KR102235622B1 - Method and Apparatus for Cooperative Edge Caching in IoT Environment - Google Patents

Abstract

Disclosed are a cooperative edge caching method in an IoT environment and a device therefor. According to an embodiment of the present invention, the cooperative edge caching method comprises: an information obtaining step of obtaining request record information from an edge server; a clustering processing step of generating at least one cluster including at least one edge server based on the request record information; and a caching processing step of performing a caching process so that the content for which a caching request of a preset level or higher occurs based on the request record information is stored in the at least one edge server.

Description

[Method and Apparatus for Cooperative Edge Caching in IoT Environment}

The present invention relates to a cooperative edge caching method and apparatus for minimizing content transmission time in consideration of a large-scale caching request in an IoT environment.

The content described in this section merely provides background information on the embodiments of the present invention and does not constitute the prior art.

Among the prior art, various studies are being conducted to improve the efficiency of content transmission by using an edge server in an IoT environment. A technology for such an edge server is described in Korean Patent Registration No. 10-1546199.

In the conventional IoT environment system, a cooperative caching method is proposed to provide a content providing service in an optimized cooperative caching environment. The proposed technology consists of collecting content information, network information, and caching server storage capacity information, generating cooperative caching setting information using the information, and transmitting the information to the caching server.

In the conventional IoT environment system, a plurality of caching servers transmit content to a user device according to network information and content information, thereby reducing the content transmission speed and network traffic.

In a conventional IoT environment system, the storage of a caching server with a storage capacity of C is divided into two regions according to the storage partitioning factor (λ). Popular content is stored in the first area, and content that is not popular or recently accessed by a user is stored in the second area. The content to be cached and the update period are network size (M), network link capacity (r), transmission packet size of user devices and caching server (l _req ), response packet size of caching server and content service provider (l _res ), content It is determined according to the number of user devices connected to the delivery network (S), the number of contents (N _g ) and the size (l). This information is used to set up cooperative caching that minimizes the content download time.

However, in the conventional IoT environment system, a method of minimizing the content download time in a cooperative caching environment has been proposed, but there is a problem that transmission efficiency is significantly reduced when the caching request in the IoT environment is greatly increased. Accordingly, there is a need for a technology to solve this problem.

The present invention relates to a technique of using a cooperative edge server to satisfy a large-scale caching request. The main object of the present invention is to provide a cooperative edge caching method in an IoT environment for improving caching content transmission efficiency and an apparatus therefor. There is this.

According to an aspect of the present invention, a cooperative edge caching method for achieving the above object includes: an information acquisition step of obtaining request record information from an edge server; A clustering processing step of creating at least one cluster including at least one edge server based on the request record information; And a caching processing step of performing a caching process to store content in which a caching request exceeding a preset criterion occurs based on the request record information in at least one edge server.

In addition, according to another aspect of the present invention, a core server for achieving the above object includes: an information acquisition unit for acquiring request record information from an edge server; A clustering processor for generating at least one cluster including at least one edge server based on the request record information; And a caching processing unit that performs caching so that content in which a caching request greater than or equal to a preset criterion occurs based on the request record information is stored in at least one edge server.

As described above, the present invention has the effect of distributing large-scale content requests and reducing a content transmission time for finally fetching caching content.

In addition, the present invention has the effect of increasing the safety of time-sensitive IoT-applications and user experience satisfaction by ensuring a stable content transmission time even in an urban area with a large IoT device and a large floating population.

