CN112583658B - Available bandwidth measuring method, storage medium and equipment - Google Patents
Available bandwidth measuring method, storage medium and equipment Download PDFInfo
- Publication number
- CN112583658B CN112583658B CN202011310372.6A CN202011310372A CN112583658B CN 112583658 B CN112583658 B CN 112583658B CN 202011310372 A CN202011310372 A CN 202011310372A CN 112583658 B CN112583658 B CN 112583658B
- Authority
- CN
- China
- Prior art keywords
- available bandwidth
- detection
- packet
- link
- packets
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- 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
Abstract
The invention discloses a method for measuring available bandwidth, a storage medium and a device, wherein the method comprises the following steps: setting different available bandwidth values in a link of a network test bed as data acquisition points, and sending m detection flows to the link at each data acquisition point, wherein each detection flow comprises n detection packets; extracting time interval information between adjacent detection packets when each detection packet in each detection stream reaches a target node to obtain m groups of time interval information; using the available bandwidth value as a label, and performing model training on m groups of time interval information by using an XGboost algorithm to obtain an available bandwidth measurement model; injecting x detection streams into a link to be detected, and extracting x groups of time interval information; and inputting x groups of time interval information into the available bandwidth measurement model to obtain the available bandwidth value of the link to be measured. The available bandwidth measuring method provided by the invention is more accurate in measuring result, faster in measuring speed and lower in intrusiveness to the link to be measured.
Description
Technical Field
The present invention relates to the field of bandwidth measurement, and in particular, to a method, a storage medium, and a device for measuring available bandwidth.
Background
The Network Available Bandwidth (Network Available Bandwidth) is one of the metrics that quantify the transmission rate of a Network link, and this parameter characterizes the unused or idle capacity of the links in the Network for a period of time.
As an important indicator of network performance parameters, network available bandwidth measurement is an important component of network performance measurement. The measurement of the available bandwidth for a network link has rich applicability in many scenarios. For example, some large network measurement platforms take the measurement of available bandwidth as an important part, and for example, M-lab. In addition, the available bandwidth measurement method and tool can be used for determining whether the service level of the internet service provider is in accordance with the contract, which can urge the internet service providers to provide more stable network service for users. The measurement of the available bandwidth of the network can also be used in the scenes of routing selection, network congestion control, network fault location and the like, so that the network resource configuration is more reasonable, and the management of the network is more efficient. Therefore, the method has important practical effect in practical scenes on the measurement and evaluation of the available bandwidth of the network.
With the capacity of network links increasing continuously, the network environment architecture becomes more and more complex, and the measurement of available high bandwidth of the network faces new challenges.
Various available bandwidth measurement methods have been proposed in the art, and can be roughly classified into two types. The first type of method is called a flood type measuring method, a representative tool is iPerf, the flood type measuring method is to continuously send detection data packets to a link to be measured until the link is full, and then an available bandwidth value is obtained through calculation. The method has high measurement accuracy, but because a large amount of detection data needs to be sent to the link to be measured continuously, the method has high intrusiveness on the link to be measured, link congestion can be caused during measurement, and the normal operation of the existing service in the network is seriously influenced.
In order to overcome the defects of the flood type measuring method, scientific research personnel in the field also provide an optimized probe type measuring method, and the method can be used for constructing detection flow in different forms without sending a large amount of detection data to a link to be detected. Such an optimized probe-based measurement method can be classified into a Probe Rate Model (PRM) and a Probe Gap Model (PGM) according to the measurement principle.
The PRM observes the change and matching condition of the input rate of the data message and the one-way delay of the detection stream by controlling the rate of sending the detection data message, under the ideal condition, when the input probe rate is less than or equal to the available bandwidth, the one-way delay of the detection stream does not have the trend of increasing, otherwise, when the input detection rate is greater than the available high bandwidth, the one-way delay of the detection stream has the trend of increasing. The PRM-based available bandwidth measurement algorithm or tool attempts to find a point where the input rate and available bandwidth are equal, and thereby the available bandwidth of the network path. In the PRM measurement mode, however, the network under test is required to satisfy several important conditions. Firstly, the background flow of the link to be measured must be constant during the measurement period, that is, the available bandwidth value of the link to be measured cannot change during the measurement period, otherwise, the accuracy of the measurement method is affected. Secondly, the intermediate node of the link to be measured must satisfy the first-in first-out principle, otherwise, the measurement method may also fail. In the current network environment, various network emergency events occur frequently, and the existing PRM-based development and measurement tools (such as Pathload, pathchirp, and the like) mostly require long-time iteration to obtain a final result, so that it is difficult to ensure that the background flow is constant during measurement, and therefore, the accuracy is low and the measurement speed is slow in the existing environment.
