KR102346417B1 - Method for selecting leader controller node in decentralized software-defined network - Google Patents

Abstract

The present invention relates to a method for selecting a leader controller in a decentralized software-defined network. The method includes the steps of: calculating a stability value for mobility of each of a plurality of controller nodes; generating a two-layer virtual loop in and out of a plurality of domains based on the calculated stability value; and synchronizing information about each node in and out of the plurality of domains based on the two-layer virtual loop. Accordingly, a load applied to a reader can be reduced.

Description

METHOD FOR SELECTING LEADER CONTROLLER NODE IN DECENTRALIZED SOFTWARE-DEFINED NETWORK

The present invention relates to a method for selecting a leader controller in a distributed software-defined network, and in particular, a distributed software-defined network capable of improving synchronization efficiency and processing performance by selecting an efficient controller leader in consideration of the mobility condition of the SDN controller. It relates to the method of selecting a leader controller in

SDN (Software Defined Networking, hereinafter SDN) technology refers to a technology that manages all devices in a network by an intelligent central management system. The SDN technology was proposed to solve the high cost incurred in network management and operation, etc. By having the control operation related to packet processing performed by the existing hardware type network equipment itself be handled by the controller provided in the form of software instead. It has the advantage of being able to develop and provide various functions compared to the existing network structure. That is, the SDN technology has a structure that separates the control plane and the data plane, unlike the existing network structure.

On the other hand, in the case of the existing SDN technology, when performing synchronization between SDN controllers, the mobility of the SDN controller is not considered, so there are problems in synchronization efficiency and processing performance.

Open Networking Foundation, "OpenFlow Specification 1.3.0" (2012.6.25.)

An object of the present invention is to provide a method for selecting a leader controller in a distributed software-defined network, which can improve synchronization efficiency and processing performance by selecting an efficient controller leader in consideration of the mobility condition of the SDN controller.

However, the problems to be solved by the present invention are not limited to the problems described above, and other problems may exist.

A method for selecting a leader controller in a distributed software-defined network according to an aspect of the present invention for solving the above-described problems includes calculating a stability value for the mobility of each of a plurality of controller nodes; generating a two-layer virtual loop in and out of a plurality of domains based on the calculated stability value; and synchronizing information about each node in and out of a plurality of domains based on the two-layer virtual loop.

In some embodiments of the present invention, the generating of a two-layer virtual ring inside and outside a plurality of domains based on the calculated stability value includes: generating a lower-level virtual ring within a domain based on the stability value; and generating an upper layer virtual ring in which leader controller nodes (hereinafter, leader nodes) in the lower layer virtual ring located in neighboring domains among the plurality of domains are set as neighbors.

In some embodiments of the present invention, the calculating of the stability value for the mobility of each of the plurality of controller nodes includes the number of past connected nodes at time t-1 based on one hop and the number of current connected nodes at time t. calculating the number of duplicate connection nodes by comparing them; and calculating the stability value based on a ratio of the number of redundantly connected nodes to the current number of connected nodes.

In some embodiments of the present invention, the generating of the lower layer virtual ring in the domain based on the stability value includes: generating a virtual ring by arranging the calculated stability values for each controller node in the domain in descending order; setting a controller node having the largest stability value in the virtual loop as a leader node; determining, at the leader node, a candidate controller node (hereinafter, a candidate node) and a following controller node (hereinafter, a following node); and configuring the lower-level virtual loop including the leader node, candidate node, and follower node.

In some embodiments of the present invention, the step of determining a candidate node and a follower node in the leader node includes: calculating a first average value of stability values for a plurality of nodes included in the domain; calculating a second average value of stability values for nodes having stability values greater than or equal to the first average value; and setting a node having a stability value equal to or greater than the second average value as a candidate node and setting the remaining nodes as follower nodes.

In some embodiments of the present invention, when the leader node has a stability value less than the second average value according to the movement of the leader node, converting the leader node to a follower node; and setting a candidate node having a next-order stability value among the candidate nodes as a leader node.

According to some embodiments of the present invention, when the candidate node is converted to a leader node, a new leader node transmitting a leader node change message to a leader node of another domain; and re-performing, by the new leader node, a process of creating a two-layer virtual ring in the domain again.

In some embodiments of the present invention, when the candidate node has a stability value less than the first average value according to the movement of the candidate node, converting the candidate node to a follower node; a step in which a leader node in a domain simultaneously performs a function of a candidate node as the candidate node is converted into a follower node; and reconfiguring the candidate node and the following node.