1 is a block diagram schematically showing a cooperative edge caching network system according to an embodiment of the present invention.
2 is a block diagram schematically showing a cooperative edge processing apparatus according to an embodiment of the present invention.
3 is a diagram for explaining a clustering processing operation according to an embodiment of the present invention.
4 is a flowchart illustrating a clustering operation of a cooperative edge processing method according to an embodiment of the present invention.
5 is a diagram for explaining a caching processing operation according to an embodiment of the present invention.
6 is a flowchart illustrating a caching operation of a cooperative edge processing method according to an embodiment of the present invention.
7 is an exemplary diagram showing a content transmission time according to cooperative edge processing according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In describing the present invention, if it is determined that a detailed description of a related known configuration or function may obscure the subject matter of the present invention, a detailed description thereof will be omitted. In addition, a preferred embodiment of the present invention will be described below, but the technical idea of the present invention is not limited or limited thereto, and may be modified and variously implemented by a person skilled in the art. Hereinafter, a cooperative edge caching method in an Internet of Things (IoT) environment proposed by the present invention and an apparatus therefor will be described in detail with reference to the drawings.

The present invention relates to a technique using a cooperative edge server to satisfy a large-scale caching request occurring in an IoT environment, and aims to improve caching content transmission efficiency. In particular, the present invention is one by using a clustering technique to overcome the limitations of computer resources of edge servers and a caching algorithm that stores content that meets the popularity and request demands of content in several edge servers. It is to solve the problem of latency caused by concentration on the edge server of the network and the problem of delay time caused by it.

The basic environment to which the present invention can be applied may be wireless communication in IoT, and the wireless communication system may include an IoT sensor, a device, and a user device such as a smartphone/tablet. In particular, the present invention can have a great effect when applied in a place such as an urban area where a lot of these devices are concentrated.

1 is a block diagram schematically showing a cooperative edge caching network system according to an embodiment of the present invention.

The cooperative edge caching network system 10 according to this embodiment is configured to include a core server 100, an edge server 110, and a user terminal 120. The cooperative edge caching network system 10 of FIG. 1 is according to an embodiment, and not all blocks shown in FIG. 1 are essential components, and some blocks included in the cooperative edge caching network system 10 in other embodiments These may be added, changed or deleted.

A cooperative edge caching network system 10 proposes a content transmission time optimization technology for cooperative edge caching, and relates to an IoT and edge server environment system.

The edge server 110 according to an embodiment of the present invention refers to a server that is located closer to the user terminal 120 than the cloud as one of the edge server elements and performs caching. Edge server 110, the edge server has significantly lower computer resources than the cloud.

The core server 100 according to an embodiment of the present invention is a server that locally manages these edge servers and performs algorithms for cluster processing and caching processing.

In the cooperative edge caching network system 10 in the IoT environment, caching requests come from a myriad of user terminals 120. In such a case, a large number of caching requests may be accumulated in one edge server, increasing the waiting delay time, and finally reducing the content transmission time.

In the cooperative edge caching network system 10 of the present invention, in order to solve this problem, a plurality of edge servers 110 cooperate with each other to cache a plurality of contents and provide them to the user terminal 120. The user terminal 120 may request content from a cluster to which the nearest edge server 110 belongs.

In order to respond to a myriad of caching requests coming to the cooperative edge server, the core server 100 performs the following algorithm.

The core server 100 processes clustering according to the average transmission time of the edge server 110.

The core server 100 prevents the user terminal 120 requesting the content from fetching the content from the edge server 110 that is excessively far away, and In order to efficiently use the storage space, the edge servers 110 in the region are clustered.

In addition, the core server 100 processes content caching according to the content arrival rate.

)을 계산하고, 변화되는 대기 지연 시간(

)을 캐싱을 처리하는데 사용한다. The core server 100 is based on the caching history from the edge servers 110 in the region, and the content j is transferred to one edge server i among the plurality of edge servers 120 in the cluster. Wait delay time that changes when saved (

), and the varying latency delay (

) Is used to handle caching.

)은 에지 서버(110)의 도착률, 콘텐츠의 인기도, 클러스터내에 저장된 콘텐츠의 수 등에 따라 결정될 수 있다. 코어 서버(100)에서, 에지 서버(110)들의 저장 공간 효율을 위하여 추가적인 캐싱이 콘텐츠 전송 시간을 일정시간 t_cl 이상만큼 감소시킬 수 없을 때 캐싱은 중단된다.In the core server 100, the waiting delay time (

) May be determined according to the arrival rate of the edge server 110, the popularity of the content, the number of content stored in the cluster, and the like. In the core server 100, caching is stopped when additional caching cannot reduce the _{content transmission time by more than a predetermined time t cl} for the storage space efficiency of the edge servers 110.

