KR101405472B1 - Method and apparatus for latency measurement based upon passive measurement - Google Patents

Abstract

A passive delay measurement method and apparatus are disclosed. According to a preferred embodiment of the present invention, the log information for the first request of the client is recorded, the data transfer time for the first request is calculated, the log information for the second request of the client is recorded, The processing time of the server for the first request and the time excluding the data transmission time for the first request are calculated as the delay in the log time for the second request and the log time interval for the first request of the client.
According to the present invention, it is possible to perform accurate delay measurement even in comparison with the active delay measurement based on a client, and more accurate delay measurement can be performed without using the original function of the server without increasing the load on the server.

Description

TECHNICAL FIELD The present invention relates to a passive delay measurement method and apparatus,

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for measuring a latency time, and more particularly, to a method and apparatus for measuring a passive delay.

There are various methods for measuring the time required for data transmission / reception between a node and a node, which are devices constituting a network.

Since the time required for data transmission and reception is an important factor for determining the quality of service, various methods for more accurately measuring the time required for data transmission and reception have been proposed.

In practice, various methods are applied to calculate the time required for data transmission and reception more accurately, so that the expression "calculation is performed rather than simply measurement" may be more accurate, but the term "measurement" is widely used and the term "calculation" It will not be used separately.

Meanwhile, the time required for data transmission and reception may be represented by RTT (Round Trip Time), latency, or delay time, which means time required for exchanging data or signals.

This delay measurement method is mainly used to generate test data and to transmit the test data generated from the network to be measured, and to monitor it in a specific server or network equipment using a protocol such as active and SNMP (Simple Network Monitoring Protocol) And passive to receive and measure data.

In other words, if the network consists of only the client and the server, the case where the data is generated for delay measurement and the measurement is performed is called the active type. The data transmission state and the transmission amount of the server and the client, It can be said to be passive when carrying out the delay measurement through.

Hereinafter, for the sake of convenience of explanation, the side of receiving data is referred to as a client and the side of giving data is referred to as a server. However, the present invention is not limited to such a name, It is obvious that measurement is possible.

Meanwhile, in the conventional active delay measurement method, a method of generating a test packet and transmitting it to the server and measuring the delay through the arrival time of the test packet is used.

Particularly, in the TCP / IP based method, a ping test method using an application program called ping is widely used as a method for checking whether data is received or not.

This active delay measurement method is advantageous in that a delay can be measured more accurately since the delay is measured based on a client provided with an actual service.

However, it is necessary to install a separate program on the client side, to generate additional traffic to the network, and in some cases, transmission of a test packet for delay measurement may be mistaken for a DDoS attack or the like, There are drawbacks to this.

In particular, there is a disadvantage in that the active delay measurement itself is not performed due to security reasons in the case of a specific network, and thus it is not applicable to a specific network other than the public network.

On the other hand, the passive delay measurement method measures the delay based on the data transmitted between the client and the server. Therefore, it is possible to eliminate the need to install a separate program on the client, generate no additional traffic to the network, There is an advantage.

However, since the delay is measured based on the server rather than the client, there may be a slight difference from the general active measurement result due to the difference in the measured position.

In the passive delay measurement method, a network connection analysis method such as a TCP connection is widely used. More specifically, a method of measuring delay through analysis of a three-way handshaking or FIN-ACK pair occurring at the end is used.

However, this network connection analysis method has a drawback in that it may cause a heavy load on the server because it is a method of measuring the delay through a detailed examination of a more detailed part of the data packet.

In order to solve the above-mentioned problems, the present invention proposes a passive delay measurement method and apparatus that enable precise delay measurement compared with an active delay measurement based on a client.

It also proposes a passive delay measurement method and apparatus that enables more accurate delay measurements without increasing the load on the server.

Other objects of the present invention will become readily apparent from the following description of the embodiments.

According to an aspect of the present invention, a passive delay measurement method is provided.

According to a preferred embodiment of the present invention, there is provided a passive latency measurement method performed by the server in a network including a client and a server, the method comprising: recording log information on a first request of the client; Calculating a data transmission time for the first request; Recording log information for a second request of the client; And a delay time excluding a processing time of the server for the first request and a data transmission time for the first request in a log time interval for the second request of the client and a log time interval for the first request of the client, And a step of calculating a passive delay is provided.

The delay calculation may be performed by calculating an average over the delays computed for each request of the client.

The first request and the second request may be received by continuing the connection between the client and the server.

The step of calculating the data transmission time for the first request may be calculated using the size of the data transmitted according to the first request included in the log information and the bandwidth between the connection of the client and the server.