In some embodiments of the present invention, when the following node leaves the corresponding domain according to the movement of the following node, neighboring nodes of the following node transmit failure information to a candidate node; and re-performing, by the candidate node, the process of creating a lower layer virtual ring of the domain by reflecting the failure information.

In some embodiments of the present invention, the step of synchronizing information on each node in and out of a plurality of domains based on the two-layer virtual loop includes: a candidate node in the domain including mobility information and connectivity information of each follower node collecting domain information; and transmitting, by the leader node in the domain, the domain information to a leader node in another domain.

In some embodiments of the present invention, when there are a plurality of candidate nodes in the domain, each of the following nodes selects the closest candidate node and transmits the mobility information and the connectivity information, and among the plurality of candidate nodes, the stability value is large. A candidate node may form and manage the lower layer virtual loop.

According to some embodiments of the present invention, a method comprising: receiving, by a predetermined node (hereinafter, a nearest node) located at a nearest distance from a new node created in a domain; information on the number of past connected nodes and the number of currently connected nodes of the new node; calculating, by the nearest node, a stability value of the new node and transmitting it to a candidate node in the domain; updating, by the candidate node, a lower layer virtual loop based on the stability value of the new node; transmitting the updated lower-level virtual ring information to a leader node; and transmitting, by the leader node, the updated domain information to a leader node in another domain.

In some embodiments of the present invention, the step of the candidate node updating the lower hierarchical virtual loop based on the stability value of the new node comprises: each node to be configured as a neighboring node of the new node; It may include transmitting the updated lower-level virtual ring information to the new node.

A computer program according to another aspect of the present invention for solving the above problems is combined with a computer that is hardware to execute a method for selecting a reader controller in the distributed software defined network, and is stored in a computer readable recording medium.

Other specific details of the invention are included in the detailed description and drawings.

First, the existing distributed SDN causes a large cost because communication of all SDN nodes is exchanged. However, according to an embodiment of the present invention, since domain information is exchanged between leader nodes, communication overhead is higher than that of the existing distributed SDN method. has the advantage of being reduced.

Second, in the existing leader node synchronization method, communication overhead and delay may occur because the leader node performs all intra/inter-domain communication. However, according to an embodiment of the present invention, failure of the leader or candidate node may occur. When this occurs, the candidate node can perform the role of the leader node, and the leader node can perform the role of the candidate node, so that self-organization is possible in maintaining the virtual loop structure in a dynamic environment.

Third, in one embodiment of the present invention, since each leader with a different role for each of a plurality of domains exists, availability is increased compared to a cluster structure in which one existing leader node is responsible for all intra/inter-domain communication, and , it has the effect of reducing the load on the leader.

Fourth, nodes in the general domain do not quickly recognize failures and state changes of other nodes, but in an embodiment of the present invention, when failures and state changes of the following nodes occur using a virtual ring structure, the nodes connected to the virtual ring Through communication between a neighbor node and a candidate node, a failure can be recognized quickly and communication can be restored.

Fifth, there is an advantage in that self-organization between nodes is possible without using a separate server in a wirelessly connected distributed SDN environment, and thus the network structure can be easily expanded.

Sixth, the candidate node synchronizes information within the domain, and the leader node synchronizes information between domains, so the communication load is low and the data transmission speed is fast. Also, the existing unidirectional loop structure is vulnerable to node failure. I have a problem that In order to solve this problem, an embodiment of the present invention can flexibly cope with a failure by maintaining information of a bidirectional neighbor node of the following node.

Finally, in one embodiment of the present invention, a hierarchical structure through leader node selection is used together with a virtual loop topology to maintain communication of wirelessly connected distributed SDN nodes in a dynamic environment. To reduce the communication load of nodes in the domain, a leader is selected based on the stability value and managed using a virtual loop structure to cope with node failure. In addition, it is possible to reduce time and resources required to perform synchronization between domains by using a hierarchical structure that separates leader/candidate/following nodes.

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

The accompanying drawings are provided to help understanding of the present embodiment, and provide embodiments together with detailed description. However, the technical features of the present embodiment are not limited to specific drawings, and features disclosed in the drawings may be combined with each other to form a new embodiment.
1 is a diagram for explaining the configuration of a software-defined network in an embodiment of the present invention.
2 is a diagram for explaining a distributed software-defined network system model in an embodiment of the present invention.
3 is a flowchart of a method for selecting a reader controller according to an embodiment of the present invention.
4 is a diagram illustrating an example of calculating a stability value based on a node connectivity table according to an embodiment of the present invention.
5 is a diagram illustrating an example of selecting a leader node based on a stability value in an embodiment of the present invention.
6 is a view for explaining the content of forming an upper layer virtual ring between leader nodes in an embodiment of the present invention.
7 is a diagram for explaining a synchronization process in a two-layer virtual ring.
8 is a diagram for explaining the contents of changing the lower layer virtual ring according to the addition of a new node in an embodiment of the present invention.
9 is a diagram for explaining a process according to a case in which a change occurs in a stability value of a leader node in an embodiment of the present invention.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only the present embodiments allow the disclosure of the present invention to be complete, and those of ordinary skill in the art to which the present invention pertains. It is provided to fully understand the scope of the present invention to those skilled in the art, and the present invention is only defined by the scope of the claims.