2 is a block diagram schematically showing a cooperative edge processing apparatus according to an embodiment of the present invention.

The cooperative edge processing apparatus 200 according to the present embodiment includes an information acquisition unit 210, a clustering processing unit 220, and a caching processing unit 230. The cooperative edge processing apparatus 200 of FIG. 2 is according to an embodiment, and not all blocks shown in FIG. 2 are essential components, and some blocks included in the cooperative edge processing apparatus 200 are added in other embodiments. , May be changed or deleted.

The cooperative edge processing apparatus 200 performs an operation of controlling the operation of the cooperative edge server. Here, the cooperative edge processing apparatus 200 controls the operation of the cooperative edge server through clustering processing and caching processing.

In FIG. 2, the cooperative edge processing device 200 is described as being a separate device from the core server 100, but is not necessarily limited thereto, and is included in the same device as the core server 100 or the core server 100. It can be implemented with some components. Hereinafter, components included in the cooperative edge processing apparatus 200 will be described.

The information acquisition unit 210 acquires request record information from the edge server.

The information acquisition unit 210 receives request record information including a request history, a cache history, and the like of a user terminal from edge servers within a boundary (a preset area) of the core server.

The request record information includes information on the transmission delay time (Transmission Delay, t _{c, bc} ) between the user terminal and the edge server and what content has been requested. In addition, the request record information may include information on the popularity of the content (p _j ), the number of contents stored in the cluster (Y _j ), and the arrival rate of the cluster (λ).

The clustering processing unit 220 performs an operation of generating at least one cluster including at least one edge server based on the request record information.

The clustering processing unit 220 creates a cluster by selecting one of the plurality of edge servers.

Thereafter, the clustering processor 220 selects a second edge server connected to the first edge server. Here, the clustering processor 220 selects a second edge server connected to the first edge server by one hop from among a plurality of edge servers connected to the first edge server.

Thereafter, the clustering processor 220 compares the second edge server with a preset cluster addition condition, and when the cluster addition condition is satisfied, adds the second edge server to the cluster.

The clustering processor 220 determines whether to add the second edge server to the cluster based on the cluster addition condition including the first condition and the second condition. Specifically, the clustering processing unit 220 determines that the first condition that the second edge server does not belong to another cluster and the time when a cache hit is made to the second edge server is longer than the time when a cache miss occurs. The second edge server is added to the cluster only when the cluster addition condition including the second condition that a difference by a set predetermined time (t _{cl) is satisfied is satisfied.}

On the other hand, the clustering processing unit 220 selects the first edge server until each of the plurality of edge servers belongs to a predetermined cluster, the operation of selecting the second edge server, and the edge server in the cluster according to the cluster addition condition. Repeatedly perform the adding operation.

Hereinafter, a process of processing clustering will be described in detail.

Each edge server records incoming requests to itself. The request record includes the transmission delay between the user and the edge server and what content was requested.

The core server can calculate _{the delay time (T hit} = t _c,bc +t _bc,c ) when hit from each edge server through the request record received from the edge server. Core server is the calculated T _hit Based on these, the following clustering process is performed. Where, t _c,bc And t _bc,c denotes a transmission delay time between the edge server and the user terminal.

(1) The core server first selects one of the edge servers belonging to its region and creates a cluster.

(2) The core server then selects the edge server (i) connected to the selected edge server by one hop, and adds the edge server to the cluster when the following two conditions are satisfied.

Condition 1: i must not belong to another cluster.

