CN107682416B - Broadcast-storage network-based fog computing architecture content collaborative distribution method and application system - Google Patents

Broadcast-storage network-based fog computing architecture content collaborative distribution method and application system Download PDF

Info

Publication number
CN107682416B
CN107682416B CN201710851632.2A CN201710851632A CN107682416B CN 107682416 B CN107682416 B CN 107682416B CN 201710851632 A CN201710851632 A CN 201710851632A CN 107682416 B CN107682416 B CN 107682416B
Authority
CN
China
Prior art keywords
content
node
controller
fog
server
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201710851632.2A
Other languages
Chinese (zh)
Other versions
CN107682416A (en
Inventor
杨鹏
扈晓娜
刘旋
李幼平
张长江
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southeast University filed Critical Southeast University
Priority to CN201710851632.2A priority Critical patent/CN107682416B/en
Publication of CN107682416A publication Critical patent/CN107682416A/en
Application granted granted Critical
Publication of CN107682416B publication Critical patent/CN107682416B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/55—Push-based network services
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/1023—Server selection for load balancing based on a hash applied to IP addresses or costs
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/56—Provisioning of proxy services
    • H04L67/568—Storing data temporarily at an intermediate stage, e.g. caching

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Information Transfer Between Computers (AREA)

Abstract

The invention discloses a broadcast and storage network-based collaborative distribution method and an application system for fog computing architecture content. The invention utilizes the broadcast storage network to improve the content active push capability of the traditional fog computing architecture and reduce the consumption of the content active push capability to the internet bandwidth, adopts the centralized control idea of the software defined network SDN, and provides a multi-granularity cooperative storage mechanism which gives consideration to the broadcast content and the Web cache content; on the basis, all nodes in the domain are maintained and controlled in a centralized way through the controller, and an efficient content forwarding strategy is formulated for the mist server in the domain; and finally, realizing the content collaborative distribution of multiple nodes in the domain through the controller. The invention can effectively enhance the content active pushing capability of the fog computing architecture, reduce the time delay of the user request and improve the user experience.

