限。
  4. 根据权利要求3所述的方法，其特征在于，所述数据处理设备根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，包括：
    所述数据处理设备确定所述目标时段中从任意节点到所述第二节点的最大常规网速、最大网速变化幅度和最高网速下限；
    所述数据处理设备基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
    所述数据处理设备根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中任意节点对于所述第二节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
  5. 根据权利要求3所述的方法，其特征在于，所述数据处理设备根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，包括：
    所述数据处理设备确定所述目标时段中从第一节点到所述任意节点的最大常规网速、最大网速变化幅度和最高网速下限；
    所述数据处理设备基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
    所述数据处理设备根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中所述第一节点对于任意节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
  6. 一种检测网络质量的***，其特征在于，所述***包括数据处理设备、中心调度设备和包含第一节点与第二节点在内的多个节点，其中：
    所述数据处理设备，用于获取预设数目个历史统计周期内目标时段中所述第一节点对于所述第二节点上文件的文件下载速率，根据所述文件下载速率确定所述目标时段中从所述第一节点到所述第二节点的多个网络质量检测参数，其中，所述多个网络质量检测参数包括常规网速、网速变化幅度和网速下限，根据所述多个网络质量检测参数和各自的预设权重，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型，并将所述网络波动模型提供给所述中心调度设备；
    所述中心调度设备，用于基于所述网络波动模型检测所述目标时段中从所述第一节点到所述第二节点的网络质量。
  7. 根据权利要求6所述的***，其特征在于，所述中心调度设备，还用于在每个节点的设备组中随机选择至少一个设备，将预设的测试文件发送至所述每个节点的至少一个设备中进行存储；
    所述第二节点的至少一个设备周期性向第一节点的至少一个设备发送所述测试文件的下载触发请求，以使所述第一节点的至少一个设备周期性从所述第二节点的至少一个设备处下载所述测试文件；
    所述数据处理设备，用于获取预设数目个历史统计周期内目标时段中所述第一节点的至少一个设备从所述第二节点的至少一个设备处下载所述测试文件的文件下载速率。
  8. 根据权利要求6所述的***，其特征在于，所述数据处理设备具体用于：
    将预设数目个历史统计周期内目标时段中所述第一节点对于所述第二节点上文件的所有文件下载速率的中值、四分位距和最小值，分别确定为所述目标时段中从所述第一节点到所述第二节点的常规网速、网速变化幅度和网速下限。
  9. 根据权利要求8所述的***，其特征在于，所述数据处理设备具体用于：
    确定所述目标时段中从任意节点到所述第二节点的最大常规网速、最大网速变化幅度和最高网速下限；
    基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
    根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中任意节点对于所述第二节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
  10. 根据权利要求8所述的***，其特征在于，所述数据处理设备具体用于：
    确定所述目标时段中从所述第一节点到所述任意节点的最大常规网速、最大网速变化幅度和最高网速下限；
    基于所述最大常规网速、最大网速变化幅度和最高网速下限对所述多个网络质量检测参数进行归一化处理；
    根据归一化处理后的多个网络质量检测参数和各自的预设权重，以及预设数目个历史统计周期内目标时段中所述第一节点对于任意节点的文件下载成功率，创建所述目标时段中从所述第一节点到所述第二节点的网络波动模型。
PCT/CN2018/091204 2018-01-25 2018-06-14 一种检测网络质量的方法和*** WO2019144560A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US16/463,212 US11108676B2 (en) 2018-01-25 2018-06-14 Method and system for detecting network quality based on a network fluctuation model
EP18893320.4A EP3550768B1 (en) 2018-01-25 2018-06-14 Network quality detection method and system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810073673.8 2018-01-25
CN201810073673.8A CN108234247B (zh) 2018-01-25 2018-01-25 一种检测网络质量的方法和***

Publications (1)

Publication Number Publication Date
WO2019144560A1 true WO2019144560A1 (zh) 2019-08-01

Family

ID=62668917

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/091204 WO2019144560A1 (zh) 2018-01-25 2018-06-14 一种检测网络质量的方法和***

