KR20180095981A - Fog Server Deployment Method Based on Host Location and Traffic Pattern in Fog Computing Environment - Google Patents

Abstract

The present invention relates to a fog server arrangement method based on a host location and a traffic pattern in a fog computing environment, which generates a logical flow for obtaining a current bandwidth usage of a link between a fog device and a core network, and arranges a fog server with respect to a link bandwidth usage. The fog server arrangement method of the present invention comprises the steps of: generating a logical flow (f), which is represented by a straight line from a data source toward a client and has a direction (Src, Dst), a bandwidth usage (B_before, B_after) before and after going through a server, and computing resource size information required for processing; obtaining a current bandwidth usage of a link (l_(d_n, c)) between a fog device and a core network using the generated logical flow (f); determining a rank of a candidate fog device in which a fog server can be arranged with respect to a link bandwidth usage between the obtained fog device and the core network; and arranging the fog server with respect to the obtained rank, and additionally arranging the fog server with respect to traffic generated in the core network and a traffic ratio generated in a local area.

Description

[0001] The present invention relates to a fog server deployment method based on host location and traffic pattern in a fog computing environment,

The present invention relates to fog computing. More particularly, the present invention relates to a fog computing environment in which, in a fog computing environment in which a logical flow is generated to obtain a current bandwidth usage of a link between a fog device and a core network and a fog server is allocated based on a link bandwidth usage, Pattern based fog server placement method.

Cloud computing refers to the way a user accesses a central data center through a network and uses hardware, software, and data.

However, there are many problems with such cloud services.

Typically, the distance between the cloud and the user is so long that the latency is long. As a result, many users are using the cloud service because it is combined with the Internet (IoT) and the whole network traffic is increased.

Recently, the use of related devices has been steadily increasing as the internet (IoT) is getting attention.

IoT devices require a service pattern that differs from the service method between the main server and the client in the cloud.

In environments that support IoT devices, local data and services are often requested. To meet these requirements, Cisco introduced the concept of fog computing for the first time.

Many researches related to fog computing show that users can receive computing resources and services from a closer location through fog computing and receive services that could not be serviced due to limit of response speed in existing cloud environment have.

In addition, computing resources and services are provided at specific locations in the fog environment, so that various data generated locally can be collected and utilized.

However, since most of the research focuses on the advantages of fog computing and the idea of fog computing architecture, further research on fog computing is needed.

In a fog computing environment, a fog server for providing computing resources, services and applications must be operating. These fog servers are deployed from the cloud into the fog area and then operate. Therefore, a mechanism for fog server deployment is needed.

There is a method based on the computing resources of the fog device to place the fog server.

This method places the fog server based on the computing resource utilization status of the target fog device in which the fog server is to be installed.

However, the fog devices lack computing resources relative to handling a large number of requests compared to a cloud server handling many users' services.

For this reason, if a service request is increased to a specific fog device, a load may be concentrated on the device, which may cause a problem in service processing.

And because users do not always use the same service, traffic on the network is constantly changing.

Even if the fog server is deployed, the flow to the service continuously changes, so running the service on the same fog server always leads to inefficient results.

Korean Patent Publication No. 10-2015-0120555 Korean Patent Publication No. 10-2014-0056382

The present invention solves the problem of the conventional arrangement of the fog computing servers. The present invention creates a logical flow for obtaining the current bandwidth usage of the link between the fog device and the core network and arranges the fog server based on the link bandwidth usage The present invention is directed to a method of arranging a fog server based on a host location and a traffic pattern in a fog computing environment.

The present invention creates a logical flow representing network traffic on an actual network topology and obtains the current bandwidth usage of the link between the fog device and the core network to determine the host location and location of the fog server in a fog computing environment, The purpose of this paper is to provide a fog server deployment method based on traffic pattern.

The present invention relates to a method and apparatus for determining the order of candidate fog devices on which a fog server can be placed based on a link bandwidth usage between a fog device and a core network and a fog computing method in which an added object is slightly changed depending on whether a flow passes through the fog server The present invention provides a method of deploying a fog server based on a host location and a traffic pattern in an environment.

The present invention determines a rank of candidate fog devices to which a fog server can be placed and determines the placement of the fog server by using the link bandwidth usage between all the fog devices and the core network, The present invention provides a fog server arrangement method based on a host location and a traffic pattern in a computing environment.