Description

Broadcast-storage network-based fog computing architecture content collaborative distribution method and application system
Technical Field
The invention relates to a broadcast-storage network-based fog computing architecture content collaborative distribution method and an application system, which can reduce the response delay of a terminal user request and improve the user experience and belong to the technical field of the Internet.
Background
The cloud computing can overcome the defects of resource monopolization, low data center service density, low system resource utilization rate and the like commonly existing in the traditional application system, but the cloud computing still has the defects of incapability of effectively supporting delay sensitive application, large consumption of network bandwidth and the like. For this reason, Cisco (Cisco) proposed the concept of Fog Computing (Fog Computing), which extends network Computing characterized by cloud Computing from the hub to the edge of the network and uses Fog Computing to describe an intermediate state between cloud Computing and terminal Computing. The fog computing architecture defines an open system-level architecture for fog computing based solutions that can provide computing, storage, control, and networking functions near data generation sources along cloud-to-object. The importance of the edge calculation is fully demonstrated by the proposal of the fog calculation. However, in the current fog computing architecture, content distribution mainly takes the form of edge caching, which is severely limited by the form of passive "user pulling", and lacks the capability of content active pushing.
The broadcast storage network is a novel network supporting network content nationwide sharing, and the basic idea is that hot content crawled from the internet and corresponding content metadata are actively pushed and stored to the edge of the network through broadcasting, so that a user terminal can directly obtain the hot content nearby conveniently. The broadcast storage network can improve the content access experience of users, shares the backbone network flow of the Internet, and has the characteristics of strong active push strength, no consumption of Internet bandwidth in content distribution and the like. In consideration of the defects that the active push capability is weak, the content distribution consumes the internet bandwidth and the like in the current fog computing architecture, the defects are difficult to overcome by adopting the traditional technology, and the broadcasting network has the advantages that the defects can be effectively overcome. On one hand, however, the industry currently has no implementation scheme of a broadcast-storage network-based fog computing architecture; on the other hand, there are many difficulties in constructing a fog computing architecture based on a broadcast storage network, such as storage coordination between Web cache content and broadcast content, efficient management and control of a fog server, a content coordinated distribution method with linkage of multiple nodes, and the like. The invention integrates the idea of the broadcast storage network into a fog computing framework, designs a fog computing framework based on the broadcast storage network, adopts the centralized control concept of a Software Defined Network (SDN) based on the novel fog computing framework, provides a multi-granularity content cooperative distribution method considering broadcast content and Web cache content, and realizes push-pull combined content push.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the problems and the defects in the prior art, the invention provides a broadcast storage network-based fog computing architecture content collaborative distribution method and an application system. The method adopts a broadcast storage network of 'broadcast distribution + ubiquitous storage' to assist a fog computing architecture to actively push contents, and on the basis, the high-efficiency collaborative distribution of the contents is realized by using a centralized control idea of a Software Defined Network (SDN), and specifically comprises the steps of multi-granularity content collaborative storage, controller-dominated information processing, multi-node collaborative distribution in a domain and the like. The method can reduce the consumption of the fog computing architecture on the internet bandwidth, reduce the request time delay of the edge user and improve the user experience.
The technical scheme is as follows: the method is different from the existing method in that the broadcast storage network is utilized to improve the content active push capability of the traditional fog computing architecture and reduce the consumption of the traditional fog computing architecture on the internet bandwidth, and the high-efficiency cooperation of broadcast content distribution and Web cache content distribution is realized by utilizing the centralized control idea of a Software Defined Network (SDN). The invention mainly comprises 3 steps:
and step 1, cooperatively storing the multi-granularity content. Firstly, the idea of the broadcast storage network is integrated into a fog computing architecture, the fog computing architecture based on the broadcast storage network is constructed, the architecture comprises a fog server, a controller, a plurality of user terminals, an internet content acquisition server and a broadcast content source, the plurality of fog servers are controlled by the controller in a centralized manner, the whole architecture adopts the broadcast content source to distribute full content texts and corresponding UCLs to the plurality of fog servers, the full content texts are acquired from the internet by the internet content acquisition server, the internet content acquisition server generates the corresponding UCLs, and meanwhile, the user terminals also receive the UCLs distributed by the broadcast. The UCL is metadata describing content, and mainly includes information such as a title, a summary, keywords, and a category to which the content belongs. The fog computing architecture simultaneously comprises two storage modes: web caching and broadcast content storage. The Web cache is responsible for storing fine-grained content (generally, constituent elements in a Web page, such as pictures and music embedded in the Web page), and the broadcast content storage is responsible for storing full-text Web page content with coarser granularity. In a fog computing architecture based on a broadcast storage network, two content storage modes with different granularities are maintained simultaneously, so that the cooperative storage of two content granularities of webpage constituent elements and the full-text webpage content is realized. First, a broadcast content store is selected with a heat value greater than a given threshold PsAnalyzing the full text of the webpage content to obtain fine-grained webpage elements, and then selecting the fine-grained webpage elementsThere are Web page elements that match the white list (the list of Web sites customized by the system administrator) and are stored in the Web cache. The heat value of the full text of the webpage content is calculated according to the formula (1):