The terminology used herein is for the purpose of describing the embodiments and is not intended to limit the present invention. In this specification, the singular also includes the plural unless specifically stated otherwise in the phrase. As used herein, “comprises” and/or “comprising” does not exclude the presence or addition of one or more other components in addition to the stated components. Like reference numerals refer to like elements throughout, and "and/or" includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various elements, these elements are not limited by these terms, of course. These terms are only used to distinguish one component from another. Accordingly, it goes without saying that the first component mentioned below may be the second component within the spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used herein will have the meaning commonly understood by those of ordinary skill in the art to which this invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless specifically defined explicitly.

Hereinafter, a method for selecting a reader controller in a distributed software defined network according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram for explaining the configuration of a software-defined network in an embodiment of the present invention.

Referring to FIG. 1 , a software defined network includes a controller 100 , a network device 200 , and a host 300 . In this case, the network device 200 and the host 300 may be referred to as a node, and a link means a connection between two nodes.

The controller 100 functions to manage the network equipment 200 , and centrally manages and controls the plurality of network equipment 200 . Specifically, the controller 100 is equipped with software that functions such as topology management, packet processing-related path management, link discovery, and packet flow flow management. It can be implemented in the form

The network device 200 performs a function of processing a packet under the control of the controller 100 . In this case, an example of the network equipment 200 may be a mobile communication base station, a base station controller, a gateway equipment, a wired network switch, a router, and the like. However, for convenience of description, the following description will focus on the case where the network equipment 200 is an OpenFlow switch, and the same reference numerals will be used. However, the network equipment 200, that is, the switch in the present invention is not necessarily limited to only the open flow switch, and is a concept including the existing legacy switch.

In a software-defined network, the controller 100 and the network equipment 200 must exchange information with each other, and the open flow protocol is widely used as a protocol for this purpose. That is, the OpenFlow protocol is a standard standard for communication between the controller 100 and the network device 200 .

More specifically, the network equipment 200 is largely divided into a software layer and a hardware layer.

The software layer exchanges information with the controller 100 through a secure channel. The secure channel is a communication channel between the network device 200 and the remote controller 100 , and information exchanged between the controller and the network device 200 is encrypted.

In the hardware layer, a flow table defines and processes a packet and includes statistical information related to the packet. The flow table consists of flow rules that define packet processing, and the flow rules are added and modified by a flow-mod message generated by the controller 100 and transmitted to the network device 200 . Or it can be deleted. The network device 200 processes the packet with reference to the flow table. On the other hand, the flow rule should be understood to mean a network policy applied by the controller 100 in a software-defined network in a normal position in the industry. That is, the flow rule is a flow rule applied when the controller 100 is connected, and a flow entry in which packet processing is centrally controlled by the controller 100 is defined.

The flow table includes packet header information defining a flow (Match Fields), action information defining packet processing (Action), and statistical information for each flow (Stats). Here, each row constituting the flow table is referred to as a flow entry.

The host 300 refers to a terminal corresponding to a lower layer of the network equipment 200 , and may be used to collectively refer to a client and a server. The host 300 may generate a packet for sending to another host through a software-defined network, and transmit the packet to the network device 200 through a port of a network interface.

For example, when the first host 300a wants to send a packet to the second host 300b, the first host 300a first generates a packet to be sent, and connects the packet to the first host 300a. It is transmitted to the network device 200 . When the network device 200 has a flow table regulating the processing of the corresponding packet, the network device 200 processes the packet according to the regulation.

However, when the network device 200 does not have a flow table related to the packet, the network device 200 transmits a packet-in message informing the controller 100 of the packet inflow.

The controller 100 generates a flow mode message regulating the processing of packets according to the operator's policy or a preset algorithm and transmits it to the network device 200, and the network device 200 reads the flow table changed by the flow mode message. Refer to and process the corresponding packet.

Meanwhile, the SDN structure according to the prior art has a problem in that it is difficult to guarantee scalability according to an increase in the network size by centrally configuring the control plane. In order to solve this control plane scalability problem, a logically centralized but physically distributed SDN controller structure consisting of a plurality of domains is used.