The log information may be information including information about a logging time of a client accessing a server, a processing time of the server for a request of the client, and a size of data transmitted in response to the request of the client.

According to another aspect of the present invention, a passive delay measurement apparatus is provided.

According to a preferred embodiment of the present invention, there is provided an apparatus for measuring passive latency by being connected to a client through a network, the apparatus comprising: a log information recording unit for recording log information on a first request and a second request of the client; A data transmission time calculation unit for calculating a data transmission time for the first request; And a processing time of the server for the first request in a log time interval for the second request of the client and a log time interval for the first request of the client and a processing time of the server for the first request calculated in the data transmission time calculation unit And a delay calculation section for calculating a time excluding the data transmission time as a delay.

The delay calculation in the delay calculator may be performed by calculating an average over the delays computed for each request of the client.

The first request and the second request can be received at the server by continuing a connection between the client and the server.

Wherein the step of calculating the data transmission time for the first request in the data transmission time calculation unit includes calculating a size of data transmitted according to the first request included in the log information for the first request, It can be calculated using the bandwidth between connections.

Wherein the log information recorded in the log information recording unit includes information on a logging time of the client connected to the server, a processing time of the server in response to the request of the client, and a size of data transmitted in response to the request of the client Information.

According to another aspect of the present invention, there is provided a recording medium on which a program in which a passive delay measurement method is implemented is recorded.

According to a preferred embodiment of the present invention, there is provided a recording medium on which a program for implementing a passive latency measurement method performed by the server in a network including a client and a server is recorded, Storing log information about the log; Calculating a data transmission time for the first request; Storing log information for a second request of the client; And a delay time excluding a processing time of the server for the first request and a data transmission time for the first request in a log time interval for the second request of the client and a log time interval for the first request of the client, There is provided a recording medium recording a program for implementing a passive delay measurement method including a step of calculating a delay time.

The delay calculation may be performed by calculating an average over the delays computed for each request of the client.

The first request and the second request may be received by continuing the connection between the client and the server.

The step of calculating the data transmission time for the first request may be calculated using the size of the data transmitted according to the first request included in the log information and the bandwidth between the connection of the client and the server.

The log information may be information including information about a logging time of a client accessing a server, a processing time of the server for a request of the client, and a size of data transmitted in response to the request of the client.

As described above, according to the method and apparatus for measuring passive delay according to the present invention, accurate delay measurement can be performed in comparison with active delay measurement based on a client.

In addition, there is an advantage that more accurate delay measurement can be performed while using the original function of the server without increasing the load on the server.

FIG. 1 is a flowchart showing a procedure of a delay measurement method according to a preferred embodiment of the present invention. FIG.
Fig. 2 is a view showing that each time for applying a delay measurement method according to a preferred embodiment of the present invention is measured. Fig.
3 is a diagram showing a configuration of a delay measurement apparatus according to a preferred embodiment of the present invention.
FIG. 4 is a graph showing a comparison between a delay by a delay measurement method and a delay by a ping test according to a preferred embodiment of the present invention.
FIG. 5 is a graph illustrating a comparison between the delay of each region according to the delay measurement method according to the preferred embodiment of the present invention and the minimum value of each region by the TCP connection analysis.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.

Like reference numerals are used for like elements in describing each drawing. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

The terms first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component.

And / or < / RTI > includes any combination of a plurality of related listed items or any of a plurality of related listed items.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, .

On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention.

The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the contextual meaning of the related art and are to be interpreted as either ideal or overly formal in the sense of the present application Do not.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings, wherein like or corresponding elements are denoted by the same reference numerals, and a duplicate description thereof will be omitted.

First, a sequence of performing a delay measurement method according to a preferred embodiment of the present invention will be described with reference to FIG.

FIG. 1 is a flowchart illustrating a procedure for performing a delay measurement method according to a preferred embodiment of the present invention.

As described above, for convenience of explanation, the client 110 is referred to as a client 100 and the server 110 is referred to as a client 110. However, the name is not limited to this name, It is apparent that the delay measurement method according to the present invention is applicable to any apparatus constituting the network.

As shown in FIG. 1, a delay measurement method according to an exemplary embodiment of the present invention firstly transmits a request for data or a signal from the client 100 (hereinafter, referred to as a first request for convenience and description) The log information is stored (S200).

When the client 100 accesses the server 110, it is called log information which is recorded every time when the client 100 accesses the server 110. In this case,

The log information is generally a logging time which is a time when the client 100 is connected to the server 110. The logging time is a time when the server 110 is connected to the client 110 100, the server process time, the data size transmitted to the client 100, and the identification information (ID) of the data are recorded.