Figure BDA0001412130200000021
wherein, PiRepresents the full-text heat of the webpage content in the ith timing period and satisfies P1＝0；NiRepresenting the number of times that the webpage content is accessed in the ith timing cycle; n is a radical ofiA weight coefficient of>1。
And 2, processing information dominated by the controller. The fog computing architecture based on the broadcast-storage network is composed of a plurality of domains, and a controller is configured in each domain and controls a plurality of fog servers in the domains. When the content in the fog server is increased, deleted and updated, the fog server reports corresponding modification information to the controller; once a terminal request reaches the fog server, the fog server reports the specific information of the request to the controller. The controller dynamically adjusts the index information table and the service information table in the local domain according to the information reported by each fog server. The index information table mainly comprises information such as nodes, ports, URLs (uniform resource locators) and Hash values, and the service information table mainly comprises information such as source identifiers, node identifiers, Hash values, types and hit times.
The controller periodically calculates a class weight ratio of each fog server according to the service information table, and maintains a class weight ratio information table for each fog server, which is used as a basis for the fog server to receive the broadcast content and the UCL next time. The user requests the content corresponding to the UCL through the UCL. Assume a total number of categories of all content is NcThen, the weighted value of any class c is calculated according to equation (2):
Figure BDA0001412130200000031
wherein R iscIndicating that during the current cycle, the UCL request issued by the user is of class cA number of times
Figure BDA0001412130200000032
Is the total number of all user UCL requests in this period. The class weight ratio is
Figure BDA0001412130200000033
The controller also determines the intra-domain forwarding policy for the fog server by calculating the forwarding value of each user UCL request. The forwarding value is calculated according to equation (3):
f ═ α W + β H + (1- α - β) R formula (3)
Wherein
Figure BDA0001412130200000034
W, H and R are weights, node hop counts and hit rates respectively calculated by the controller according to the request parameters, x and y are constant parameters representing the proportion of W, H, R three factors, and the specific values can be dynamically adjusted according to actual requirements, but the condition x + y is less than or equal to 0.
And 3, performing intra-domain multi-node cooperative distribution. The mist computing architecture based on the broadcast storage network adopts a centralized control idea to realize the cooperation of nodes in the domain, so that the mist server does not need to maintain complex data information and control information and only needs to be responsible for realizing service logic, and the method is quicker and more efficient compared with completely distributed processing. When the content resource requested by the user is locally stored in the fog server, the fog server directly responds to the content request of the terminal user; and if the requested content resource is not locally stored in the fog server, the fog server directly sends an intra-domain collaborative query request to the controller. When the controller receives a content query request sent by the fog server, for other fog servers which are in the range controlled by the controller and store the requested content, the fog server with the highest forwarding value is selected as the optimal forwarding node by calculating the forwarding values of the fog servers for the UCL request of the user, and the optimal node information is returned to the requesting node. After receiving the information returned by the controller, the requesting node directly forwards the content request to the IP and the port corresponding to the optimal node, acquires the content resource and returns the content resource to the terminal user.
A content collaborative distribution application system based on a broadcast-storage network fog computing architecture for executing the method mainly comprises the following nodes: the system comprises a crawler server for collecting hot content from the Internet and indexing content metadata UCL, a broadcast server for broadcasting the hot content, a fog server node (node identifiers are A, B and C respectively), a controller node and a client (node identifier is D). The crawler server collects hot web pages from the Internet, performs semantic analysis on the hot web pages to obtain corresponding UCLs, and then transmits the hot web pages and the UCLs to the broadcast server; the broadcast server distributes the hot web pages and the UCL to each fog server through broadcasting and radiation.
For the hot web pages received by broadcasting, the fog server firstly calculates the heat value P of each hot web page according to the formula (1)i. Next, P is judgediWhether the condition is satisfied: pi≥PsAnd if the webpage content meets the requirement, webpage element analysis is carried out on the popular webpage to obtain webpage elements with fine granularity, and then all the webpage elements matched with the white list are selected. Then, for the selected web page element, the SHA1 algorithm is used to Hash the original URL of the web page element. And finally, storing the webpage elements to a corresponding directory of the Web cache according to the Hash value.
And each fog server can actively report corresponding information to the controller when receiving the hot webpage and the UCL sent by the broadcast server. And the UCL content metadata information is displayed on a user terminal browsing interface, a user accesses the corresponding full-text webpage or webpage elements by clicking the UCL, the contents are supposed to be stored in each fog server, and once a user request reaches the fog server, the fog server reports the hit condition of the request to the controller. The contents stored in the fog server are added, deleted or updated, and the fog server also reports corresponding modification information to the controller. And the controller will maintain the index information table and the service information table within the domain.
The controller will calculate each fog clothes according to the formula (2) in each timing periodWeight value W for each category c of content in servercFurther, the class weight ratio is obtained
Figure BDA0001412130200000041
(3) And multi-node cooperative distribution in the domain. If a UCL request issued by the user corresponds to category 1, the URL corresponds to a Hash value of 4fghadf234hg23 bc. The UCL request reaches one of the fog servers a, and if the fog server a misses, the fog server a sends an intra-domain query request to the controller.