The SDN architecture with a distributed controller divides the management network into clusters, which are several sub-networks, and each controller manages the switches of the cluster. However, in order to synchronize the information of the switches in the cluster collected by the distributed controller with other controllers, a large traffic load may be generated, which acts as a factor degrading the processing performance of the entire network system. Therefore, a controller reader technique to alleviate the traffic load in distributed SDN controller synchronization has been proposed.

The controller leader selection technique aims to reduce traffic load and improve stability and efficiency by selecting a leader controller among multiple distributed SDN controllers in one domain and synchronizing the leaders.

In particular, in the case of an SDN controller with mobility, it is important to select a controller with relatively low mobility as a leader and perform synchronization because synchronization between controllers must occur closely. For example, if a controller leader is randomly selected, the mobility of the controller is not taken into account correctly, resulting in decreased synchronization efficiency and processing performance between controllers. Accordingly, an inaccurate controller selection procedure causes a traffic load of the controller due to the movement of the distributed controller, and has a problem in that the accuracy of synchronization is lowered.

In order to solve this problem, an embodiment of the present invention aims to select an efficient controller leader in consideration of the mobility condition of the SDN controller. In particular, the present invention proposes a Mobility-Aware Leader Controller Selection (MLCS) method for determining a suitable controller for controller synchronization.

That is, according to an embodiment of the present invention, a stability value of the controller is calculated based on a change in the topology table, and a leader and a candidate controller are determined based on the stability value. Thereafter, efficient communication between controller nodes is performed by forming a 2-tier virtual ring within/between each domain.

Through such selection of a leader controller and communication between neighboring controllers connected through a virtual ring, an embodiment of the present invention can reduce communication overhead between all controllers.

In addition, self-organization between controllers is required as the insertion/deletion of controller nodes occurs frequently in a mobile distributed SDN environment. At the same time, it provides a fault-tolerant technology in the dynamic environment of the controller.

Hereinafter, a method for selecting a reader controller in a distributed software defined network (hereinafter, a method for selecting a reader controller) according to an embodiment of the present invention will be described with reference to FIGS. 2 to 9 .

2 is a diagram for explaining a distributed software-defined network system model in an embodiment of the present invention. 3 is a flowchart of a method for selecting a reader controller according to an embodiment of the present invention.

A distributed software-defined network system model connected to a wireless network according to an embodiment of the present invention is configured to include a plurality of controller nodes. Each controller node moves with a random speed and direction, and each node has the number of nodes connected in the past (hereinafter referred to as the number of past connected nodes, N _p ) and the number of currently connected nodes (hereinafter referred to as the current number of nodes) based on 1-hop. The number of connected nodes, N _c ), and the number of redundantly connected nodes at the previous and current time points (hereafter, the number of redundantly connected nodes, N _d ) are stored and managed in the database.

Each controller node calculates its own Stability Value based on the current number of connected nodes ( N _c ) and the number of redundantly connected nodes ( N _{c ) in the table.} At this time, node information of one node at time t-1 and node information updated at time t are stored in a table.

Thereafter, the node with the largest stability value is selected as a leader controller node (hereinafter, leader node), and the stability values are arranged in descending order to form a lower virtual ring. In addition, each domain information is synchronized by forming an upper virtual ring between leader nodes in each domain.

Hereinafter, an embodiment of the present invention will be described in more detail.

An embodiment of the present invention first calculates a stability value for the mobility of each of a plurality of controller nodes (S110).

In a wireless environment, distributed SDN controller nodes C _{d and n} are composed of a plurality of domain sets ( D ) and node sets ( N ). In step S110, in a dynamic environment in which the controller nodes move randomly, whenever there is a table update from time t-1 to time t, the number of past connected nodes at time t-1 ( N _p ) and the current connection node at time t By comparing the number ( N _c ), the number of redundantly connected nodes ( N _d ) is calculated. And based on the ratio of the number of redundantly connected nodes ( N _d ) to the current number of connected nodes ( N _c ), a stability value m _d,n for the mobility of the controller node is calculated. The initial stability value is set to 0, and the stability value is also updated when there is an update of the table over time.

4 is a diagram illustrating an example of calculating a stability value based on a node connectivity table according to an embodiment of the present invention.

으로 계산된다. 각 컨트롤러 노드는 자신의 테이블 업데이트 시점에 맞춰 안정성 값을 산출한다.Referring to FIG. 4 , the number of past connected nodes ( N _p ) of the controller nodes C _{1 and 6} is 13, the current number of connected nodes ( N _c ) is 12, and the number of redundantly connected nodes ( N _d ) is 10. Therefore, the stability values m _{1, 6} are

is calculated as Each controller node calculates a stability value according to its own table update time.