Since the size of the data transmitted to the client 100 is known through the log information and the connection between the client 100 and the server 110 is established, the bandwidth of the connection between the client 100 and the server 110 and information on the bandwidth.

Therefore, the transmission time of the arithmetic data can be calculated as a data size divided by the bandwidth size (S102).

Next, the server 110 records log information about the next data or signal request (hereinafter referred to as a second request for convenience and description) from the client 100 (S104).

The log information for the second request also includes a logging time which is a time when the client 100 is connected to the server 110 and a logging time when the server 110 transmits data to the client 100 A sever process time, a data size transmitted to the client 100, and identification information (ID).

Next, a latency is calculated using the log information for the first request, the data transmission time, and the log information for the second request (S106).

The delay calculation is first performed by subtracting the server processing period and the data transmission time from the logging time for the second request and the logging time interval for the first request.

A detailed description will now be made with reference to FIG. 2, in which a time is calculated as a delay by subtracting a server processing period and a data transmission time from a logging time for the second request and a logging time interval for the first request.

FIG. 2 is a view showing that each time for applying the delay measurement method according to the preferred embodiment of the present invention is measured.

As shown in FIG. 2, the logging time when the client 100 accesses the server 110 for the first request is referred to as Tlog1.

The time required for the client 100 to transmit data or signals to the client 100 and completed is the server processing time, which is referred to as T1.

Next, the data transmission time calculated according to the data size and bandwidth is referred to as T2.

That is, T2 may be a data transmission start time and a data transmission completion time in the server 100 that are calculated arithmetically.

Meanwhile, in the case of the data transfer start time and the data transfer completion time, the server 110 may record and store data separately, but it is generally not included in the log information to be recorded as a function of the general server 110.

That is, since it is possible to know only Tlog1, Tlog2, T1 and the size of data through the log information recorded as a function of the general server 110, the data transmission time is calculated separately using data size and bandwidth information.

By doing this, there is no need to include additional functions and data transmission / reception in addition to the functions inherent to the conventional server 110, and there is an advantage that only functions inherent to the server 10 can be used.

On the other hand, if the server 110 includes a function of separately recording and storing the data transmission start time and the data transmission completion time, it may not be necessary to separately calculate the time T2.

Next, when the time for receiving and processing the data in the client 100 is t3, the client 100 requests the next data again after the processing of the received data is completed, and the log information is also recorded for the second request And the logging time for the second request is Tlog2.

On the other hand, the delay computes the time, excluding T1 and T2, as the delay in the time interval between Tlog2 and Tlog1.

That is, the time excluding the processing time of the server T1 and the data transmission time T2 is calculated as a delay in the logging time for the first request and the logging time interval for the second request.

The passive delay measurement method according to the present invention will be described with reference to FIG. 2 and a conventional active delay measurement method.

Since the conventional active delay measurement method is based on the client 100, t1, which is the time from when the client 100 performs the first request to the server 110 until the start of receiving the data, is calculated as the delay.

On the other hand, according to the delay measurement method of the present invention, the delay time of the first request and the logging time interval of the second request are delayed by T3, which is a time excluding the processing time of the server and T1, .

In order to determine whether t1 and T3 are equal or approximate to each other, the request time in the client 100 and the connection time in the server 110 are determined so that the time required for the request to move from the client 100 to the server 110 It can be assumed that it is the same time.

Also, since the data arrival time in the client 100 and the data transmission start time in the server 110 are eventually transferred from the server to the client 100, it can be assumed that they are the same time, except for the data arrival time.

According to this assumption, it can be said that t1 + t2 + t3 = T1 + T2 + T3.

On the other hand, since t2 and T2 can be substantially the same or at least substantially the same time as the data transmission time, they can be the same or approximate values. Consequently, the delay T3 by the passive delay measurement method and the delay t1 by the active delay measurement method are equal , T3 and T1 may be the same or have approximate values.

In the case of t3, the time for data processing in the client 100 is T2, and the time for data processing in the server 110 is T2, and the data to be processed is the same data.

Therefore, it can be seen that the time difference between t3 and T2 is not large, and the smaller the data size is, the more the difference is.

This can be verified by judging based on the RTT, which is one type of delay measurement.

The RTT measured based on the client 100 becomes a value obtained by adding the processing time of the server 110 to the time when the client 100 accesses the server 110 and the response is returned to the client 100 again.

On the other hand, in the case of the RTT measured based on the server 110, a value obtained by adding the processing time of the client 100 to the time when the server 110 accesses the client 100 and the response is returned to the server 110 do.