When the controller receives the intra-domain content query request sent by the fog server A, the controller firstly finds out the set Node of all nodes storing the requested content from the Node identification table according to the Hash value1,Node2,...,Nodei}. Then, for each Node in the set NodeiAcquiring Node from class weight table according to class value of contentiClass weight W ofiAccording to the node identification, the hop count H between the nodes is obtained from the topology information tableiAnd calculates Node according to the service information tableiHit rate R on requested categoryiFinally, calculate Node according to formula (3)iForward value of Fi. After the forwarding values of all the nodes in the set nodes are calculated, the controller sorts the nodes from high to low according to the forwarding values, selects the Node arranged at the top as the optimal forwarding Node, and returns the optimal forwarding Node information to the request Node. And after receiving the information returned by the controller, the user directly sends the UCL request to the optimal forwarding node, and finally the optimal forwarding node returns the webpage content full text or the webpage elements corresponding to the UCL request of the user to the terminal user.
Has the advantages that: compared with the prior art, the method for cooperatively distributing the content of the fog computing architecture based on the broadcast storage network and the application system have the following advantages that:
1. the broadcast storage network with strong active push force and good user experience is integrated into the fog computing architecture, and the broadcast distribution and ubiquitous storage are adopted to assist the fog computing architecture in actively pushing the content, so that the defect of the traditional fog computing architecture in the aspect of content active push capability can be effectively improved, and the consumption of the fog computing architecture on the internet bandwidth can be reduced.
2. The centralized control idea of the software defined network SDN is combined with the fog computing architecture, so that the high-efficiency cooperation of broadcast content distribution and Web cache content distribution is realized, the complexity of content cooperative distribution is simplified, the distribution performance of the fog computing architecture is improved, and the user experience can be improved.
Drawings
Fig. 1 is a diagram of a broadcast-deposit network-based fog computing architecture.
Detailed Description
The present invention is further illustrated by the following examples, which are intended to be purely exemplary and are not intended to limit the scope of the invention, as various equivalent modifications of the invention will occur to those skilled in the art upon reading the present disclosure and fall within the scope of the appended claims.
The method is different from the existing method in that the broadcast-storage network is utilized to improve the content active push capability of the traditional fog computing architecture and reduce the consumption of the traditional fog computing architecture on the internet bandwidth, and the high-efficiency cooperation of broadcast content distribution and Web cache content distribution is realized by utilizing the centralized control idea of a Software Defined Network (SDN). The invention mainly comprises 3 steps:
and step 1, cooperatively storing the multi-granularity content. Firstly, the idea of the broadcast storage network is integrated into a fog computing architecture, the fog computing architecture based on the broadcast storage network is constructed, the architecture comprises a fog server, a controller, a plurality of user terminals, an internet content acquisition server and a broadcast content source, the plurality of fog servers are controlled by the controller in a centralized manner, the whole architecture adopts the broadcast content source to distribute full content texts and corresponding UCLs to the plurality of fog servers, the full content texts are acquired from the internet by the internet content acquisition server, the internet content acquisition server generates the corresponding UCLs, and meanwhile, the user terminals also receive the UCLs distributed by the broadcast. The UCL is metadata describing content and mainly comprises content targetsTopic, abstract, keyword, and category to which the content belongs. The fog computing architecture simultaneously comprises two storage modes: web caching and broadcast content storage. The Web cache is responsible for storing fine-grained content (generally, constituent elements in a Web page, such as pictures and music embedded in the Web page), and the broadcast content storage is responsible for storing full-text Web page content with coarser granularity. In a fog computing architecture based on a broadcast storage network, two content storage modes with different granularities are maintained simultaneously, so that the cooperative storage of two content granularities of webpage constituent elements and the full-text webpage content is realized. First, a broadcast content store is selected with a heat value greater than a given threshold PsAnd analyzing the full text of the webpage content to obtain fine-grained webpage elements, then selecting all webpage elements matched with a white list (a website list customized by a system manager) from the fine-grained webpage elements, and storing the webpage elements in a Web cache. The heat value of the full text of the webpage content is calculated according to the formula (1):
Figure BDA0001412130200000061
wherein, PiRepresents the full-text heat of the webpage content in the ith timing period and satisfies P1＝0；NiRepresenting the number of times that the webpage content is accessed in the ith timing cycle; n is a radical ofiA weight coefficient of>1。
And 2, processing information dominated by the controller. The fog computing architecture based on the broadcast-storage network is composed of a plurality of domains, and a controller is configured in each domain and controls a plurality of fog servers in the domains. When the content in the fog server is increased, deleted and updated, the fog server reports corresponding modification information to the controller; once a terminal request reaches the fog server, the fog server reports the specific information of the request to the controller. The controller dynamically adjusts the index information table and the service information table in the local domain according to the information reported by each fog server. The index information table mainly comprises information such as nodes, ports, URLs (uniform resource locators) and Hash values, and the service information table mainly comprises information such as source identifiers, node identifiers, Hash values, types and hit times.
The controller periodically calculates a class weight ratio of each fog server according to the service information table, and maintains a class weight ratio information table for each fog server, which is used as a basis for the fog server to receive the broadcast content and the UCL next time. The user requests the content corresponding to the UCL through the UCL. Assume a total number of categories of all content is NcThen, the weighted value of any class c is calculated according to equation (2):
Figure BDA0001412130200000071
wherein R iscIndicates the number of times that the UCL request issued by the user is of the category c in the current period, and