In the present invention, the order of a candidate fog device to which a fog server can be placed is determined, a first fog server is arranged on the basis of a rank, a rank is redefined, The present invention provides a fog server arrangement method based on a host location and a traffic pattern in a fog computing environment for determining whether or not a fog server exists.

The present invention provides a fog computing environment capable of distributing traffic generated on a core network to a local area by arranging a fog server in a corresponding location even if congestion occurs in the case of deploying the fog server based on traffic generated on the network The present invention provides a method of arranging a fog server based on a host location and a traffic pattern.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood by those skilled in the art from the following description.

), 처리 위해 필요한 컴퓨팅 리소스 크기 정보를 갖는 로지컬 플로우(f)를 생성하는 단계;생성된 로지컬 플로우(f)를 이용하여 포그 디바이스와 코어 네트워크 사이 링크(

)의 현재 대역폭 사용량을 구하는 단계;구해진 포그 디바이스와 코어 네트워크 사이 링크 대역폭 사용량을 기준으로 하여 포그 서버가 배치될 수 있는 후보 포그 디바이스의 순서(Rank)를 정하는 단계;구해진 순서(Rank)를 기준으로 포그 서버를 배치하고, 코어 네트워크에서 발생하는 트래픽과 로컬 영역에서 발생하는 트래픽 비율을 기준으로 추가로 포그 서버를 배치하는 단계;를 포함하는 것을 특징으로 한다.In order to achieve the above object, in a fog computing environment according to the present invention, a fog server arrangement method based on a host location and a traffic pattern is represented by a straight line from a data source to a client and includes directions (Src, Dst) Post-bandwidth usage (

), Generating a logical flow ( f ) with computing resource size information required for processing, linking the fog device and the core network using the generated logical flow ( f )

Determining a rank of a candidate foggy device to which the fog server can be placed based on the link bandwidth usage between the obtained fog device and the core network based on the obtained rank; Arranging a fog server, and disposing a fog server in addition to the traffic generated in the core network and the traffic ratio occurring in the local area.

Here, in the case where the fog server is arranged on the basis of the obtained order (Rank), when there is no client adjacent to the fog device, the fog device is excluded from the order and the in- The fog device is excluded from the order.

If there is a sufficient computing resource of Rank [0] in the process of arranging the fog server based on the obtained rank, the fog server is arranged in the fog device corresponding to Rank [0] And determines whether or not to further arrange the fog server.

The fog server arrangement method based on the host location and the traffic pattern in the fog computing environment according to the present invention has the following effects.

First, it is possible to increase the fog computing efficiency by creating a logical flow to obtain the current bandwidth usage of the link between the fog device and the core network and arranging the fog server based on the link bandwidth usage.

Second, by creating a logical flow, it is possible to obtain the current bandwidth usage of the link between the fog device and the core network, and determine the position of the fog server at which position it is preferable to increase the efficiency of the fog server placement.

Third, the order of the candidate fog devices in which the fog server can be arranged is determined, and the object to be added is slightly changed depending on whether the flow passes through the fog server, thereby increasing the fog computing efficiency.

Fourth, it is necessary to determine the order of the candidate fog devices in which the fog server can be placed, to arrange the first fog server based on the rank, to reorder the order, and then to arrange the additional fog server It is possible to reduce the waste of computing resources and improve the network efficiency.

Fifth, if a fog server is arranged based on the traffic generated on the network, even if congestion occurs, the fog server is disposed at the corresponding location, so that the traffic generated on the core network can be distributed to the local area.

) 등의 정보를 담고 있는 로지컬 플로우 구성도
도 3은 네트워크상의 플로우가 지나가는 각 링크의 대역폭을 비율로 나타낸 구성도
도 4는 본 발명에 따른 포그 서버 배치 후보 포그 디바이스의 순위를 나타내는 우선순위(Rank) 도출 과정을 나타낸 플로우 차트
도 5는 포그 서버를 추가 배치하기 전 네트워크상의 모든 플로우 모습과 각 링크 대역폭 비율을 나타낸 구성도
도 6은 포그 서버 배치 및 추가 배치를 위한 과정을 나타낸 플로우 차트1 is a block diagram of a fog computing environment according to the present invention;
FIG. 2 shows the direction (Src, Dst), the bandwidth used

), And so on.
3 is a diagram showing a ratio of the bandwidth of each link to which a flow on a network passes,
FIG. 4 is a flowchart showing a process of deriving a rank indicating a ranking of a fog device candidate fog device according to the present invention.
FIG. 5 is a diagram showing all the flows on the network and the configuration of each link bandwidth ratio before further deploying the fog server.
6 is a flowchart showing a process for fog server placement and further placement

Hereinafter, a preferred embodiment of a fog server arrangement method based on a host location and a traffic pattern in a fog computing environment according to the present invention will be described in detail.