At this time, when the mobility of the node is high, not only the table is updated frequently, but also the node list of the table is frequently changed. Therefore, the stability value increases as there is no change in the node according to the table update, which means that the mobility value of the corresponding node is low and stable. On the other hand, when mobility is high, table node changes occur frequently. This means that the number of duplicate nodes in the previous and current tables is small, and since the stability value is calculated low, mobility is high.

Next, based on the calculated stability value, a two-tier virtual ring in and out of a plurality of domains is generated ( S120 ).

Since each distributed node is connected to each other via wireless links, there is no fixed infrastructure. Therefore, a topology is needed for the network to keep communication self-organized. To this end, in an embodiment of the present invention, a hierarchical structure is formed through leader node selection, and a topology suitable for a dynamic environment is configured by combining the circular topology.

In one embodiment, in the present invention, a lower virtual ring within a domain is created based on a stability value, and a leader node in a lower layer virtual ring located in a neighboring domain among a plurality of domains is set as a neighbor to each other. You can create an upper virtual ring.

5 is a diagram illustrating an example of selecting a leader node based on a stability value in an embodiment of the present invention.

The process of delivering the number of past connected nodes ( N _p ), the current number of connected nodes ( N _c ), and the number of redundantly connected nodes ( N _d ) to each node in the domain since the stability values of all controller nodes in the domain are initially 0 carry out Thereafter, the calculated stability values for each controller node in the domain are arranged in descending order to sequentially create virtual rings.

At this time, the controller node having the largest stability value in the virtual loop is set as the leader node. When a leader node is set, the leader node performs a process of selecting a candidate node.

)을 산출한다. 그리고 식 2와 같이 제1 평균값 이상의 안정성 값을 같는 노드들에 대한 안정성 값의 제2 평균값(

)을 재계산한다.In order to select a candidate node, first, as in Equation 1 in the domain d, the first average value (

) is calculated. And as shown in Equation 2, the second average value (

) is recalculated.

[Equation 1]

[Equation 2]

Then, a node having a stability value equal to or greater than the second average value is selected as candidate nodes, and unselected nodes are set as follower nodes. In this case, the candidate node transmits to the following nodes that it is a candidate node in the corresponding domain by broadcasting.

Nodes with a stability value higher than the average value have relatively little movement in the corresponding domain. However, when the mobility of nodes within a domain is generally high, even if the stability value is higher than the average within the domain, the stability may be low compared to other domains. Therefore, it is necessary to recalculate the average of nodes having a value greater than or equal to the average stability and select more stable nodes as candidate nodes. Through this, the leader node can maintain stable nodes in the domain as candidate nodes.

Through this process, the leader node can determine a candidate node and a follower node, and a lower-level virtual loop including the leader node, the candidate node, and the follower node can be configured.

미만을 갖게 되면 리더를 변경한다. 즉, 후보 노드 중 차순위 안정성 값을 갖는 후보 노드를 리더 노드로 설정하여, 후보 노드가 리더 노드의 역할을 동시에 수행하며, 기존의 리더 노드는 추종 노드로 전환된다.On the other hand, as the leader node moves, the stability value of the leader node changes to the second average value (

Change the leader when you have less. That is, by setting the candidate node having the next highest stability value among the candidate nodes as the leader node, the candidate node simultaneously performs the role of the leader node, and the existing leader node is converted into a follower node.

) 미만을 갖게 되면 후보 노드의 자격을 상실하고, 후보 노드는 추종 노드로 전환된다. 이때, 후보 노드가 추종 노드로 전환됨에 따라 도메인 내 리더 노드가 후보 노드의 기능을 동시에 수행하게 된다. 그리고 후보 노드 및 추종 노드를 다시 설정하는 과정을 수행하게 된다.In addition, as the candidate node also moves, the stability value changes to the first average value (

), the qualification of a candidate node is lost, and the candidate node is converted to a follower node. At this time, as the candidate node is converted to a follower node, the leader node in the domain simultaneously performs the function of the candidate node. Then, a process of re-establishing a candidate node and a follower node is performed.

Follower nodes have information about nodes in the 1-hop neighbor of the lower layer virtual loop. Since it has information on the following nodes that exist on both sides of the ring rather than the existing unidirectional ring structure, it can respond flexibly to failures even when the following nodes fail.