Figure BDA0001412130200000072
is the total number of all user UCL requests in this period. The class weight ratio is
Figure BDA0001412130200000073
The controller also determines the intra-domain forwarding policy for the fog server by calculating the forwarding value of each user UCL request. The forwarding value is calculated according to equation (3):
f ═ α W + β H + (1- α - β) R formula (3)
Wherein
Figure BDA0001412130200000074
W, H and R are weights, node hop counts and hit rates respectively calculated by the controller according to the request parameters, x and y are constant parameters representing the proportion of W, H, R three factors, and the specific values can be dynamically adjusted according to actual requirements, but the condition x + y is less than or equal to 0.
And 3, performing intra-domain multi-node cooperative distribution. The mist computing architecture based on the broadcast storage network adopts a centralized control idea to realize the cooperation of nodes in the domain, so that the mist server does not need to maintain complex data information and control information and only needs to be responsible for realizing service logic, and the method is quicker and more efficient compared with completely distributed processing. When the content resource requested by the user is locally stored in the fog server, the fog server directly responds to the content request of the terminal user; and if the requested content resource is not locally stored in the fog server, the fog server directly sends an intra-domain collaborative query request to the controller. When the controller receives a content query request sent by the fog server, for other fog servers which are in the range controlled by the controller and store the requested content, the fog servers with the highest forwarding value are selected as the optimal forwarding nodes by calculating the forwarding values of the fog servers for the UCL request of the user, and the optimal node information is returned to the requesting nodes. After receiving the information returned by the controller, the requesting node directly forwards the content request to the IP and the port corresponding to the optimal node, acquires the content resource and returns the content resource to the terminal user.
When the method is implemented, a content collaborative distribution application system conforming to the broadcast-storage network-based fog computing architecture shown in fig. 1 is constructed first. Without loss of generality, the content collaborative distribution application system is assumed to mainly comprise 7 nodes: the system comprises 1 crawler server for collecting hot content from the Internet and indexing content metadata UCL, 1 broadcast server for broadcasting the hot content, 3 fog server nodes (node identifiers are A, B and C respectively), 1 controller node and 1 client (node identifier is D). Assume that a crawler server has collected 100 popular web pages from the internet, and performs semantic parsing on the popular web pages to obtain 100 corresponding UCLs, and then transmits 100 popular web pages and 100 UCLs to a broadcast server. The broadcast server distributes 100 popular web pages and 100 UCLs to 3 fog servers through broadcasting and radiation, and the specific implementation process of the invention is as follows:
(1) multi-granularity content storage collaboration. For 100 hot web pages received by broadcasting, the 3 fog servers firstly calculate the heat value P of each hot web page according to the formula (1)i. Next, P is judgediWhether the condition is satisfied: pi≥PsIf yes, webpage element analysis is carried out on the hot webpage to obtain webpage elements with fine granularity (such as pictures embedded in the webpage)Music, etc.) and then select out all web page elements that match the white list. Then, for the selected web page element, the SHA1 algorithm is used to Hash the original URL of the web page element. And finally, storing the webpage elements to a corresponding directory of the Web cache according to the Hash value.
(2) Controller-directed information processing. When receiving 100 hot web pages and 100 UCLs sent by the broadcast server, each fog server can actively report corresponding information to the controller. And 100 pieces of UCL content metadata information are displayed on a user terminal browsing interface, a user accesses corresponding full-text web pages or web page elements by clicking UCL, the contents are supposed to be stored in the fog servers A, B and C, and once a user request arrives at the fog server, the fog server reports the hit condition of the request to the controller. In addition, in the dynamic process, the content stored in the fog server may be added, deleted or updated, and the fog server may also report corresponding modification information to the controller. The controller will maintain the index information table and the service information table in the domain, and the structures of the index information table and the service information table are respectively shown in table 1 and table 2:
table 1 indexing information table
Node point Port(s) URL Hash value
A 8080 www.sina.com.cn 52df5gd2fg5dvbn6
B 9009 www.baidu.com 4fghadf234hg23bc
┆ ┆ ┆ ┆
C 9009 www.jd.com ac45b5b52c2gfg0
Table 2 service information table
Source identification Node identification Hash value Type (B) Hit in
D A 1g34h121h32lc145 1 0
D B efgl19034hg2abcd 5 1
┆ ┆ ┆ ┆ ┆
D C 5acb632fg4987de 4 2
The controller will calculate the weight value W of each category c of the content in the 3 fog servers according to the formula (2) in each timing periodcFurther, the class weight ratio is obtained
Figure BDA0001412130200000091
(3) And multi-node cooperative distribution in the domain. Assuming that a UCL request issued by the user corresponds to category 1, the URL corresponds to a Hash value of 4fghadf234hg23 bc. The UCL request reaches fog server a, and if fog server a misses, fog server a sends an intra-domain query request to the controller.
When the controller receives the intra-domain content query request sent by the fog server A, the controller firstly finds out the set Node of all nodes storing the requested content from the Node identification table according to the Hash value1,Node2,...,Nodei}. Then, for each Node in the set NodeiAcquiring Node from class weight table according to class value of contentiClass weight W ofiAccording to the node identification, the hop count H between the nodes is obtained from the topology information tableiAnd calculates Node according to the service information tableiHit rate R on requested categoryiFinally, calculate Node according to formula (3)iForward value of Fi. After the forwarding values of all the nodes in the set nodes are calculated, the controller sorts the nodes from high to low according to the forwarding values, selects the Node arranged at the top as the optimal forwarding Node, and returns the optimal forwarding Node information to the request Node. And after receiving the information returned by the controller, the user directly sends the UCL request to the optimal forwarding node, and finally the optimal forwarding node returns the webpage content full text or the webpage elements corresponding to the UCL request of the user to the terminal user.