Country Status (4)

Country Link
US (1) US11108676B2 (zh)
EP (1) EP3550768B1 (zh)
CN (1) CN108234247B (zh)
WO (1) WO2019144560A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115134277A (zh) * 2022-06-24 2022-09-30 山东信通电子股份有限公司 一种动态调整网络连接数的宽带网络速率测试方法及设备
CN117135709A (zh) * 2023-08-01 2023-11-28 苏州旭智设计营造有限公司 一种展厅网络快速切换控制***

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109561138B (zh) * 2018-11-19 2021-04-13 Oppo广东移动通信有限公司 文件下载方法及相关产品
CN109413688B (zh) * 2018-11-28 2022-03-15 南京林洋电力科技有限公司 一种基于gprs信道特征的智能化管理链路通道的方法
JP7109391B2 (ja) * 2019-02-26 2022-07-29 株式会社日立製作所 不正通信検知装置および不正通信検知プログラム
CN111147327B (zh) * 2019-12-23 2021-09-14 中国联合网络通信集团有限公司 网络质量的评估方法及装置
CN114039855B (zh) * 2021-11-04 2024-05-14 中国通信建设第三工程局有限公司 一种基于大数据的网络下载速度自动调整***及方法
CN115134274B (zh) * 2022-05-19 2023-08-22 赛尔网络有限公司 网络测速方法、装置、电子设备及存储介质
CN115314424B (zh) * 2022-08-08 2024-03-26 湖南三湘银行股份有限公司 一种快速检测网络信号方法及装置

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020080726A1 (en) * 2000-12-21 2002-06-27 International Business Machines Corporation System and method for determining network throughput speed and streaming utilization
CN101431473A (zh) * 2008-12-31 2009-05-13 深圳市迅雷网络技术有限公司 一种实现网络限速的方法及装置
US20140215059A1 (en) * 2011-05-12 2014-07-31 Telefonica, S.A. Method and a tracker for content delivery through a content delivery network
CN103986621A (zh) * 2014-05-23 2014-08-13 中国联合网络通信集团有限公司 一种评估移动互联网性能的方法和***
CN105050133A (zh) * 2015-07-31 2015-11-11 腾讯科技（深圳）有限公司 检测Wi-Fi热点下的网速的方法及装置
CN105634784A (zh) * 2014-11-06 2016-06-01 阿里巴巴集团控股有限公司 控制数据分发方法、装置及***
CN106534216A (zh) * 2016-12-30 2017-03-22 上海幻电信息科技有限公司 一种服务器部署结构
CN107241771A (zh) * 2017-06-16 2017-10-10 广东欧珀移动通信有限公司 网络切换方法、移动终端及计算机可读存储介质

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040024859A1 (en) * 2002-08-05 2004-02-05 Gerald Bloch Method and apparatus for communications network resource utilization assessment
FI122670B (fi) * 2006-09-15 2012-05-15 Tellabs Oy Menetelmä ja järjestely datakehysten välisen aikajakson tuottamiseksi
US8600425B1 (en) * 2010-02-17 2013-12-03 Sprint Communications Company L.P. TCP-aware power control in wireless networks
US9385938B2 (en) * 2010-06-22 2016-07-05 Blackberry Limited Information distribution in a wireless communication system
CN104348647B (zh) 2013-07-31 2019-04-12 腾讯科技（深圳）有限公司 多源带宽调度方法、装置及***
CN104683182B (zh) * 2015-02-15 2018-04-10 上海帝联信息科技股份有限公司 Idc机房网络服务质量检测方法及装置
CN105897838A (zh) * 2015-12-07 2016-08-24 乐视云计算有限公司 一种网络服务节点选择方法及装置
CN105847380A (zh) * 2016-04-18 2016-08-10 乐视控股（北京）有限公司 内容分发网络中的udp加速方法和***