The features and advantages of the fog server placement method based on the host location and traffic pattern in the fog computing environment according to the present invention will be apparent from the detailed description of each embodiment below.

) 등의 정보를 담고 있는 로지컬 플로우 구성도이다.FIG. 1 is a configuration diagram of a fog computing environment according to the present invention. FIG. 2 is a diagram illustrating a direction (Src, Dst)

), And the like.

The present invention generates a logical flow expressing network traffic on an actual network topology to obtain a current bandwidth usage of a link between the fog device and the core network to rank a fog server in which position it should be placed, The fog server deployment is based on traffic patterns.

In the present invention, the order of a candidate fog device to which a fog server can be placed is determined, a first fog server is arranged on the basis of a rank, a rank is redefined, So that it is possible to reduce the waste of computing resources and improve the network efficiency.

)를 기준으로 하여 포그 서버를 배치한다.Algorithms that deploy fog servers based on computing resources require that the device's preference value (

) As a reference.

)을 표시한다. 그리고 나서 디바이스의 자원 사용이 증가하였음(

++)을 표시하고. 그에 따라 선호도 값(

--)은 줄어들었다고 표시한다. When the arrangement satisfies Equation (1) indicating that the resources of the target device are sufficient, the fog server is placed in the corresponding device and the fog server is placed in the device

). Then the resource usage of the device increased (

++). Accordingly, the preference value (

-) indicates that it has shrunk.

In this method, a fog server is determined based on the available computing resources of the fog device.

However, the method of simply deploying based on a computing resource does not reflect the traffic generated on the network, so that a bottleneck phenomenon may cause problems such as delay in service response.

To solve this problem, it would be ideal to have a fog server on all fog devices.

However, placing fog servers everywhere is a waste of computing resources.

Accordingly, the present invention implements a fog server arrangement method that shows performance similar to an ideal placement method while reducing wasting computing resources by reflecting traffic generated on a network.

FIG. 1 illustrates a configuration of a fog computing environment according to an embodiment of the present invention. In FIG. 1, there is shown a foggy computing environment in which a fog device 300 for processing services of the cloud at an edge as an intermediate switch, a client 100 for requesting a service to a fog server, And the order of the candidate fog devices in which the fog server can be arranged based on the link bandwidth usage between the fog device 300 and the core network .

As shown in FIG. 1, the present invention provides a fog server arrangement method that configures a fog computing environment and uses traffic patterns that vary depending on a location of a client and a data source, and selects meaningful information from network traffic generated on an actual topology as a logical flow.

It is expressed as a straight line connecting the client and the data source and contains information about the bandwidth usage before and after processing by the server, the source / destination, and the computing resources required for processing.

That is, the present invention represents a logical flow ( f ) of data traffic on a current network in order to efficiently deploy a fog server based on a traffic pattern on a network in a fog computing environment, as shown in FIG.

), 처리 위해 필요한 컴퓨팅 리소스 크기 정보를 담고 있다.The logical flow is represented by a straight line from the data source to the client and is expressed as direction (Src, Dst), bandwidth usage before / after server

), And computing resource size information required for processing.

Then, the generated logical flow is used to prioritize the fog device to which the fog server is to be placed and to place the fog server in the selected fog device. We also consider how to deploy additional fog servers.

First, the logical flow generation will be described as follows.

Network traffic on a real network topology is represented by several logical flows. This can be used to determine the current bandwidth usage of the link between the fog device and the core network.

The bandwidth usage is then used to rank in which locations fog servers should be deployed.

)의 현재 대역폭 사용량을 비율로 나타낸 예이다.Figure 3 according to the logical flow shows the link between the fog device and the core network

) Is the ratio of the current bandwidth usage.

The fog device ranking for fog server placement will be described as follows.

Here, the order of the candidate fog devices in which the fog server can be placed is determined based on the link bandwidth usage between the fog device and the core network.

Depending on whether or not the flow passes through the server, the objects to be added are slightly different.