Claims (9)

1. A fog computing architecture content collaborative distribution method based on broadcast storage network is characterized in that the method mainly comprises 3 steps:
step 1, multi-granularity content collaborative storage; firstly, a fog computing architecture based on a broadcast storage network is constructed, and the fog computing architecture simultaneously comprises two storage modes: web cache and broadcast content storage, realize the cooperative storage of two kinds of granularity contents of webpage constituent element and webpage content full text;
step 2, processing information dominated by the controller; the fog computing architecture based on the broadcast storage network is composed of a plurality of domains, and each domain is provided with a controller which controls a plurality of fog servers in the domain;
step 3, multi-node cooperative distribution in the domain; when the content resource requested by the user is locally stored in the fog server, the fog server directly responds to the content request of the terminal user; if the requested content resource is not stored locally in the fog server, the fog server directly sends an intra-domain collaborative query request to the controller; when the controller receives a content query request sent by the fog server, the controller calculates the optimal node information of the requested content stored in the domain under jurisdiction according to the corresponding forwarding strategy, and returns the optimal node information to the requesting node; after receiving the information returned by the controller, the requesting node directly forwards the content request to the IP and the port corresponding to the optimal node, acquires the content resource and returns the content resource to the terminal user.
2. The method for the cooperative distribution of the content in the mist computing architecture based on the broadcast-storage network as claimed in claim 1, wherein the Web cache is responsible for storing fine-grained contents of the Web page constituent elements, and the broadcast content storage is responsible for storing full-text contents with coarser-grained contents.
3. The mist computing architecture based on broadcast storage network content cooperative distribution method according to claim 1, characterized in that in the mist computing architecture based on broadcast storage network, two different granularity content storage manners are maintained simultaneously, thereby realizing cooperative storage of two granularity contents of web page constituent elements and web page content full texts; first, a broadcast content store is selected with a heat value greater than a given threshold PsAnalyzing the full text of the webpage content to obtain webpage elements with fine granularity, then selecting all the webpage elements matched with the white list from the webpage elements, and storing the webpage elements in a Web cache; the heat value of the full text of the webpage content is calculated according to the formula (1):
Figure FDA0002413005940000011
wherein, PiRepresents the full-text heat of the webpage content in the ith timing period and satisfies P1＝0；NiRepresenting the number of times that the webpage content is accessed in the ith timing cycle; n is a radical ofiA weight coefficient of>1。
4. The cooperative distribution method for the contents in the mist computing architecture based on the broadcast-storage network as claimed in claim 1, wherein when the contents in the mist server are added, deleted and updated, the mist server reports the corresponding modification information to the controller; once a terminal request reaches the fog server, the fog server reports the specific information of the request to the controller; the controller dynamically adjusts the index information table and the service information table in the local domain according to the information reported by each fog server; the index information table comprises nodes, ports, URLs and Hash values, and the service information table comprises source identifiers, node identifiers, Hash values, types and hit times.
5. The mist computing architecture content collaborative distribution method based on the broadcast-storage network as claimed in claim 1, wherein the controller periodically calculates a category weight ratio of each mist server according to a service information table, wherein the service information table includes a source identifier, a node identifier, a Hash value, a type and a number of hits; and maintaining a category weight ratio information table for each fog server, wherein the category weight ratio information table is used as a basis for the fog server to receive the broadcast content and UCL next time; a user requests content corresponding to UCL through UCL; assume a total number of categories of all content is NcThen, the weighted value of any class c is calculated according to equation (2):
Figure FDA0002413005940000021
wherein R iscIndicates the number of times that the UCL request issued by the user is of the category c in the current period, and
Figure FDA0002413005940000022
the total number of all user UCL requests in the period; then the class weight ratio is W ═ W1:W2:…WNc。
6. The mist computing architecture content collaborative distribution method based on the broadcast-storage network as claimed in claim 1, wherein the controller decides an intra-domain forwarding policy of the mist server by calculating a forwarding value of each user UCL request; the forwarding value is calculated according to equation (3):
f ═ α W + β H + (1- α - β) R formula (3)
Wherein
Figure FDA0002413005940000023
W, H and R are weight, node hop count and hit rate calculated by the controller according to the request parameters, x and y are represented by W, H, RA constant parameter of the proportion of each factor.