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020080726A1 (en) * 2000-12-21 2002-06-27 International Business Machines Corporation System and method for determining network throughput speed and streaming utilization
CN101431473A (zh) * 2008-12-31 2009-05-13 深圳市迅雷网络技术有限公司 一种实现网络限速的方法及装置
US20140215059A1 (en) * 2011-05-12 2014-07-31 Telefonica, S.A. Method and a tracker for content delivery through a content delivery network
CN103986621A (zh) * 2014-05-23 2014-08-13 中国联合网络通信集团有限公司 一种评估移动互联网性能的方法和***
CN105634784A (zh) * 2014-11-06 2016-06-01 阿里巴巴集团控股有限公司 控制数据分发方法、装置及***
CN105050133A (zh) * 2015-07-31 2015-11-11 腾讯科技（深圳）有限公司 检测Wi-Fi热点下的网速的方法及装置
CN106534216A (zh) * 2016-12-30 2017-03-22 上海幻电信息科技有限公司 一种服务器部署结构
CN107241771A (zh) * 2017-06-16 2017-10-10 广东欧珀移动通信有限公司 网络切换方法、移动终端及计算机可读存储介质

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115134277A (zh) * 2022-06-24 2022-09-30 山东信通电子股份有限公司 一种动态调整网络连接数的宽带网络速率测试方法及设备
CN115134277B (zh) * 2022-06-24 2023-10-20 山东信通电子股份有限公司 一种动态调整网络连接数的宽带网络速率测试方法及设备
CN117135709A (zh) * 2023-08-01 2023-11-28 苏州旭智设计营造有限公司 一种展厅网络快速切换控制***

Also Published As

Publication number Publication date
CN108234247A (zh) 2018-06-29
US11108676B2 (en) 2021-08-31
EP3550768A4 (en) 2020-01-15
EP3550768A1 (en) 2019-10-09
EP3550768B1 (en) 2021-01-13
CN108234247B (zh) 2020-06-23
US20210105201A1 (en) 2021-04-08

Similar Documents

Publication Publication Date Title
WO2019144560A1 (zh) 一种检测网络质量的方法和***
US20180007121A1 (en) Performance-based content delivery
US10027739B1 (en) Performance-based content delivery
Liu et al. Monitoring and analyzing big traffic data of a large-scale cellular network with Hadoop
US10338967B2 (en) Systems and methods for predicting performance of applications on an internet of things (IoT) platform
CN102195992B (zh) 用于对从网络下载的数据进行反病毒扫描的***及方法
US8627147B2 (en) Method and computer program product for system tuning based on performance measurements and historical problem data and system thereof
CN104102875B (zh) 基于加权朴素贝叶斯分类器的软件服务质量监控方法及***
CN110428127B (zh) 自动化分析方法、用户设备、存储介质及装置
JP2019521427A (ja) 人工知能に基づくネットワークアドバイザー
Picoreti et al. Multilevel observability in cloud orchestration
CN107967488B (zh) 一种服务器的分类方法及分类***
CN110311812B (zh) 一种网络分析方法、装置和存储介质
CN104965784B (zh) 自动测试方法及装置
JP2010204966A (ja) サンプリング装置、サンプリング方法、サンプリングプログラム、クラス判別装置およびクラス判別システム。
CN103873293B (zh) 一种健康探测装置及方法
WO2020093502A1 (zh) 一种调整额定带宽的方法和装置
US10057155B2 (en) Method and apparatus for determining automatic scanning action
Xu et al. Lightweight and adaptive service api performance monitoring in highly dynamic cloud environment
CN116781563A (zh) 一种网页服务细粒度的评估装置、评估方法和评估***
US7617313B1 (en) Metric transport and database load
CN111124830A (zh) 一种微服务的监控方法及装置
CN113360376A (zh) 埋点测试方法和装置
CN111901134B (zh) 一种基于循环神经网络模型rnn的预测网络质量的方法和装置
WO2019239004A1 (en) Method for determining a performance indicator of a wireless telecommunications network

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2018893320

Country of ref document: EP

Effective date: 20190701

NENP Non-entry into the national phase

Ref country code: DE