Condition 2: T _{hit +} t _{cl <} T _miss

T _miss Refers to the transmission delay time when a cache miss occurs and content is received from the cloud.In order to prevent edge servers that are too far apart from each other from becoming a cluster, the time when the cache hit is compared to when the cache miss occurs, t _cl The edge server (i) is added to the cluster only when the difference occurs. Here, t _cl may be a threshold for determining the size of the cluster.

(3) The core server repeats steps (1) to (2) until there is no edge server added to the cluster. The core server returns to (1) and repeats the clustering process until all edge servers belong to the cluster.

The caching processing unit 230 performs caching so that content in which a caching request exceeding a preset criterion occurs based on the request record information is stored in at least one edge server.

)을 산출한다. The caching processing unit 230 changes the change waiting time when the content is stored for each of the plurality of edge servers (

) Is calculated.

)을 산출한다. The caching processing unit 230 caches the content in the edge server using the popularity of the content included in the request record information (p _j ), the number of contents stored in the cluster (Y _j ), and the arrival rate of the cluster (λ). Calculate the request arrival rate (Δλ _i ) that changes when Thereafter, the caching processing unit 230 uses the size of the content (b _j ), the bandwidth between the user terminal and the edge server (r _j ), and the arrival rate of the content at the edge server (w _{i, j} ). Calculate the total service time, and use the changed request arrival rate and the total service time of the edge server.

) Is calculated.

_{Thereafter, the caching processing unit 230 calculates an average waiting time (T w} ) of each cluster based on the change waiting time for each of the plurality of edge servers.

Thereafter, the caching processing unit 230 selects a cluster with the lowest average waiting time (T _w ) among the clusters, selects a combination of contents on one edge server among a plurality of edge servers included in the cluster, and caches the contents to be stored. Perform. Here, the caching processing unit 230 checks a storage space for caching the content of the edge server, and if there is a storage space, caching is performed so that the content is stored in the corresponding edge server.

On the other hand, when the content cannot be stored due to the storage space, the caching processing unit 230 performs caching so that the content is stored in another edge server by combining the content with another edge server with a lower waiting time after the edge server in the cluster. .

Hereinafter, a process of processing caching will be described in detail.

After clustering, the core server performs a caching algorithm that stores high-demand content in one or more edge servers in order to reduce the waiting time caused by large-scale caching requests.

For this, the core server first calculates a request arrival rate that is changed when the new content j is cached in the edge server i. Here, the request arrival rate may be calculated using the popularity (p _j ) of the content (j), the number of contents (j) stored in the cluster (Y _j ), and the arrival rate of the cluster (λ). The request arrival rate can be calculated by [Equation 1]. The variables in [Equation 1] can be obtained through the request record received in the process prior to the clustering process.

In addition, the core server uses the size of the content (b _j ), the bandwidth between the user terminal and the edge server (r _i ), and the arrival rate of the content at the edge server (w _{i, j} ) to provide the entire service of the edge server (i). Calculate the total service time. Here, the total service time may be calculated through [Equation 2].

)을 얻을 수 있으며, 변경된 후의 도착률(

)와 [수학식 2]를 사용하여 콘텐츠 복제본이 클러스터에 하나 추가되었을 때 변화되는 대기시간(waiting time,

)을 산출할 수 있다. 여기서, 코어 서버는 [수학식 3]을 통해 변화되는 대기시간(

)을 산출할 수 있다. Finally, the core server uses the changed arrival rate derived from [Equation 1], and the arrival rate after the change (

), and the arrival rate after the change (

) And [Equation 2] to change the waiting time when one content replica is added to the cluster.

) Can be calculated. Here, the core server is the waiting time changed through [Equation 3] (

) Can be calculated.

In [Equation 3], λ _i means the arrival rate before change.

The core server performs caching by selecting the content (j) and the edge server (i) when the change of the waiting time is the greatest through the changed waiting time obtained through [Equation 3].

The core server performs a caching process based on the equation of [Equation 3].

)들을 산출하고, 이에 따른 클러스터의 평균 대기시간(T_w)을 산출한다. First, for all the edge servers (i) and content (j), the core server has the varying latency of each edge server (i) when the content (j) is stored in the edge server (i) (

) Are calculated, and the average waiting time (T _w ) of the cluster is calculated accordingly.

At this time, the core server selects the _{edge server (i) and content (j) that induces the lowest average latency (T w} ), and then checks whether the edge server (i) has enough space to store the content (j). Check whether or not.

The core server stores the content j in the edge server i when there is enough space to store the content j in the edge server i. If there is no space to store the content j on the edge server i, the core server selects the next best combination of the edge server and content j. The second-best edge server may be an edge server that induces _{the lowest average latency (T w} ) after the edge server (i) in the cluster.

If the core server cannot store the contents j in the edge server for all combinations, the caching process is stopped. That is, the core server performs caching by selecting the best combination of the edge server (i) and content (j) until caching is stopped due to the storage space of all edge servers in the cluster.

3 is a diagram for explaining a clustering processing operation according to an embodiment of the present invention.

_{Referring to FIG. 3, the core server 100 receives information of a delay time (T hit} ) between the edge server 110 and the user terminals 120 from the edge servers 110 and performs clustering based on the information.

During the clustering process, the core server 100 selects one edge server 120 to create the cluster 310, and then the edge server that satisfies _{T hit +} t _{cl <} T _{miss among the edge servers connected to the cluster by 1-hop.} Add 120 to the cluster 310.

If there is no edge server satisfied, the core server 100 returns to the process of selecting one edge server, selects another edge server, and repeats the operation of adding the selected other edge servers to the cluster 310. .

4 is a flowchart illustrating a clustering operation of a cooperative edge processing method according to an embodiment of the present invention.

The core server 100 creates a cluster (v) by selecting one of the edge servers belonging to its region (S410).

The core server 100 then selects the edge server i connected to the selected edge server by one hop (S420).

The core server 100 checks whether the edge server i does not belong to another cluster (S430).

If the edge server (i) does not belong to another cluster, the core server 100 compares the transmission delay time when the content is received from the cloud due to a cache miss and the delay time when the cache hits, and a predetermined time t _cl It is checked whether there is a difference by as much as (S440, S450).

The core server 100 compares the transmission delay time when the content is received from the cloud due to a cache miss and the delay time when the cache hits, and only when a difference by _{a certain amount of time t cl occurs.} It is added to the cluster (v) (S460).

The core server 100 repeats steps S420 to S460 until there is no edge server added to the cluster (S470).

In FIG. 4, it is described that each step is sequentially executed, but is not limited thereto. In other words, since it may be applicable to changing and executing the steps illustrated in FIG. 4 or executing one or more steps in parallel, FIG. 4 is not limited to a time-series order.

The cooperative edge processing method according to the present embodiment illustrated in FIG. 4 may be implemented as an application (or program) and recorded on a recording medium that can be read by a terminal device (or computer). An application (or program) for implementing the cooperative edge processing method according to the present embodiment is recorded, and the recording medium that can be read by the terminal device (or computer) is all types of recording devices that store data that can be read by the computing system. Or media.

5 is a diagram for explaining a caching processing operation according to an embodiment of the present invention.

Referring to FIG. 5, each edge server 110 has a caching record (Cache History) of a caching request received from the user terminal 120 to them. The caching record contains information about what data is requested and how much.

), 데이터의 인기도 등을 계산할 수 있다. 이 때, 코어 서버(100)는

를 이용하여 콘텐츠(j)에 대한 인기도를 계산할 수 있다. The core server 100 receives the caching record from each edge server 110 and receives the request arrival rate (λ _i ) of each edge server, and the service time (

), the popularity of the data, etc. can be calculated. At this time, the core server 100

The popularity of the content j can be calculated using.

The core server 100 can derive [Equation 1] and [Equation 2] using the calculated arrival rate, service time, and popularity, and finally derive [Equation 3] to [Equation 3]. ], the edge server (i) and content (j) are selected and caching is performed when the result value (changing waiting time) is the largest.

6 is a flowchart illustrating a caching operation of a cooperative edge processing method according to an embodiment of the present invention.

)들을 산출한다(S610).The core server 100 for all of the edge servers (i) and content (j), when the content (j) is stored in the edge server (i), the varying latency of each edge server (i) (

) Are calculated (S610).

)들을 기반으로 클러스터의 평균 대기시간(T_w)을 산출하고, 가장 낮은 평균 대기시간(T_w)을 유도하는 에지 서버(i) 및 콘텐츠(j)를 선택한다(S620).The core server 100 has the varying latency of each edge server (i) (

), the average waiting time (T _w ) of the cluster is calculated, and the edge server (i) and the content (j) that induce the lowest average waiting time (T _{w) are selected (S620).}

The core server 100 checks whether there is enough space in the edge server i to store the content j (S630).

If there is enough space in the edge server i to store the content j, the core server 100 stores the content j in the edge server i and performs caching (S640).

If there is no space to store the content (j) in the edge server (i), the core server 100 selects a combination of another edge server and the content (j) of the next best position, or selects the content (j) for all combinations. If it is not possible to store in, the caching process is stopped (S650).

In FIG. 6, it is described that each step is sequentially executed, but is not limited thereto. In other words, since it is possible to change and execute the steps illustrated in FIG. 6 or execute one or more steps in parallel, FIG. 6 is not limited to a time-series order.

The cooperative edge processing method according to the present embodiment illustrated in FIG. 6 may be implemented as an application (or program) and recorded on a recording medium that can be read by a terminal device (or computer). An application (or program) for implementing the cooperative edge processing method according to the present embodiment is recorded, and the recording medium that can be read by the terminal device (or computer) is all types of recording devices that store data that can be read by the computing system. Or media.

7 is an exemplary diagram showing a content transmission time according to cooperative edge processing according to an embodiment of the present invention.

Simulation was performed to show the effect of the content caching method according to the edge server clustering and the varying waiting time proposed by the cooperative edge caching network system 10 according to the present invention. It was confirmed that the delay time for content transmission was reduced compared to caching stored in the server.

In order to evaluate a content caching method that minimizes latency through the cooperative caching algorithm proposed by the cooperative edge caching network system 10, the content transmission delay time was measured by varying the number of caching requests per second (λ) arriving at the edge server.

7 shows the content transmission delay time for each algorithm according to the number of caching requests.

Compared to the conventional algorithm marked as original, it can be seen that when the algorithm marked as proposed in the present invention is used, a lower content transmission delay time is shown in all sections.

In FIG. 7, the delay time is compared in situations where the degree of the Zipf variable is additionally given differently as 0.3 (Fig. 7 (a)), 0.7 (Fig. 7 (b)), and 1.0 (Fig. 7 (c)). I did. In the result of FIG. 7, as the zipf variable increases, the caching content transmission time increases, and it can be seen that the proposed technology transmits the content at a lower content transmission time than the prior art in all zipf variable change intervals.

The above description is merely illustrative of the technical idea of the embodiments of the present invention, and those of ordinary skill in the technical field to which the embodiments of the present invention belong to, various modifications and modifications without departing from the essential characteristics of the embodiments of the present invention. Transformation will be possible. Accordingly, the embodiments of the present invention are not intended to limit the technical idea of the embodiment of the present invention, but to explain the technical idea, and the scope of the technical idea of the embodiment of the present invention is not limited by these embodiments. The scope of protection of the embodiments of the present invention should be interpreted by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of the rights of the embodiments of the present invention.

10: cooperative edge caching network system
100: core server 110: edge server
120: user terminal
200: cooperative edge processing device
210: information acquisition unit 220: clustering processing unit
230: caching processing unit

Claims (11)

In the method of caching the cooperating edge in the core server,
An information acquisition step of obtaining request record information from an edge server;
A clustering processing step of creating at least one cluster including at least one edge server based on the request record information; And
A caching processing step of performing a caching process so that content in which a caching request exceeding a preset criterion occurs based on the request record information is stored in at least one edge server
Collaborative edge caching method comprising a. The method of claim 1,
The clustering processing step,
A first edge server selection step of selecting one of a plurality of edge servers to create a cluster;
A second edge server selection step of selecting a second edge server connected to the first edge server; And
Cluster adding step of comparing the second edge server with a preset cluster addition condition, and adding the second edge server to the cluster when the cluster addition condition is satisfied
Collaborative edge caching method comprising a. The method of claim 2,
The second edge server selection step,
Collaborative edge caching method, characterized in that selecting the second edge server connected to the first edge server and one hop from among a plurality of edge servers connected to the first edge server. The method of claim 2,
The cluster addition step,
_{The first condition that the second edge server does not belong to another cluster and a predetermined time (t cl} ) when the time when a cache hit is made to the second edge server is a time when a cache miss occurs (t cl) And adding the second edge server to the cluster only when the cluster addition condition including a second condition that a difference must occur by as much as is satisfied. The method of claim 4,
The clustering processing step,
The cooperative edge caching method, characterized in that repeatedly performing the first edge server selection step, the second edge server selection step, and the cluster addition step until each of the plurality of edge servers belongs to a predetermined cluster. The method of claim 1,
The caching processing step,
Change waiting time when the above content is stored for each of the plurality of edge servers (

A change waiting time calculation step of calculating );
An average waiting time calculation step of calculating an average waiting time _{(T w} ) of each cluster based on the change waiting time for each of a plurality of edge servers;
Caching to perform caching so that the content is stored by selecting a cluster with the lowest average waiting time (T _w ) among the clusters, selecting a combination of the content on one edge server among a plurality of edge servers included in the cluster Steps to perform
Collaborative edge caching method comprising a. The method of claim 6,
The step of calculating the waiting time for change,
Using the popularity of the content (p _j ), the number of contents stored in the cluster (Y _j ), and the arrival rate of the cluster (λ), the request arrival rate that is changed when the content is cached in the edge server (Δλ) _{calculating i} );
The total service time of the edge server is calculated using the size of the content (b _j ), the bandwidth between the user terminal and the edge server (r _j ), and the arrival rate of the content at the edge server (w _{i,j ).} Calculating; And
Using the changed request arrival rate and the total service time of the edge server, the changed waiting time (

Steps to calculate)
Collaborative edge caching method comprising a. The method of claim 6,
The step of performing the caching,
Check the storage space for caching the content of the edge server, and if there is the storage space, store the content,
When the content cannot be stored due to the storage space, caching is performed so that the content is stored in the other edge server by combining the content with another edge server having a lower waiting time after the edge server in the cluster. Collaborative edge caching method, characterized in that. In a core server interworking with a plurality of edge servers in a cooperative edge caching network,
An information acquisition unit that acquires request record information from an edge server;
A clustering processor for generating at least one cluster including at least one edge server based on the request record information; And
A caching processing unit that performs caching so that content in which a caching request exceeding a preset criterion occurs based on the request record information is stored in at least one edge server.
Core server comprising a. The method of claim 9,
The clustering processing unit,
Selecting one of the plurality of edge servers to create a cluster, and selecting a second edge server connected to the first edge server,
And comparing the second edge server with a preset cluster addition condition, and adding the second edge server to the cluster when the cluster addition condition is satisfied. The method of claim 9,
The caching processing unit,
Change waiting time when the above content is stored for each of the plurality of edge servers

_{), and calculates the average waiting time (T w} ) of each cluster based on the change waiting time for each of the plurality of edge servers,
Selecting a cluster with the lowest average waiting time (T _w ) among the clusters, selecting a combination of the contents in one of a plurality of edge servers included in the cluster, and performing caching so that the contents are stored. Core server characterized by.

Images