As a result, considering the time period between the client 100 and the server 110, the difference between the processing time in the client 100 and the processing time in the server 110 is not limited to the active type, the passive type RTT, Is the time difference.

Therefore, there is no significant difference in whether the time of data and signal coming and going between the client 100 and the server 110 is based on the client 100, on the basis of the server 110, If the difference in processing time in the server 110 and the server 110 is not large, it means that both the active type and the passive RTT measurement can have similar values. Particularly, the smaller the data size is, the larger the difference is.

According to the present invention, in the passive delay measurement method of measuring the delay in the server 110, the processing time information of the server and the size information of the data are included in the log information, which is information originally recorded according to the function of the server 100 The delay measurement can be performed without the need to transmit / receive other information or analyze data.

On the other hand, the log information recorded in the server 110 is continuously recorded continuously when the connection between the client 100 and the server 110 is continued.

Therefore, in calculating the delay, it is also possible to calculate the delays from a plurality of pieces of log information, and to calculate the delays by averaging these delays.

For example, if there are 10 log information, and the connection between the client 100 and the server 110 is continued while 10 log information is being recorded, eventually, 9 time interval information can be acquired, Delay information can be obtained.

By calculating the average of the nine pieces of delay information as one piece of delay information, it is possible to make the measurement of the delay with higher probability more accurate.

In particular, due to the characteristics of the network, there may be a case where the delay value greatly changes due to a malfunction at a specific time or a change of a network state. Therefore, the measurement of the delay through the average value can improve the accuracy of the delay.

In addition, for example, a method of excluding a delay having a maximum value or a minimum value in the calculation of the average value through sampling, or a method of calculating a delay except for a maximum value and a delay that is twice the minimum value of the delay It is also possible to further increase the accuracy of the delay measured through.

Since the server 110 is connected to an unspecified number of the clients 100, the server 110 allows the server 110 to continue the connection with the specific client 100 for delay measurement, It may be necessary to have the second request be received immediately after the processing for the second request is completed.

For this purpose, it is necessary to establish a connection maintenance through setting such as 'keep alive' based on TCP / IP (Transfer Control Protocol / Internet Protocol), but the present invention is not limited thereto.

It is obvious that the delay measurement method according to the present invention can be implemented by a program and stored in a recording medium or the like, and a program is installed in the digital processing apparatus to implement the delay measurement method according to the present invention.

Hereinafter, a configuration of a delay measurement apparatus for performing delay measurement according to the present invention will be described with reference to FIG.

3 is a diagram illustrating a configuration of a delay measurement apparatus according to an embodiment of the present invention.

3, the delay measurement apparatus according to an exemplary embodiment of the present invention may include a log information recording unit 300, a data transmission time calculation unit 310, and a delay calculation unit 320.

The log information recording unit 300 records the logging time, the processing time of the server, the size of the data, and the like, which are the times when the client 100 accesses the server 110.

In addition, it may include identification information of data and code information which is information responded to the client 100.

It is obvious that the log information recording unit 300 can be implemented by a function originally included in the server 100. [

The data transmission time calculation unit 310 calculates the data transmission time using the information about the size of the data recorded by the log information recording unit 300 and the bandwidth between the client 100 and the server 110. [

As described above, if there is a case where the server 110 separately records and stores the data transmission start time and the data transmission completion time, the data transmission time calculation unit 310 may be implemented by a function included in the server 110 .

However, since data transmission start time and data transmission completion time are not included in log information, data transmission time is calculated separately using data size and bandwidth information.

Particularly, in the present invention, it is desirable to reduce the load of the server 110 without including additional functions other than the original functions of the conventional server 110, so that only the log information stored in accordance with the function of the server 110 is used Lt; / RTI >

The delay calculator 320 calculates the delay by subtracting the processing time and the data transmission time of the server from the logging time for the second request and the logging time interval for the first request.

The reason why the processing time of the server and the time excluding the data transmission time in the logging time interval for the second request and the logging time interval for the first request can be calculated as a delay is as described above.

Meanwhile, it is apparent that the delay measurement apparatus according to the preferred embodiment of the present invention can be implemented by being included in a server or through a separate device connected to a server.

In addition, the delay measurement method according to the present invention may be implemented in the form of a program and may be implemented in the form of a digital processing apparatus such as a server in which a program is installed.

Hereinafter, a delay measurement according to a preferred embodiment of the present invention will be described with reference to FIGS. 4 and 5, in which a delay measured by actually applying a delay measurement method according to a preferred embodiment of the present invention is displayed and compared.

Meanwhile, the conventional RTT measurement method is widely used for delay measurement by ping test and TCP connection analysis.