4 is a flowchart illustrating a process of deriving a rank indicating a ranking of a fog device candidate fog device according to the present invention.

If there is a fog server on the fog device adjacent to the flow direction Src (S401), the link bandwidth usage between the fog device and the core network is obtained as in Equation (2) (S411)

If there is a fog server on the fog device adjacent to the flow direction Dst (S402), the link bandwidth usage between the fog device and the core network is obtained as in Equation (3) (S412)

In the case where the fog server exists on the fog device adjacent to the flow direction Src and the case where the fog server exists on the fog device adjacent to the flow direction Dst is excluded (S403) The link bandwidth usage between the core networks is calculated (S413)

If one or more fog servers are arranged, the link bandwidth usage between the fog device and the core network is obtained as shown in Equation (5) (S414)

과 Core Network 사이의 링크

의 대역폭 사용량을 구하기 위한 단계이다.In the above procedure, each Fog Device

And Core Network

Is the step for obtaining the bandwidth usage of the network.

별 대역폭 사용량을 기준으로

의 Rank를 정한다.Here,

Based on Star Bandwidth Usage

Of the total.

)에는 해당 Fog Device를 Rank에서 제외한다.In the present invention, when the Fog Server is arranged based on the obtained Rank, when there is no Client adjacent to the Fog Device

), The corresponding Fog Device is excluded from the Rank.

에 이미 Fog Server가 배치되었을 경우에도 Rank에서 제외한다.Also, the corresponding Fog Device

Even if Fog Server is already deployed in the Rank.

The fog server arrangement method will be described as follows.

FIG. 5 is a block diagram showing all flow views and respective link bandwidth ratios on a network before further placement of a fog server, and FIG. 6 is a flowchart showing a procedure for fog server placement and additional placement.

If the candidate fog device is determined to be placed in the previous step, the determined rank value is received (S601) and the actual fog server is placed in the fog device at this stage.

When placing the first fog server based on a predetermined rank, if the computing resources of Rank [0] are sufficient, the fog server is placed in the fog device corresponding to Rank [0].

After determining the rank again, it is judged whether or not to further deploy the fog server.

The additional placement of the fog server proceeds if Equation (6) is satisfied.

는 Fog Device

와 Core Network 사이를 연결하는 링크(

)의 현재 대역폭 사용량이다. 따라서

는 현재 토폴로지 상 flow의 전체 대역폭 사용량을 뜻한다.here

Fog Device

And the Core Network (

) Is the current bandwidth usage. therefore

Refers to the total bandwidth usage of the flow in the current topology.

은 포그 서버가 배치된 포그 디바이스의 수를 뜻하며

이 커질수록 전체 컴퓨팅 리소스 사용량이 증가한다.

은 Fog Server의 추가 배치에 따라 컴퓨팅 리소스 사용량이 증가한 비율만큼은 이론적으로 Core Network의 트래픽이 줄어야 함을 뜻한다(

= 1/2, 1/3, 1/4 ... ).

Means the number of fog devices in which the fog server is located

The greater the overall computing resource usage.

Means that as the amount of computing resource usage increases theoretically, the traffic of the Core Network should decrease as the Fog Server is further deployed

= 1/2, 1/3, 1/4 ...).

은 Fog Device에 Fog Server가 배치될 경우 Fog Device가 있는 local 영역에서 처리되는 트래픽을 나타낸다.

Shows the traffic processed in the local area where the Fog Device is located when the Fog Server is placed in the Fog Device.

보다 클 경우 Fog Server를 추가 배치함을 뜻한다. In summary, Equation (6) shows that when the Fog Server is additionally deployed, the traffic rate in the local network

If it is bigger, it means to deploy Fog Server.

와

가 모두 1이라고 가정했을 때, 도 5에서는 Fog Server가 1개가 이미 배치되어 있을 경우 현재 네트워크상의 대역폭 사용량 합(8)과 추가로 Fog Server가 배치되었을 경우 Fog Device로 분산될

(3 * 2 = 6, [Fog Device 4의 client 수] * [링크를 지나는 횟수])을 비교해볼 수 있다(client가 아닌 data Source가 local에 존재할 경우, [해당 Fog Device의 data Source 수] * [링크를 지나는 횟수]가 분산될 트래픽임). For example,

Wow

Is 1, in FIG. 5, if one Fog Server is already arranged, the sum of the bandwidth usage on the present network (8) and the Fog Device when the Fog Server is further deployed

(3 * 2 = 6, [Number of clients of Fog Device 4] * [Number of times of passing link]) (If data source other than client is local, [The number of times the link passes] is the traffic to be distributed.