That is, when a communication failure with a neighboring node occurs outside the corresponding domain due to the movement of the following node, the neighboring nodes of the following node notify the candidate node, and the candidate node transmits information about the new lower layer virtual loop connection to the neighbor of the failed node. pass to the node. A new connection is made between the neighboring nodes that have received this, and according to this process, a robust virtual loop structure can be formed even in a dynamic environment in which insertion/deletion of nodes occurs frequently.

Each process performed in the event of such a failure will be described again with reference to the drawings.

6 is a view for explaining the content of forming an upper layer virtual ring between leader nodes in an embodiment of the present invention.

6 illustrates a process in which a leader node selected in one domain forms an upper layer virtual loop for information synchronization with a lido node in another domain. It has a two-layer virtual ring structure containing All two-layer virtual loops are formed as logical virtual loops.

As described above, the lower layer virtual loop is formed based on the stability values of nodes in the domain, whereas the upper layer virtual loop is formed based on the distance between each leader node.

Since a node with less movement in each domain is selected as a leader, the leader node in each domain has relatively high stability. Each leader node forms an upper layer virtual loop with other leader nodes in the nearest domain as its neighbors. Each leader node stores the value of the upper layer virtual loop in the database.

Meanwhile, the leader node may include a configuration of a monitoring agent, a reachability agent, a connectivity agent, etc. for communication between domains. In addition, it is possible to perform functions such as path computation, monitoring manager, service manager, event processing for communication within the domain.

The leader node performs synchronization by transmitting mobility and status information of the corresponding domain nodes only to the neighboring leader nodes connected in a loop. In other words, by performing synchronization using an upper layer virtual loop rather than full synchronization of the leader nodes, an effect of reducing the communication load can be expected even in synchronization between the leader nodes. When the neighbor leader node transmits its domain-related information back to the neighbor leader node, it also transmits information received from the previous neighbor leader node.

After configuring the two-layer virtual ring in this way, information on each node in and out of a plurality of domains is synchronized based on the two-layer virtual ring (S130).

7 is a diagram for explaining a synchronization process in a two-layer virtual ring.

7 is a diagram illustrating information synchronization between a following node and a candidate node in a domain and information synchronization between a leader node between domains, wherein the candidate node in the domain collects domain information including mobility information and connectivity information of each follower node; A leader node in a domain can synchronize information by transferring the collected information to a leader node in another domain.

Specifically, the following nodes such as C _1,1 and C _1,2 transmit information such as their mobility and connectivity to the candidate node C _{1,4 in the domain.} Candidate nodes ( C _1,4 ) play a role in forming and managing a lower-level virtual loop based on the information collected from each follower node. If there are a plurality of candidate nodes, the following nodes can select a candidate node closest to them to transmit mobility information and connectivity information, and the candidate node with the greatest stability value among the plurality of candidate nodes forms a lower-level virtual loop. and manage.

The leader node ( C _1,6 ) plays a role of synchronizing information with leader nodes of other domains based on the lower layer virtual ring information received from the candidate node. That is, the candidate node collects information in the domain and serves to form lower-level virtual ring information, and the leader node performs synchronization by transferring information collected from the lower node to other domains.

As a result, in one embodiment of the present invention, since a plurality of leaders with different roles exist for each domain, compared to a cluster structure in which one existing leader node is exclusively responsible for communication and management within/between a plurality of domains, the availability There is a peak that this increase and the load on the leader can be reduced. Also, when a leader is absent, the candidate node becomes a leader node, but since the following nodes already consider the candidate node as a leader, there is no need to separately notify the follower node of the change to the leader node.

Hereinafter, a process when a new node is added or a failure occurs in the virtual ring will be described with reference to FIGS. 8 and 9 .

8 is a diagram for explaining the contents of changing the lower layer virtual ring according to the addition of a new node in an embodiment of the present invention.

In one embodiment, when a new node is created in the domain, the new node provides information on the number of past connected nodes ( N _p ) and the number of current connected nodes ( N _c ) to the nearest node through broadcasting. to convey

The local node ( C _1,1 ) located closest to the new node calculates the stability value of the new node and transmits it to the candidate node ( C _{1,4 ) in the domain.} Receiving a reliability value candidate node updates the virtual sub-layer ring is based on the new stable value of the new node, each be configured to update the information of the virtual sub-layer adjacent to the ring of the new node following the node (C _1,5, C _1,7 ), the leader node ( C _1,6 ), and the new node.

The new node sends a permission request to each follower node ( C _{1,5 ,} C _1,7 ) to be configured as a neighbor, and each requested follower node transmits an accepting message to the new node.