FIG. 4 is a graph comparing the results of the delay measurement method according to the preferred embodiment of the present invention with the results of the delay measurement method using the ping test. FIG. 5 is a diagram illustrating a delay measurement method according to a preferred embodiment of the present invention. And the minimum value of delay for each region by TCP connection analysis.

The delay measurement method according to a preferred embodiment of the present invention uses a delay measurement method using a sampling method for calculating an average value excluding a maximum value.

First, FIG. 4 is a comparison of delay measurement by the delay measurement method and delay measurement by the ping test according to a preferred embodiment of the present invention. As a method of knowing the latency, Graph.

As shown in FIG. 4, when comparing the delay by the delay measurement method and the delay by the ping test according to the preferred embodiment of the present invention, it can be seen that no significant difference occurs in the average value.

In the case of the maximum value, a smaller value appears rather than the case of the ping test, and a minimum value shows a larger value than the ping test, but the difference is not large.

Next, FIG. 5 is a graph illustrating a comparison of RTT as a delay based on the delay measurement by TCP and the delay measurement by the world region in Seoul by the delay measurement method according to a preferred embodiment of the present invention.

As shown in FIG. 5, it can be seen that the delay according to the present invention is larger than the delay caused by TCP connection analysis on average, but the difference is not large.

That is, it can be seen that the delay measurement according to the present invention does not cause a large difference in comparison with the delay measurement through the RTT measurement by the conventional ping test or the RTT measurement through the TCP connection analysis.

Therefore, according to the passive delay measurement method of the present invention, there is no need to install a separate program on the client side as in the active delay measurement method, and even when no additional traffic is generated in the network, a large error is generated compared with the active delay measurement method So that the delay can be measured.

Even in the case of the server, since the function of recording the log information implemented as the original function of the server is used, the load of the server is not greatly increased even though it is a passive delay measurement method.

It will be apparent to those skilled in the relevant art that various modifications, additions and substitutions are possible, without departing from the spirit and scope of the invention as defined by the appended claims. The appended claims are to be considered as falling within the scope of the following claims.

100: client 110: server
300: log information recording unit 310: data transmission time calculation unit
320: delay calculation unit

Claims (11)

1. A passive latency measurement method performed by the server in a network including a client and a server,
Recording log information for a first request of the client;
Calculating a data transmission time for the first request;
Recording log information for a second request of the client, wherein the first request and the second request are received while the connection between the client and the server continues; And
The processing time for the first request of the server included in the log information for the first request in the log time for the second request of the client and the log time interval for the first request of the client, And calculating a time, excluding a data transmission time, for the delay time as a delay.
The method according to claim 1,
Wherein the delay calculation is performed by calculating an average over the delays computed for each request of the client.
delete The method according to claim 1,
The step of calculating the data transmission time for the first request comprises:
Wherein the bandwidth is calculated using a size of data transmitted according to the first request included in the log information and a bandwidth between the connection of the client and the server.
The method according to claim 1,
Wherein the log information is information including information on a logging time of a client connected to a server, a processing time of the server in response to a client request, and a size of data transmitted in response to a request of the client. Way.
1. An apparatus for measuring passive latency by being connected to a client through a network,
A connection between the log information recording unit and the server for recording log information for the first request and the second request of the client is continuously received;
A data transmission time calculation unit for calculating a data transmission time for the first request; And
Wherein the processing time of the server for the first request of the server included in the log information for the first request in the log time of the second request of the client and the log time interval of the first request of the client, And a delay calculation unit for calculating, as a delay, a time calculated by the time calculation unit, excluding a data transmission time for the first request.
The method according to claim 6,
Wherein the delay calculation in the delay calculator is performed by calculating an average over the delays calculated for each request of the client.
delete The method according to claim 6,
Wherein the step of calculating the data transmission time for the first request in the data transmission time calculation unit includes calculating a size of data transmitted according to the first request included in the log information for the first request, Wherein the bandwidth is calculated using a bandwidth between connections.
The method according to claim 6,
Wherein the log information recorded in the log information recording unit includes information on a logging time of the client connected to the server, a processing time of the server in response to the request of the client, and a size of data transmitted in response to the request of the client Wherein the information is information.
A recording medium on which a program for implementing a passive latency measurement method performed by the server in a network including a client and a server is recorded,
Storing log information for a first request of the client;
Calculating a data transmission time for the first request;
Storing log information for a second request of the client, wherein the first request and the second request are received while the connection between the client and the server continues; And
The processing time for the first request of the server included in the log information for the first request in the log time for the second request of the client and the log time interval for the first request of the client, And calculating a time, excluding a data transmission time, for each of the plurality of channels, as a delay.

Images