The basic idea of PGM is that, when data is transmitted, due to the existence of background traffic, when a tight link is passed, the interval between probe packets changes due to the insertion of background traffic, and at this time, if the capacity of the tight link is known, the background traffic can be calculated according to the change of the input/output interval of the probe packets, so as to obtain the available bandwidth. Similar to the PRM measurement model, the PGM measurement model also requires that the link to be measured meets the principles of constant background traffic and first-in first-out. Also PGM requires that the background tight link capacity is known. Sponce and igi/ptr are representative of PGM-based available bandwidth measurement tools. Wherein, the Spruce requires the user to input the tight link capacity, which will severely limit the usage scenario, while the igi/ptr has built-in tight link capacity measurement method, which will cause extra error.
Therefore, the existing methods are difficult to satisfy the available bandwidth measurement in the current network environment, and the existing technologies are still to be improved and developed.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide an available bandwidth measuring method, a storage medium and a device for solving the deficiencies of the prior art, and to solve the problems of poor measuring accuracy, slow measuring speed and high intrusiveness on a link to be measured in the prior bandwidth measuring method.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
an available bandwidth measuring method, comprising the steps of:
setting different available bandwidth values as data acquisition points in a link of a network test bed, and sending m detection flows to the link at each data acquisition point, wherein each detection flow comprises n detection packets, m is an integer greater than or equal to 1, and n is an integer greater than or equal to 2;
extracting time interval information between adjacent detection packets when each detection packet in each detection stream reaches a target node to obtain m groups of time interval information;
performing model training on the m groups of time interval information by using the XGboost algorithm by taking the available bandwidth value as a label to obtain an available bandwidth measurement model;
injecting x detection streams into a link to be detected, and extracting x groups of time interval information;
and inputting the x groups of time interval information into the available bandwidth measurement model to obtain the available bandwidth value of the link to be measured.
In the method for measuring the available bandwidth, each detection flow is divided into two stages in the sending process, a first detection packet queue with the inter-packet intervals distributed in an exponential manner and gradually reduced is sent in the first stage, and a second detection packet queue with the inter-packet intervals being fixed values is sent in the second stage.
The available bandwidth measuring method, wherein the first probe packet queue comprises 5-100 full-size probe packets; the second probe packet queue contains 5-100 small size probe packets.
The method for measuring available bandwidth, wherein the first probe packet queue comprises 20 full-size probe packets; the second probe packet queue contains 10 small-size probe packets, wherein the payload of the full-size probe packets is 1472 bytes; the payload of the small-sized probe packet is 4 bytes.
The method for measuring available bandwidth is characterized in that the payload contents in the full-size probe packet and the small-size probe packet are different.
The method for measuring available bandwidth, wherein the step of setting different available bandwidth values in the link of the network test bed as data acquisition points comprises:
setting link capacity of a control bottleneck link for a port on a router;
background flow with different rates is sent to a measuring receiving end server through a background flow generator, and different available bandwidth values are used as data acquisition points by controlling the background flow.
In the step of taking different available bandwidth values as data acquisition points, the different available bandwidth values are 5Mbps to 1000Mbps, and one data acquisition point is set every 5 Mbps.
The method for measuring available bandwidth, wherein the step of inputting the time interval information between the x groups of probe packets into the available bandwidth measurement model to obtain the available bandwidth value of the link to be measured includes:
inputting the time interval information among the x groups of detection packets into the available bandwidth measurement model to obtain a plurality of available bandwidth values;
and taking the value with the highest occurrence frequency in the plurality of available bandwidth values, or the median of the plurality of available bandwidth values, or the average value of the plurality of available bandwidth values as the available bandwidth value of the link to be tested.
A storage medium storing one or more programs, the one or more programs being executable by one or more processors to implement the steps of the available bandwidth measuring method according to any one of the present invention.
An available bandwidth measuring device, comprising a processor adapted to implement instructions; and a storage medium adapted to store a plurality of instructions adapted to be loaded by a processor and to perform the steps of the available bandwidth measuring method according to any of the present invention.
Has the beneficial effects that: compared with the prior art, the available bandwidth measuring method provided by the invention has the following advantages: the measurement result is more accurate than that of the prior method; the measurement speed is higher, so that the influence of an emergency in the network on a measurement result is weakened; the method has low invasion to the link to be tested, and cannot cause long-time congestion of the link to be tested to influence the normal operation of other network services.
Drawings
Fig. 1 is a first flowchart of a method for measuring available bandwidth according to a preferred embodiment of the present invention.
Fig. 2 is a second flowchart of a method for measuring available bandwidth according to a preferred embodiment of the present invention.
FIG. 3 is a schematic diagram of the probe flow composition of the present invention.
FIG. 4 is a topology structure diagram of the test platform of the present invention.
FIG. 5 is a graph of a single probe flow measurement error profile according to the present invention.
FIG. 6 is a graph of a single measurement error profile according to the present invention.
Fig. 7 is a schematic diagram of an available bandwidth measuring apparatus according to the present invention.
Detailed Description
The present invention provides a method, a storage medium and a device for measuring available bandwidth, which will be described in further detail below with reference to the accompanying drawings and examples in order to make the objects, technical solutions and effects of the present invention clearer and clearer. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. As used herein, the term "and/or" includes all or any element and all combinations of one or more of the associated listed items.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The invention will be further explained by the description of the embodiments with reference to the drawings.
Referring to fig. 1, fig. 1 is a flowchart of a method for measuring available bandwidth according to a preferred embodiment of the present invention, as shown, the method includes the following steps:
s10, setting different available bandwidth values in a link of a network test bed as data acquisition points, and sending m detection flows to the link at each data acquisition point, wherein each detection flow comprises n detection packets, m is an integer greater than or equal to 1, and n is an integer greater than or equal to 2;
s20, extracting time interval information between adjacent detection packets when the detection packet of each detection stream reaches a target node to obtain m groups of time interval information;
s30, taking the available bandwidth value as a label, and performing model training on the m groups of time interval information by using an XGboost algorithm to obtain an available bandwidth measurement model;
s40, injecting x detection streams into a link to be detected, and extracting x groups of time interval information;
and S50, inputting the x groups of time interval information into the available bandwidth measurement model to obtain the available bandwidth value of the link to be measured.
Specifically, the present embodiment first constructs an available bandwidth measurement model according to the extracted features, and then measures the available bandwidth value of the link to be measured in real time based on the available bandwidth measurement model. As shown in fig. 2, the upper half of fig. 2 corresponds to the construction process of the available bandwidth measurement model: setting different available bandwidth values in a link of a network test bed as data acquisition points, and sending m detection flows to the link at each data acquisition point, wherein each detection flow comprises n detection packets; extracting time interval information between messages when a detection packet of each detection flow reaches a target node (n-1 message time interval information can be obtained by each detection flow), and obtaining m groups of message time interval information; and taking the currently known available bandwidth value as a label, and performing model training on the m groups of time interval information by using an XGboost algorithm to obtain a model for distinguishing the label according to characteristics, namely an available bandwidth measurement model, wherein m is an integer greater than or equal to 1, and n is an integer greater than or equal to 2.
The lower half in fig. 2 corresponds to the measurement process of the available bandwidth value of the link to be measured: injecting x detection streams into a link to be detected, and extracting x groups of detection packet time interval information, wherein x is an integer greater than or equal to 1; respectively substituting the obtained multiple groups of time interval information into the available bandwidth measurement model obtained by the previous training to obtain multiple available bandwidth values; and taking the value with the highest occurrence frequency in the plurality of available bandwidth values, or the median of the plurality of available bandwidth values, or the average value of the plurality of available bandwidth values as the available bandwidth value of the link to be tested.
The embodiment provides an available bandwidth measuring method combining a machine learning algorithm XGboost and a PGM measuring model, and the advantage of using the machine learning algorithm XGboost is that the problem solving mode of machine learning is to train existing data and model the data to form a model with generalization. For new problems, the result can be directly obtained by only extracting the characteristics of new data and bringing the characteristics into a trained model. Therefore, the embodiment takes the known available bandwidth value as a label, and trains the extracted m groups of time interval information by using a machine learning algorithm XGboost to obtain an available bandwidth measurement model; in the subsequent measurement process, the required characteristics (detection packet time interval information) are proposed and brought into the available bandwidth measurement model to obtain the available bandwidth value of the link to be measured only by the process of receiving and sending packets in a plurality of rounds, and the whole process is very quick and does not need repeated iteration. The rapid measurement is beneficial to reducing the influence of the change of the background flow on the measurement result. In addition, from the research on the existing available bandwidth measurement, it is found that the PGM-based measurement method performs better in both the measurement speed and the intrusiveness into the link to be measured because the number and size of packets are more controllable.
The available bandwidth measuring method provided by the embodiment is more accurate in measuring result than the previous method; the measurement speed is higher, so that the influence of an emergency in a network on a measurement result can be effectively weakened; the method has low invasion to the link to be tested, and cannot cause long-time congestion of the link to be tested to influence the normal operation of other network services.
In some embodiments, each probe flow is divided into two stages in the transmission process, where the first stage transmits a first probe packet queue whose inter-packet intervals are distributed exponentially and gradually decrease, and the second stage transmits a second probe packet queue whose inter-packet intervals are fixed values.
Specifically, in the aspect of probe flow, two probe packet queues (probe packet trains) with different inter-packet intervals are sent in this embodiment, and probe packet queues are adopted instead of probe packet pairs (probe packet pairs) because it can be known from the principle analysis and experimental tests of the foregoing distances and igi/ptrs that a probe packet queue has a better capturing capability for background traffic in a network, and the change of the interval between probe packets can reflect the actual situation of a network link to be tested. The first probe packet queue contains 5-100 full-size probe packets, but is not so limited; the second probe packet queue contains 5-100 small-size probe packets, but is not limited thereto, wherein Payload of the full-size probe packet is 1472 bytes, and Payload of the small-size probe packet is 4 bytes. The probe stream composition and the transmission mode thereof designed by the embodiment have the following advantages: firstly, in a first stage, the sending interval of the detection packets is gradually reduced in an exponential distribution mode, namely, the first detection packet queue presents different sending rates (from low speed to high speed), so that no matter the background flow is in a high-speed or low-speed state, the detection flow has a higher probability of being captured; secondly, the exponentially distributed and reduced sending intervals can greatly compress the sending time of the first detection packet queue, so that the completion time of the whole measurement is reduced; thirdly, because the full-size detection packet is sent in the first stage and the sending rate is higher as the first stage approaches the second stage, the queue length of the network equipment in the bottleneck link is increased, and the interval of the detection packet passing through the bottleneck link has a higher probability of being larger than the sending interval of the sending end; and the small-size detection packets with fixed inter-packet intervals are sent in the second stage, so that the queue length of the bottleneck link network equipment is reduced after the second stage is started, and the intervals of the detection packets passing through the bottleneck link are distributed in another way with higher probability, so that more inter-packet interval characteristics are captured in the whole detection process to be trained and learned by a machine learning module.
In some embodiments, as shown in fig. 3, for example, the first probe packet queue includes 20 full-size probe packets, the second probe packet queue includes 10 small-size probe packets, the full-size probe packets have a size of 1500 bytes, and the payload has 1472 bytes; the small size probe packet has a size of 32 bytes with a payload of 4 bytes. Each detection flow is divided into two stages in the sending process, the first stage sends a first detection packet queue with the inter-packet intervals distributed exponentially and gradually reduced, and the second stage sends a second detection packet queue with the inter-packet intervals being fixed values. Since the inter-packet spacing of the probe packets varies, the entire probe stream will exhibit variations in rate when transmitted. Through setting, the low-speed part of the detection flow is about 256Kbps, and the high-speed part of the detection flow can reach about 1Gbps, so that the detection flow can basically meet the link measurement requirement below 1 Gbps.
In some embodiments, to identify a possible packet loss location, payload of each probe packet may be set to be differentiated according to different contents, so as to quickly locate the specific packet loss location.
In some embodiments, the step of setting different available bandwidth values in the links of the network test bed as data collection points comprises: setting link capacity of a control bottleneck link for a port on a router; background flow with different rates is sent to a measuring receiving end server through a background flow generator, and different available bandwidth values are used as data acquisition points by controlling the background flow.
Specifically, in terms of data acquisition, the data acquisition environment is performed in a high-precision and high-reliability test platform, and the topology of the test platform is shown in fig. 4. In the data acquisition process, setting a network card on a Router (Soft Router) to control the link capacity of a bottleneck link; sending background traffic with different rates to a Measurement receiving end server (Measurement set) through a background traffic Generator (Cross-traffic Generator); by controlling background traffic, different available bandwidth environments are used as data acquisition points, and each available bandwidth value is a 'label' used for the training of an available bandwidth measurement model. And simultaneously sending the detection flow, simultaneously using a DAG card of a high-precision packet grabber to grab packets at the receiving and sending ends, thus obtaining the most original data, and then carrying out data cleaning to extract the characteristics for model training and testing.
Based on PGM's measurement principle, because the existence of background flow, the interval between the detection package can change between sending end and receiving end, because the interval between the detection package is invariable basically at the sending end, so only need to draw out the time stamp of the detection package that the receiving end snatched, the time stamp of two adjacent detection packages subtracts, obtains a plurality of time interval information, regards this as a set of characteristics of model training, and it is the value of a certain available bandwidth to correspond with it.
Because the high-speed part of the detection flow and the background flow are added together to fully occupy the link or even exceed the link capacity, packet loss can occur in the packet sending process, for the lost data packet, the interval between the detection packet and the two detection packets before and after the detection packet is replaced by the '-1', and because the packet loss rate is also related to the available bandwidth, the lost packet rate under each available bandwidth can be reflected in data concentration by using the '-1' to process the lost data, thereby being beneficial to the training of the model. In the model training stage, 80% of data in a data set is used as a training set, and an XGboost algorithm is used for training a model; the other 20% of the data tested the accuracy of the model as a test set.
As an example, in the data collection process, the link capacity of the bottleneck link is controlled by setting a network card on the software router. And sending background flow with different rates to a measuring receiving terminal server through a background flow generator. By controlling background flow, the set available bandwidth value is 5Mbps-1000Mbps, and one acquisition point is set every 5Mbps, namely a 'label' for machine learning model training later. 100 probe streams are sent from the measurement sender under each available bandwidth condition. Since there are 30 probe packets in a probe stream, 29 output intervals can be obtained, so there are 100 sets of 29-dimensional features for each tag, 80% of which are used for model training and 20% for model testing.
In some embodiments, after the available bandwidth measurement model is trained, 20% of the test set data is put into the available bandwidth measurement model for testing, and the results are shown in fig. 5 and 6. As can be seen from fig. 5, for the available bandwidth measurement model trained by using the XGBoost algorithm, for the measurement result of a single detection stream, the accuracy of the test is improved by a good margin with respect to the CART model, and the probability that the measurement value and the true value of the model are completely equal can reach more than 60%. If the accuracy is relaxed a little, the probability that the model measurement is positioned 5M around the true value (error between measured value and true value is less than or equal to 5M) can approach 90%.
Similarly, this is only the measurement result for a single probe flow, and in a complete measurement process, taking sending 10 probe flows as an example, the result is as shown in fig. 6, for the complete measurement, the accuracy of the measurement result is greatly improved, and the probability that the measured value and the true value are completely equal can reach about 85%. If the accuracy is relaxed a little, the probability that the model measurement is positioned 5M (error between the measured value and the actual value is less than or equal to 5M) around the actual value can reach more than 90%.
From the aspect of measurement speed, taking an example of sending 10 probe streams in total in one complete measurement, each probe stream is composed of 20 probe packets of 1500 bytes and 10 probe packets of 32 bytes, and performing one measurement injects about 2.4Mbit of traffic into a link to be measured, so that even on a link with an available bandwidth of only 5Mbps, transmission of such multiple traffic can be completed within 0.5 second. Compared with the traditional measuring method, the method has great improvement on the measuring speed.
From the aspect of measuring intrusiveness, in each measurement, the probe flow sent by the method is fixed and does not change due to different environments of the network to be measured, so that the measuring intrusiveness is controllable. However, because the interval of the probe packets in the probe stream is not constant, and the high-speed part of one probe stream can reach the giga level, for the link with a smaller available bandwidth, a short-time blocking of the link to be detected may be caused in the probing process.