Meanwhile, the leader node transmits the changed domain information to the leader node in another domain to perform synchronization. At this time, except for each following node ( C _{1,5 ,} C _1,7 ) to be configured as a neighbor of the new node, other following nodes do not need to know information of the new node because information exchange is possible through the neighbor node. As such, an embodiment of the present invention has an advantage in that communication costs can be reduced by transmitting new node information only to a neighbor node, a candidate node, and a leader node based on a two-layer virtual loop.

9 is a diagram for explaining a process according to a case in which a change occurs in a stability value of a leader node in an embodiment of the present invention.

If a failure occurs in the leader node ( C _1,6 ), the candidate node ( C _1,4 ) recognizes this and converts itself to a leader node to notify the leader node of another domain that the leader has been changed. In this case, the existing leader node temporarily performs the role of the corresponding candidate node.

Since the follower nodes only need to communicate with the candidate node smoothly, it is not necessary to transmit the change of the leader node to the follower node, thereby reducing the communication cost.

The leader node simultaneously performs the role of a candidate node, and a new candidate node ( C _1,1 ) is selected, and the newly selected candidate node delivers its information to a follower node in the domain.

Also, when the stability value of the leader node is less than the average stability value in the domain, a change process of the leader node occurs. In this case, as described above, the candidate node also performs the role of the leader node, and notifies the leader node of another domain that the leader has been changed.

Thereafter, the candidate node converted to the leader node selects a new candidate node by re-performing the 2-layer virtual ring creation process in the domain, and delivers the corresponding information to each node in the lower layer virtual ring.

On the other hand, when a failure occurs in the candidate node, the leader node recognizes it and notifies the nodes in the domain. In this case, since it is not necessary to notify the leader node of the other domain of the absence of the candidate node, communication cost can be reduced. The leader node performs a process of reselecting the candidate node while simultaneously performing the role of the candidate node. Afterwards, when a new candidate node is selected, the selected candidate node notifies the following node in the domain of its information.

In addition, when a failure and/or mobility of a follower node occurs and leaves the corresponding domain, neighboring nodes of the corresponding follower node transmit failure information to the candidate node. The candidate node recognizes this and re-performs the lower-level virtual ring creation process to create a new lower-level virtual ring. As such, in an embodiment of the present invention, even if a failure occurs in the following node, the lower layer virtual ring is configured, so that the candidate node does not need to inform all nodes of the new lower layer virtual ring information. In other words, it is enough to transmit information only to nodes in both its neighbors to form a new lower-level virtual loop.

Meanwhile, in the above description, steps S110 to S130 may be further divided into additional steps or combined into fewer steps according to an embodiment of the present invention. In addition, some steps may be omitted if necessary, and the order between the steps may be changed.

The method for selecting a reader controller in a distributed software defined network according to an embodiment of the present invention described above may be implemented as a program (or application) and stored in a medium to be executed in combination with a computer that is hardware.

The above-mentioned program, in order for the computer to read the program and execute the methods implemented as a program, C, C++, JAVA, Ruby, which the processor (CPU) of the computer can read through the device interface of the computer; It may include code coded in a computer language such as machine language. Such code may include functional code related to a function defining functions necessary for executing the methods, etc., and includes an execution procedure related control code necessary for the processor of the computer to execute the functions according to a predetermined procedure. can do. In addition, the code may further include additional information necessary for the processor of the computer to execute the functions or code related to memory reference for which location (address address) in the internal or external memory of the computer to be referenced. have. In addition, when the processor of the computer needs to communicate with any other computer or server located remotely in order to execute the above functions, the code uses the communication module of the computer to determine how to communicate with any other computer or server remotely. It may further include a communication-related code for whether to communicate and what information or media to transmit and receive during communication.

The storage medium is not a medium that stores data for a short moment, such as a register, a cache, a memory, etc., but a medium that stores data semi-permanently and can be read by a device. Specifically, examples of the storage medium include, but are not limited to, ROM, RAM, CD-ROM, magnetic tape, floppy disk, and an optical data storage device. That is, the program may be stored in various recording media on various servers accessible by the computer or in various recording media on the computer of the user. In addition, the medium may be distributed in a computer system connected to a network, and a computer-readable code may be stored in a distributed manner.