은 8이고

은 6이다.In sum,

Is 8

Is 6.

이므로 추가 배치 요건을 만족한다.And if Fog Server is added, the total number of Fog Servers is 2. therefore

So that additional layout requirements are met.

This process is shown in FIG. 6, which shows a process for fog server placement and further placement.

The method of deploying the Fog Server based on the computing resources of the previous Fog Device does not reflect the network traffic. Therefore, if congestion occurs in the network, it can not respond to the congestion, and the response time increases accordingly.

The present invention can distribute the traffic generated on the core network to the local area by arranging the Fog Server at the corresponding location even if the Fog Server is arranged based on the traffic generated on the network. Therefore, it is appropriate to reduce waste of computing resources and improve network efficiency as a way to better reflect the situation of the network.

As described above, it will be understood that the present invention is implemented in a modified form without departing from the essential characteristics of the present invention.

It is therefore to be understood that the specified embodiments are to be considered in an illustrative rather than a restrictive sense and that the scope of the invention is indicated by the appended claims rather than by the foregoing description and that all such differences falling within the scope of equivalents thereof are intended to be embraced therein It should be interpreted.

100. Client 200. Data Source
300. Fog device 400. Core network

Claims (9)

It is represented by a straight line from the data source to the client, and indicates the direction (Src, Dst), before and after the server,

), Generating a logical flow ( f ) with computing resource size information required for processing;
Using the generated logical flow ( f ), the link between the fog device and the core network

Obtaining a current bandwidth usage of the mobile terminal;
Determining a rank of a candidate foggy device to which the fog server can be placed based on the link bandwidth usage between the obtained fog device and the core network;
Disposing a fog server based on the obtained rank, and disposing a fog server additionally on the basis of the traffic generated in the core network and the traffic ratio occurring in the local area. A method of deploying fog servers based on host location and traffic patterns. 2. The method of claim 1, wherein in obtaining the current bandwidth usage,
If there is a fog server on the fog device adjacent to the direction Src of the logical flow

Wherein the link bandwidth usage between the fog device and the core network is obtained by using the host location and traffic pattern in the fog computing environment. 2. The method of claim 1, wherein in obtaining the current bandwidth usage,
If there is a fog server on the fog device adjacent to the direction Dst of the logical flow

Wherein the link bandwidth usage between the fog device and the core network is obtained by using the host location and traffic pattern in the fog computing environment. 2. The method of claim 1, wherein in obtaining the current bandwidth usage,
Except for the case where the fog server exists on the fog device adjacent to the direction Src of the logical flow and the case where the fog server exists on the fog device adjacent to the flow direction Dst

Wherein the link bandwidth usage between the fog device and the core network is obtained by using the host location and traffic pattern in the fog computing environment. 2. The method of claim 1, wherein in obtaining the current bandwidth usage,
If more than one fog server is deployed

Wherein the link bandwidth usage between the fog device and the core network is obtained by using the host location and traffic pattern in the fog computing environment. The method according to claim 1, wherein, in the fixing in which the fog server is arranged based on the obtained rank,
If there is no client adjacent to the fog device, the fog device is excluded from the order,
And if the in-fog server is disposed in the fog device, the fog device is excluded from the order. The fog server arrangement method based on the host location and the traffic pattern in the fog computing environment. The method according to claim 1, wherein, in the process of arranging the fog server on the basis of the obtained rank,
If the computing resources of Rank [0] are sufficient, a fog server is arranged in the fog device corresponding to Rank [0], and it is determined whether or not the fog server is to be further arranged after the order is set again A fog server deployment method based on host location and traffic pattern in a fog computing environment. 8. The method of claim 7, wherein the further placement of the fog server

Is satisfied,
here

Lt; RTI ID = 0.0 &

And the link between the core network (

), ≪ / RTI >

Is the total bandwidth usage of the current topology flow,

Wherein the number of fog servers is a number of fog devices in which the fog server is located. 9. The method of claim 8, wherein when the fog server is additionally disposed,

Wherein the fog server is further configured to allocate a fog server based on the host location and the traffic pattern in the fog computing environment.

Images