In some embodiments, a storage medium is further provided, wherein the storage medium stores one or more programs, and the one or more programs are executable by one or more processors to implement the steps in the available bandwidth measuring method according to any one of the present invention.
In some embodiments, there is also provided an available bandwidth measuring device, as shown in fig. 7, which includes at least one processor (processor) 20; a display screen 21; and a memory (memory) 22, and may further include a communication Interface (Communications Interface) 23 and a bus 24. The processor 20, the display 21, the memory 22 and the communication interface 23 can communicate with each other through the bus 24. The display screen 21 is configured to display a user guidance interface preset in the initial setting mode. The communication interface 23 may transmit information. Processor 20 may call logic instructions in memory 22 to perform the methods in the embodiments described above.
Furthermore, the logic instructions in the memory 22 may be implemented in software functional units and stored in a computer readable storage medium when sold or used as a stand-alone product.
The memory 22, which is a computer-readable storage medium, may be configured to store a software program, a computer-executable program, such as program instructions or modules corresponding to the methods in the embodiments of the present disclosure. The processor 20 executes the functional application and data processing, i.e. implements the method in the above-described embodiments, by executing the software program, instructions or modules stored in the memory 22.
The memory 22 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal device, and the like. Further, the memory 22 may include a high speed random access memory and may also include a non-volatile memory. For example, a variety of media that can store program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, may also be transient storage media.
In addition, the specific processes loaded and executed by the storage medium and the instruction processors in the terminal device are described in detail in the method, and are not stated herein.
In summary, the available bandwidth measuring method provided by the present invention has the following advantages: the measurement result is more accurate than that of the prior method; the measurement speed is higher, so that the influence of an emergency in the network on a measurement result is weakened; the method has low invasion to the link to be tested, and cannot cause long-time congestion of the link to be tested to influence the normal operation of other network services.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (5)
1. An available bandwidth measuring method, comprising the steps of:
setting link capacity of a control bottleneck link for a port on a router;
background flows with different rates are sent to a measuring receiving end server through a background flow generator, different available bandwidth values are used as data acquisition points by controlling the background flows, the different available bandwidth values are 5Mbps-1000Mbps, and one data acquisition point is arranged at every 5 Mbps; sending m detection flows to the bottleneck link at each data acquisition point, wherein each detection flow comprises n detection packets, m is an integer greater than or equal to 1, and n is an integer greater than or equal to 2; each detection flow is divided into two stages in the sending process, a first detection packet queue with the inter-packet intervals distributed in an exponential manner and gradually reduced is sent in the first stage, and a second detection packet queue with the inter-packet intervals being fixed values is sent in the second stage; the first probe packet queue contains 5-100 full size probe packets; the second detection packet queue comprises 5-100 small-size detection packets; when the detection flow is sent, a high-precision packet grabbing device DAG card is used for grabbing packets at the receiving and sending ends at the same time, and for a lost detection packet, a space between the detection packet and the two detection packets before and after the detection packet is replaced by a '-1'; the Payload of each detection packet is set as different contents to be distinguished so as to quickly locate a specific packet loss position;
extracting time interval information between adjacent detection packets when each detection packet in each detection stream reaches a target node to obtain m groups of time interval information;
using the available bandwidth value as a label, and performing model training on the m groups of time interval information by using an XGboost algorithm to obtain an available bandwidth measurement model;
injecting x detection streams into a link to be detected, and extracting x groups of time interval information;
inputting the x groups of time interval information into the available bandwidth measurement model to obtain a plurality of available bandwidth values;
and taking the value with the highest occurrence frequency in the plurality of available bandwidth values, or the median of the plurality of available bandwidth values, or the average value of the plurality of available bandwidth values as the available bandwidth value of the link to be tested.
2. The method of claim 1, wherein the first probe packet queue comprises 20 full-size probe packets; the second probe packet queue contains 10 small-size probe packets, wherein the payload of the full-size probe packets is 1472 bytes; the payload of the small-sized probe packet is 4 bytes.
3. The method of claim 2, wherein the payload content in the full-size sounding packet and the small-size sounding packet are different.
4. A storage medium storing one or more programs, the one or more programs being executable by one or more processors to perform the steps in the available bandwidth measuring method according to any one of claims 1 to 3.
5. An available bandwidth measuring device comprising a processor adapted to implement instructions; and a storage medium adapted to store a plurality of instructions adapted to be loaded by a processor and to perform the steps of the available bandwidth measuring method according to any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011310372.6A CN112583658B (en) | 2020-11-20 | 2020-11-20 | Available bandwidth measuring method, storage medium and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011310372.6A CN112583658B (en) | 2020-11-20 | 2020-11-20 | Available bandwidth measuring method, storage medium and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112583658A CN112583658A (en) | 2021-03-30 |
| CN112583658B true CN112583658B (en) | 2022-12-23 |
Family
ID=75123065
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011310372.6A Active CN112583658B (en) | 2020-11-20 | 2020-11-20 | Available bandwidth measuring method, storage medium and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112583658B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115632727A (en) * | 2022-09-15 | 2023-01-20 | 鹏城实验室 | Spectrum sensing method and device |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101018161A (en) * | 2006-09-08 | 2007-08-15 | 中山大学 | A link, path, and network availability bandwidth measurement method |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101026509B (en) * | 2007-02-28 | 2011-07-20 | 西南科技大学 | End-to-end low available bandwidth measuring method |
| CN101895417B (en) * | 2010-07-06 | 2013-02-13 | 同济大学 | Positioning method for tight link based on available bandwidth of subpaths |
| CN102045219B (en) * | 2010-12-14 | 2012-07-25 | 同济大学 | High-efficiency single-end available bandwidth measuring method |
| US10679131B2 (en) * | 2012-07-12 | 2020-06-09 | Eaton Intelligent Power Limited | System and method for efficient data collection in distributed sensor measurement systems |
| CN111464323A (en) * | 2019-01-18 | 2020-07-28 | 北京沃东天骏信息技术有限公司 | Node bandwidth scheduling method and scheduling device |
-
2020
- 2020-11-20 CN CN202011310372.6A patent/CN112583658B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101018161A (en) * | 2006-09-08 | 2007-08-15 | 中山大学 | A link, path, and network availability bandwidth measurement method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112583658A (en) | 2021-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100369424C (en) | Method and apparatus for estimating terminal to terminal service grade protocol | |
| US7881221B2 (en) | Hardware implementation of network testing and performance monitoring in a network device | |
| US9251032B2 (en) | Method, computer program, and information processing apparatus for analyzing performance of computer system | |
| CN102047717B (en) | Method and system for telecommunications using layer 3 packets obtained from a sequence of layer 2 radio link control layer data frames | |
| US8774008B2 (en) | Real-time network measurement | |
| CN102045222B (en) | Real-time overall test method of network system | |
| EP2903213B1 (en) | Throughput test method and apparatus | |
| EP2713270A1 (en) | Resource scheduling method and device | |
| CN106713074A (en) | Data network quality piecewise detection method and system based on service content | |
| US8165011B2 (en) | Real-time network measurement | |
| CN102104502A (en) | Ethernet equipment performance test platform based on Linux system | |
| US8879403B2 (en) | Link microbenchmarking with idle link correction | |
| CN112583658B (en) | Available bandwidth measuring method, storage medium and equipment | |
| US8036132B1 (en) | Systems, devices, and methods for determining network failures | |
| CN112688837B (en) | Network measurement method and device based on time sliding window | |
| CN108183840A (en) | Verification method, device and the realization device of switch performance | |
| CN114189480B (en) | Flow sampling method, device, electronic equipment and medium | |
| CN114244737A (en) | Method, device and system for evaluating network quality | |
| JP3996085B2 (en) | Packet network output bandwidth design method and apparatus | |
| BAKAR et al. | SIMULATION OF WIRELESS ESTIMATION BANDWIDTH FOR NETWORK TECHNOLOGY | |
| JP3753704B2 (en) | Communication quality measuring apparatus and method | |
| CN117544537A (en) | Switch self-test streaming method, system, equipment and storage medium | |
| CN116016253A (en) | Network sub-health detection method, device and medium | |
| CN115550217A (en) | Network diagnosis method and device for seven-layer load balancing scene in cloud network | |
| Phanekham et al. | Accurate Bulk Throughput Benchmarks Using Reduced Network Resources |
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 |