The above description of the present invention is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a dispersed form, and likewise components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

100: controller
200: network equipment
300: host

Claims (13)

A method for selecting a leader controller in a distributed software defined network, the method comprising:
calculating a stability value for the mobility of each of the plurality of controller nodes;
generating a two-layer virtual loop in and out of a plurality of domains based on the calculated stability value; and
Comprising the step of synchronizing information about each node in and out of a plurality of domains based on the two-layer virtual loop,
How to choose a leader controller in a decentralized software-defined network.
According to claim 1,
The step of generating a two-layer virtual ring within and outside a plurality of domains based on the calculated stability value,
generating a lower-level virtual loop within the domain based on the stability value; and
and generating an upper layer virtual ring in which leader controller nodes (hereinafter, leader nodes) located in neighboring domains among the plurality of domains are set as neighbors to each other,
How to choose a leader controller in a decentralized software-defined network.
3. The method of claim 2,
Calculating the stability value for the mobility of each of the plurality of controller nodes comprises:
calculating the number of redundantly connected nodes by comparing the number of past connected nodes at time t-1 based on one hop and the number of current connected nodes at time t; and
Comprising the step of calculating the stability value based on the ratio of the number of redundantly connected nodes to the number of current connected nodes,
How to choose a leader controller in a decentralized software-defined network.
3. The method of claim 2,
The step of creating a lower-level virtual ring in the domain based on the stability value comprises:
generating a virtual loop by arranging the calculated stability values for each controller node in the domain in descending order;
setting a controller node having the largest stability value in the virtual loop as a leader node;
determining, at the leader node, a candidate controller node (hereinafter, a candidate node) and a following controller node (hereinafter, a following node); and
Constructing the lower-level virtual loop including the leader node, candidate node and follower node;
How to choose a leader controller in a decentralized software-defined network.
5. The method of claim 4,
The step of determining a candidate node and a follower node in the leader node comprises:
calculating a first average value of stability values for a plurality of nodes included in the domain;
calculating a second average value of stability values for nodes having stability values greater than or equal to the first average value; and
setting a node having a stability value equal to or greater than the second average value as a candidate node and setting the remainder as a follower node,
How to choose a leader controller in a decentralized software-defined network.
6. The method of claim 5,
converting the leader node to a follower node when the leader node has a stability value less than the second average value according to the movement of the leader node; and
Further comprising the step of setting a candidate node having a next-order stability value among the candidate nodes as a leader node,
How to choose a leader controller in a decentralized software-defined network.
7. The method of claim 6,
transmitting, by a new leader node, a leader node change message to a leader node in another domain, as the candidate node is converted to a leader node; and
Further comprising the step of the new leader node re-performing the two-layer virtual ring virtual ring creation process in the domain,
How to choose a leader controller in a decentralized software-defined network.
6. The method of claim 5,
converting the candidate node into a follower node when the candidate node has a stability value less than the first average value according to the movement of the candidate node;
a step in which a leader node in a domain simultaneously performs a function of a candidate node as the candidate node is converted into a follower node; and
Further comprising the step of re-establishing the candidate node and the following node,
How to choose a leader controller in a decentralized software-defined network.
6. The method of claim 5,
transmitting failure information to a candidate node by neighboring nodes of the following node when the following node leaves the corresponding domain according to the movement of the follower node; and
Further comprising the step of the candidate node re-performing the lower layer virtual ring creation process of the domain by reflecting the failure information,
How to choose a leader controller in a decentralized software-defined network.
6. The method of claim 5,
The step of synchronizing information about each node in and out of a plurality of domains based on the two-layer virtual loop comprises:
collecting, by the candidate node in the domain, domain information including mobility information and connectivity information of each following node; and
and transmitting, by a leader node in the domain, the domain information to a leader node in another domain.
How to choose a leader controller in a decentralized software-defined network.
11. The method of claim 10,
When there are a plurality of candidate nodes in the domain, each of the following nodes selects the closest candidate node and transmits the mobility information and the connectivity information;
A candidate node having a large stability value among the plurality of candidate nodes forms and manages the lower layer virtual loop,
How to choose a leader controller in a decentralized software-defined network.
3. The method of claim 2,
a step of receiving, by a predetermined node (hereinafter, a nearest node) located at a nearest distance from a new node created in the domain, information on the number of past connected nodes and the number of currently connected nodes of the new node;
calculating, by the nearest node, a stability value of the new node and transmitting it to a candidate node in the domain;
updating, by the candidate node, a lower layer virtual loop based on the stability value of the new node;
transmitting the updated lower-level virtual ring information to a leader node; and
Further comprising the step of the leader node transmitting the updated domain information to a leader node in another domain,
How to choose a leader controller in a decentralized software-defined network.
13. The method of claim 12,
The step of the candidate node updating the lower layer virtual loop based on the stability value of the new node,
Each node to be configured as a neighboring node of the new node, and transmitting the updated lower-level virtual ring information to the nearest node and the new node,
How to choose a leader controller in a decentralized software-defined network.

Images