7. A content collaborative distribution application system based on a broadcast storage network fog computing architecture is characterized by mainly comprising the following nodes: the system comprises a crawler server, a broadcast server, a fog server node, a controller node and a client, wherein the crawler server collects hot content from the Internet and indexes content metadata UCL (content metadata); the crawler server collects hot web pages from the Internet, performs semantic analysis on the hot web pages to obtain corresponding UCLs, and then transmits the hot web pages and the UCLs to the broadcast server; the broadcasting server distributes the hot webpage and the UCL to each fog server in a radiation mode through broadcasting;
the fog server firstly calculates the heat value P of each hot webpage for the hot webpages received by broadcastingi(ii) a Wherein the heat value is calculated according to formula (1):
Figure FDA0002413005940000031
wherein, PiRepresents the full-text heat of the webpage content in the ith timing period and satisfies P1＝0；NiRepresenting the number of times that the webpage content is accessed in the ith timing cycle; n is a radical ofiA weight coefficient of>1; next, P is judgediWhether the condition is satisfied: pi≥PsIf yes, webpage elements of the popular webpage are analyzed to obtain fine-grained webpage elements, and then all webpage elements matched with the white list are selected; then, for the selected webpage elements, the Hash value of the original URL of the webpage elements is calculated by using the SHA1 algorithm; and finally, storing the webpage elements to a corresponding directory of the Web cache according to the Hash value.
8. The system according to claim 7, wherein each fog server actively reports the corresponding information to the controller when receiving the hit web page and the UCL sent from the broadcast server; the UCL content metadata information is displayed on a user terminal browsing interface, a user accesses corresponding full-text web pages or web page elements by clicking UCL, the contents are supposed to be stored in each fog server, and once a user request reaches the fog server, the fog server reports the hit condition of the request to the controller; if the content stored in the fog server is increased, deleted or updated, the fog server can report corresponding modification information to the controller; the controller maintains the index information table and the service information table in the domain; the index information table comprises nodes, ports, URLs and Hash values, and the service information table comprises source identifiers, node identifiers, Hash values, types and hit times.
9. The system according to claim 7, wherein the controller calculates a weight value W for each category c of the content in each fog server in each timing cyclecFurther, the class weight ratio W is obtained1:W2:…WNc(ii) a Wherein WcThe calculation is performed according to equation (2):
Figure FDA0002413005940000032
wherein R iscIndicates the number of times that the UCL request issued by the user is of the category c in the current period, and
Figure FDA0002413005940000033
the total number of all user UCL requests in the period;
if a UCL request sent by a user corresponds to a category, the UCL request reaches one of the fog servers A, and the fog server A does not hit the fog server A, the fog server A sends an intra-domain query request to the controller; when the controller receives the intra-domain content query request sent by the fog server A, the controller firstly finds out the set Node of all nodes storing the requested content from the Node identification table according to the Hash value1,Node2,...,Nodei}; however, the device is not suitable for use in a kitchenThereafter, for each Node in the set NodeiAcquiring Node from class weight table according to class value of contentiClass weight W ofiAccording to the node identification, the hop count H between the nodes is obtained from the topology information tableiAnd calculates Node according to the service information tableiHit rate R on requested categoryiFinally, calculate NodeiForward value of Fi(ii) a The service information table comprises a source identifier, a node identifier, a Hash value, a type and hit times, and the forwarding value F is calculated according to a formula (3):
f ═ α W + β H + (1- α - β) R formula (3)
Wherein
Figure FDA0002413005940000041
W, H and R are weights, node hop counts and hit rates respectively calculated by the controller according to the request parameters, and x and y are constant parameters representing the proportion of W, H, R three factors; after the forwarding values of all the nodes in the set nodes are calculated, the controller sorts the nodes from high to low according to the forwarding values, selects the Node arranged at the top as the optimal forwarding Node, and returns the optimal forwarding Node information to the request Node; and after receiving the information returned by the controller, the user directly sends the UCL request to the optimal forwarding node, and finally the optimal forwarding node returns the webpage content full text or the webpage elements corresponding to the UCL request of the user to the terminal user.
CN201710851632.2A 2017-09-19 2017-09-19 Broadcast-storage network-based fog computing architecture content collaborative distribution method and application system Active CN107682416B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710851632.2A CN107682416B (en) 2017-09-19 2017-09-19 Broadcast-storage network-based fog computing architecture content collaborative distribution method and application system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710851632.2A CN107682416B (en) 2017-09-19 2017-09-19 Broadcast-storage network-based fog computing architecture content collaborative distribution method and application system

Publications (2)

Publication Number Publication Date
CN107682416A CN107682416A (en) 2018-02-09
CN107682416B true CN107682416B (en) 2020-05-19

Family

ID=61137383

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710851632.2A Active CN107682416B (en) 2017-09-19 2017-09-19 Broadcast-storage network-based fog computing architecture content collaborative distribution method and application system

Country Status (1)

Country Link
CN (1) CN107682416B (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108156267B (en) * 2018-03-22 2020-12-29 山东大学 Method for improving website access time delay by using cache in fog computing architecture
CN109167842B (en) * 2018-10-24 2020-04-03 清华大学 Content distribution and push business service system and method based on mixed broadcast mode
CN109802998B (en) * 2018-12-28 2021-09-17 上海无线通信研究中心 Game-based fog network cooperative scheduling excitation method and system
CN110457575B (en) * 2019-07-08 2023-08-18 视联动力信息技术股份有限公司 File pushing method, device and storage medium
CN112287102B (en) * 2019-08-29 2024-04-16 北京京东尚科信息技术有限公司 Data mining method and device
CN110601812B (en) * 2019-09-17 2020-06-30 电子科技大学 Privacy protection encrypted data query method based on fog assistance
US11481259B2 (en) 2020-01-07 2022-10-25 International Business Machines Corporation Distributing computation workloads based on calculated compute gravity within differing computing paradigms

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105794268A (en) * 2013-09-19 2016-07-20 思科技术公司 High-speed mobile broadband network access via programmed tracking of a sequence of wireless broadband data links
CN106357743A (en) * 2016-08-29 2017-01-25 北京邮电大学 Fog computing network service transmission method based on grading caching

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170048308A1 (en) * 2015-08-13 2017-02-16 Saad Bin Qaisar System and Apparatus for Network Conscious Edge to Cloud Sensing, Analytics, Actuation and Virtualization

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105794268A (en) * 2013-09-19 2016-07-20 思科技术公司 High-speed mobile broadband network access via programmed tracking of a sequence of wireless broadband data links
CN106357743A (en) * 2016-08-29 2017-01-25 北京邮电大学 Fog computing network service transmission method based on grading caching

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Fog computing and its role in the internet of things;Flavio Bonomi 等;《MCC workshop on mobile cloud computing》;20120831;第13-15页 *
一种播存网络环境下的UCL协同过滤推荐算法;顾梁 等;《计算机研究与发展》;20151231;第52卷(第2期);第477-第485页 *
播存网络体系结构普适模型及实现模式;杨鹏 等;《电子学报》;20150531;第43卷(第5期);第974-978页 *

Also Published As

Publication number Publication date
CN107682416A (en) 2018-02-09

Similar Documents

Publication Publication Date Title
CN107682416B (en) Broadcast-storage network-based fog computing architecture content collaborative distribution method and application system
US8001146B2 (en) Routing XML queries
Lv et al. RISC: ICN routing mechanism incorporating SDN and community division
Romano et al. A neural network proxy cache replacement strategy and its implementation in the Squid proxy server
US20140280606A1 (en) Method and Apparatus for Content Management
CN104717304B (en) A kind of CDN P2P content optimizations select system
CN103428267B (en) A kind of wisdom caching system and the method distinguishing user preferences dependency thereof
CN103152396A (en) Data placement method and device applied to content distribution network system
Chunlin et al. Dynamic cooperative caching strategy for delay-sensitive applications in edge computing environment
Zhao et al. Dynamic replica creation strategy based on file heat and node load in hybrid cloud
Paltoglou et al. Collection-integral source selection for uncooperative distributed information retrieval environments
Rao et al. STAIRS: Towards efficient full-text filtering and dissemination in DHT environments
Li et al. Efficient cooperative cache management for latency-aware data intelligent processing in edge environment
Zhang et al. DENA: An intelligent content discovery system used in named data networking
Reshadinezhad et al. An efficient adaptive cache management scheme for named data networks
Horincar et al. Best-effort refresh strategies for content-based rss feed aggregation
Liu et al. An overview of world wide web caching
Sarkar et al. Content delivery networks: Insights and recent advancement
CN106790421B (en) ICN two-step caching method based on community
Wang et al. Low-cost web service discovery based on distributed decision tree in P2P environments
Xu et al. An adaptive per-application storage management scheme based on manifold learning in information centric networks
Sarkar et al. Population-based clustering to enhance the utilization of surrogate in Content Delivery Networks
Han et al. Energy-efficiency aware cooperative caching strategy for content-centric networks
Chang et al. An Information‐Centric Network Caching Method Based on Popularity Rating and Topology Weighting
Gao et al. Collaborative Caching in P2P Streaming Systems

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant