WO2015065003A1

WO2015065003A1 - Method for controlling qos by handling traffic depending on service

Info

Publication number: WO2015065003A1
Application number: PCT/KR2014/010169
Authority: WO
Inventors: 안태진; 김형수; 이세희; 한경아
Original assignee: 주식회사 케이티
Priority date: 2013-10-28
Filing date: 2014-10-28
Publication date: 2015-05-07

Abstract

A method for controlling QoS by handling traffic for each service in a software defined network environment is disclosed. The method of a controller controlling QoS comprises the steps of: determining a queue configuration change that includes performing one or more of adding, deleting and changing one or more queues configured to an output port of a network apparatus; and performing the configuration change by transmitting to the network apparatus a queue status change message including information about the determined queue configuration change. As a result, various services can be processed differently according to their traffic characteristics, thereby providing sufficient quality for each service and efficiently using network resources.

Description

OS control method using traffic processing according to service

The present invention relates to QoS control, and more particularly, to a QoS control method through traffic processing for each service in a software defined networking environment.

Recently, standardization of technologies that efficiently operate communication systems by separating the traffic forwarding function of the switch and the control function of the switch has been carried out by the Open Networking Foundation (ONF), Internet Engineering Task Force (IETF), and European Telecommunications Standards Institute ISG Network Function Virtualization (NFV) and ITU-T (International Telecommunications Union Telecommunication) are being conducted mainly.

Software Defined Network (SDN) refers to a user-oriented network in which a user has control regardless of basic network devices such as a router or a switch, and a separate software controller controls traffic flow.

Among standardization organizations that promote standardization of OpenFlow technology, one of SDN's technologies, Open Networking Foundation (ONF) defines an interface between hardware (switch) and controller (Network OS). This is a protocol for interacting with the data plane by separating the control plane from the physical network to control how data packets are delivered through the network.

Meanwhile, in the Internet, as various services such as voice, video, data, wired and wireless are integrated into a single network, techniques that can efficiently use network resources while ensuring service quality for their individual characteristics have become very important.

Accordingly, various QoS control methods have been developed and used in the existing switch to handle traffic differently according to the characteristics of services. However, only basic QoS control methods are defined in the current ONF OpenFlow standard. There is a problem in that it is difficult to efficiently control the quality of service (QoS) for traffic.

An object of the present invention for solving the above problems is to provide a QoS control method in an SDN environment.

Another object of the present invention for solving the above problems is to provide a QoS control method using traffic processing according to a service.

Another object of the present invention for solving the above problems is to provide a method for adding or deleting a queue for processing traffic according to a service, and a QoS control method using traffic processing according to queue statistics and network types.

A method for controlling QoS by a controller according to an aspect of the present invention for achieving the above object includes performing at least one of adding, deleting, and changing at least one queue configured at an output port of a network device. Determining a setting change of the queue for; And transmitting a queue state change message including information on the setting change of the determined queue to a network device, to perform setting change of the queue.

Here, in the determining of the setting change of the queue, the setting change of the queue may be determined based on a service requirement or a QoS policy change.

Here, the method includes receiving queue characteristic information for at least one queue set at an output port of a network device; Generating a flow entry for mapping between a service and at least one queue based on the queue characteristic information; The method may further include controlling the network device to process traffic for each service according to the generated flow entry.

In the receiving of the queue characteristic information, the queue status message including the queue characteristic information may be received from a network device.

In the generating of the flow entry, each service may be mapped to at least one queue set separately by being divided into a priority queue and at least one weighted round robin (WRR) queue.

The generating of the flow entry may include mapping traffic for the highest rank service to be assigned to a priority queue and mapping traffic for the next rank service to be sequentially assigned to at least one WRR queue. have.

Herein, the generating of the flow entry may sequentially process traffic for the service of the highest rank and sequentially perform traffic for the service of the next rank based on a weight applied to each of the at least one WRR queue in the remaining traffic range. Can be mapped to be assigned.

In the generating of the flow entry, traffic allocated to at least one critical server may be mapped to different WRR queues.

The controlling of the network device may include: transmitting a flow table change message including a generated flow entry to the network device; Updating the flow table of the network device using the flow table change message; The method may include controlling the network device to process traffic for each service based on the updated flow table.

QoS control method using the traffic processing according to the service according to another aspect of the present invention for achieving the above object, in the method for the controller to control the QoS, the queue that can be supported by the output port of the network device from the network device (Queue) Receiving a queue setting information message including information about; Mapping between a service and at least one queue based on a queue establishment information message and a QoS policy; Sending a queue state change message to the network device to establish at least one queue based on a mapping relationship between a service and at least one queue.

Here, mapping between the service and at least one queue comprises: setting each of the at least one queue to a Priority queue or at least one Weighted Round Robin (WRR) queue; Mapping each service to at least one queue based on the priority of each service.

The mapping of the services to the at least one queue based on the priority of each service may include mapping traffic for the highest priority service to be allocated to the priority queue and at least traffic for the next rank service. It can be mapped to be sequentially assigned to one WRR queue.

Here, mapping the services to the at least one queue based on the priority of each service may be performed by processing the traffic for the highest priority service and applying a weight to each of the at least one WRR queue in the remaining traffic range. Based on this, traffic for next-order services may be mapped sequentially.

Here, the queue state change message may include a command for adding or deleting a queue to an output port of the network device.

Wherein the method comprises the steps of: receiving information about a lost packet from a network device to calculate a packet loss rate; The method may further include changing a weight applied to each of the at least one WRR queue based on the packet loss rate.

Here, the changing of the weights applied to each of the at least one WRR queue may be performed by including a command for changing a weight in the queue state change message and transmitting the weight change command to the network device.

Here, the method comprises the steps of: receiving a queue status message comprising queue characteristic information for at least one queue set based on the queue status change message; Generating a flow entry for mapping between a service and at least one queue based on the queue characteristic information; The method may further include controlling the network device to process traffic for each service according to the generated flow entry.

QoS control method using the traffic processing according to the service according to another aspect of the present invention for achieving the above object, in the method for controlling the QoS by the network device, a queue (support) supported by the output port of the network device (Queue) Transmitting a queue setting information message including information on the controller to the controller; Receiving a queue state change message generated by the controller based on the queue setting information message and the QoS policy; Setting at least one queue to an output port according to the queue state change message.

Here, the queue state change message may include information generated by a controller by mapping between a service and at least one queue based on a queue setting information message and a QoS policy.

Here, the queue state change message may include information generated by the controller by mapping between a network and at least one queue according to a network type priority based on a queue setting information message and a QoS policy.

Here, the setting of the output port may be configured such that traffic for a service of the highest rank is allocated to a priority queue and traffic for a service of a next rank may be sequentially assigned to at least one WRR queue. .

Herein, the setting of the output port may include sequentially processing traffic for a service of the highest priority and sequentially processing traffic for a service of the next rank based on a weight applied to each of the at least one WRR queue in the remaining traffic range. Can be set to be assigned.

The QoS control method using the traffic processing according to the service according to the present invention as described above, it is possible to efficiently use the network resources while satisfying the service quality by differentiating and processing the various services according to the traffic characteristics.

In addition, the controller can accurately determine the queue status of the switch in real time so that it can issue appropriate QoS control commands to the switch.

1 is a block diagram illustrating a configuration of a controller and a network apparatus for performing a QoS control method using traffic processing according to a service according to an embodiment of the present invention.

2 is a conceptual diagram illustrating a priority queue and a weighted round robin queue according to an embodiment of the present invention.

3 is a flowchart illustrating a method of controlling QoS by mapping a queue for each service according to an embodiment of the present invention.

4 is an exemplary diagram for explaining a queue status message according to an embodiment of the present invention.

5 is a flowchart illustrating a method of controlling QoS by changing a queue state according to an embodiment of the present invention.

6 is a flowchart illustrating a procedure of adding a queue according to an embodiment of the present invention.

7 is a flowchart illustrating a procedure of a method of deleting a queue according to an embodiment of the present invention.

8 is a flowchart illustrating a procedure of changing a queue characteristic based on a lost packet rate according to an embodiment of the present invention.

9 is a flowchart illustrating a procedure of mapping traffic allocated to a critical server to different queues according to an embodiment of the present invention.

10 is a flowchart illustrating a method of controlling QoS by mapping queues for each network type according to an embodiment of the present invention.

11 is a conceptual diagram illustrating a case where a queue is mapped for each network type according to an embodiment of the present invention.

As the invention allows for various changes and numerous embodiments, particular embodiments will be illustrated in the drawings and described in detail in the written description. However, this is not intended to limit the present invention to specific embodiments, it should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention. In describing the drawings, similar reference numerals are used for similar elements.

Terms such as first, second, A, and B may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as the second component, and similarly, the second component may also be referred to as the first component. The term and / or includes a combination of a plurality of related items or any item of a plurality of related items.

When a component is referred to as being "connected" or "connected" to another component, it may be directly connected to or connected to that other component, but it may be understood that other components may be present in between. Should be. On the other hand, when a component is said to be "directly connected" or "directly connected" to another component, it should be understood that there is no other component in between.

The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, the terms "comprise" or "have" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, components, or a combination thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art. Terms such as those defined in the commonly used dictionaries should be construed as having meanings consistent with the meanings in the context of the related art and shall not be construed in ideal or excessively formal meanings unless expressly defined in this application. Do not.

Hereinafter, the controller referred to in the present invention is an integrated SDN controller, which may mean a function element for controlling related components (eg, a switch, a router, etc.) to control the flow of traffic. Can be.

In addition, the controller is not limited to the physical implementation form or implementation location. For example, a controller refers to a controller function entity defined by OpenFlow (ONF), Internet Engineering Task Force (IETF), European Telecommunication Standards Institute (ETSI) and / or International Telecommunication Union Telecommunication (ITU-T). can do.

The network device referred to in the present invention may refer to a functional element that substantially forwards, switches, or routes traffic (or packets), such as 'switch' or 'router'. Therefore, in the present invention, the network device may be referred to as a switch or a router.

For example, a network device may mean a switch, a router, a switching element, a routing element, a forwarding element, and the like defined in ONF, IETF, ETSI, and / or ITU-T. Can be.

In the embodiments of the present invention, for example, the parameter and / or message type (for example, flow table entry) defined in the ONF is used in the operation for QoS control in the openflow switch. The mapping is not limited to the contents defined in ONF, and various parameters that can distinguish a controller from a switch can be used in transmitting a QoS control command between a controller and a switch, and messages used in an operation process for QoS control. Nor is it limited to the specific message mentioned later.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Referring to FIG. 1, a QoS control method using traffic processing according to a service according to an embodiment of the present invention may be performed by a controller and a switch.

First, the controller 100 according to an exemplary embodiment of the present invention includes a QoS policy management unit 110, a switch control unit 120, a flow table management unit 130, and a switch interworking processing unit 140.

The QoS policy manager 110 may manage a QoS policy for a domain managed by the controller 100.

The switch controller 120 may control the switch 200 to communicate with the controller 100 through the switch interworking processor 140. For example, the switch controller 120 may generate a control command for adding, modifying, and deleting a flow entry based on the QoS policy managed by the QoS policy manager 110, and may generate a control command. The switch 200 can be controlled.

The flow table manager 130 may store and manage a parameter and a flow table for synchronizing the flow table.

The switch interworking processor 140 may process a protocol so that the switch 200 and the controller 100 can communicate.

Next, the switch 200 according to the embodiment of the present invention includes a Qos controller 210, a switch controller 220, a flow table manager 230, and a controller interworking processor 240.

The QoS controller 210 may actually perform QoS control on the packet delivered to the output port.

The switch controller 220 may control the switch such as setting a queue characteristic for the output port. In addition, the switch controller 220 may receive and execute a control command from the controller 100.

The flow table manager 230 may manage the flow table received from the controller 100. That is, the flow table manager 230 may store and manage the parameters and the flow table for synchronizing the flow table for the switch.

The controller interworking processor 240 processes the protocol so that the controller 100 and the switch 200 can communicate.

Referring to FIG. 2, at least one queue may be set at an output port of a network device.

The case where there are four queues at the output port of the network device will be described in more detail as an example.

If there are four queues at the output port of the network device, one queue is used as a priority queue and the other three queues are assigned as weighted round robin (WRR) queues.

Service traffic can be classified into gold, silver, bronze, and best effort (BE) according to their characteristics.

Gold traffic is assigned to the priority queue and can be processed with the highest priority. In other words, if a packet is in a Priority queue, it can always be processed first, even if there is a packet in another queue.

Silver traffic has priority after Gold traffic. However, not all Silver traffic will always be processed before other Bronze or Best effort traffic, but will be processed by the assigned weight and then Bronze traffic.

Bronze traffic may also be processed by the assigned weight, followed by Best effort traffic.

For example, Silver traffic can use 50% of the remaining bands used by Gold traffic, Bronze traffic can use 30% of the remaining bands used by Gold traffic, and Best effort traffic can use the remaining bands left by Gold traffic. It may be set to use 20% of the band. Therefore, silver traffic may be processed with a higher priority than Bronze traffic, and Bronze traffic may be processed with a higher priority than Best effort traffic.

Referring to FIG. 3, the switch may set a queue property for the output port x by using a command line interface (CLI) or a configuration protocol (S310).

More specifically, the switch sets four queues for output port x, one queue (Q1) to Priority queue and the other three queues (Q2, Q3, Q4) to WRR (weighted round robin) queues. Can be set. For example, the switch sets four queues for output port x.However, the number of queues that can be set for an output port may vary depending on the functions or capabilities provided by the switch. The number of possible queues may also be three or more.

Weights may be set to w1% (Q2), w2% (Q3), and w3% (Q4) for the three queues Q2, Q3, and Q4 set to WRR, respectively.

For example, high weights may be set for Q2, medium weights for Q3, and low weights for Q4.

The switch may newly set the queue characteristic for the output port x or, if the parameter value of the previously set queue characteristic is changed, may immediately notify the controller through the queue status message (S320).

The queue status message may include queue characteristic information such as output port x where the queue status change occurred, queue list, and queue characteristic values for the queue included in the queue list. Here, the queue characteristic value may include information on whether the corresponding queue is a priority queue or a WRR queue, and information on weight when the queue is a WRR queue.

Since the controller receives the queue status message from the switch and knows the characteristics of each queue from the queue characteristic information, the controller configures a service-specific queue mapping table to determine which queues to assign flows to each service characteristic. It may be (S330).

Which queue to allocate for each service is determined by establishing QoS policy considering the importance of each service and sensitivity to delay or loss of packets according to network operation policy.

For example, very important packets used for Broadband convergence Network (BcN) service or routing protocol can be assigned to Q1 queue as Gold service and important packets such as VoIP or IPTV service can be assigned to Q2 queue. have.

In addition, packets such as VoD or VPN service may be assigned to the Q3 queue by defining it as a Bronze service, and general Internet service may be assigned to the Q4 queue by setting it as a best effort service.

After configuring the queue mapping table for each service, the controller may configure or generate a flow entry so that the actual service-specific flow may be mapped to the corresponding queue (S340).

A flow entry may be divided into a match field for distinguishing flows and an action field indicating what action to perform on a corresponding flow packet.

For example, the controller sends Flow 1 for BcN or routing protocol packets to Q1, Flow 2 for VoIP or real-time IPTV packets to Q2, and Flow 3 for VoD or VPN packets to Q3. Flow 4, which is a normal Internet packet, can configure a flow entry to send a command to the switch to send to Q4.

After the controller generates the flow entries, the controller may include them in a flow table change message and transmit them to the switch (S350).

The switch may receive a flow table change message from the controller and update the flow table of the switch using a flow entry list included therein (S360).

Therefore, when a packet matching the corresponding flow arrives, QoS control according to queue characteristics may be performed by transmitting to the corresponding queue of the corresponding output port according to the action content set in the action field. That is, the switch may perform QoS control by allocating flows to different queues for each service (for example, a priority queue or a WRR queue) (S370).

Referring to FIG. 4, a queue status message transmitted by a switch to a controller will be described as an example.

Four queues are assigned to output port x, of which Q1 can assign a property of Priority Q and a value of Null.

Q2 can assign property to WRR and assign value to 50%, Q3 can assign property to WRR and assign value to 30%, and Q4 can assign property to WRR. And value to 20%.

In addition, in the WRR method, the value allocated to each queue can be determined by predicting the priority and traffic amount for each service, and the value can be changed by periodically collecting traffic during operation.

Referring to FIG. 5, the controller may control the QoS by transmitting a queue state change message to the switch.

Specifically, in order for the controller to change (e.g., add / modify / delete) the queue state of output switch x of the switch, it is necessary to provide configuration information such as the maximum number of queues that can be set on output port x or the supported queue characteristics. You must know.

The controller may transmit a queue setting information request message to the switch in order to know the queue setting information of the switch (S510). In this case, if the output port parameter is set to x, the switch may transmit a queue setting information message for the output port x to the controller (S520). Here, the queue setting information message may include setting information such as the maximum number of queues that can be set in the output port x or the queue characteristics that can be supported.

In addition, if the output port parameter is set to 'ANY' instead of x, it may mean that the queue setting information for all output ports of the switch is sent.

In detail, the switch may receive the queue configuration information request message and include the configuration information such as the maximum number of queues that can be set for the output port x and the queue characteristics that can be supported in the queue configuration information message and transmit the same to the controller.

Similarly, if the output port parameter is set to 'ANY' in the queue configuration information request message, the switch can configure the configuration information such as the maximum queue count and the supported queue characteristics of all output ports in a list form and send them to the controller. .

Depending on the switch, configuration information such as the maximum queue number that can be set for all output ports and the queue characteristics that can be supported may be identical. For these switches, you can use 'ANY' for the output port parameter and send the configuration information such as maximum queue count and supported queue characteristics to one.

After receiving the queue configuration information message, the controller may determine to which service to map the corresponding queues according to the QoS policy (S530).

The switch may set four queues for output port x, and one of the queues Q1 as a priority queue and the other three queues Q2, Q3, and Q4 as weighted round robin (WRR) queues.

For example, the controller maps the very important packets used for BcN service or routing protocol to Q1 queue by mapping Gold service, and important packets such as VoIP or IPTV service are mapped to Silver service, and high weight among WRR queues. Can be assigned to the Q2 queue with.

In addition, packets such as VoD or VPN service are mapped to the Bronze service and assigned to the Q3 queue having the middle weight among the WRR queues, and general Internet services are mapped to the best effort service to the Q4 queue having the lowest weight among the WRR queues. Can be.

After performing the queue mapping for each service according to the QoS policy, the controller may transmit a queue state change message to the switch to reflect this to the switch (S540).

The queue state change message may include an output port x, a command, a queue list, and property information for each queue as an output port parameter. Here, the command may indicate a command to add, delete, or change the queue. Further, the queue characteristic information can be understood with reference to the table shown in FIG.

The switch may set the queue characteristic for the output port x by receiving a queue state change message from the controller (S550).

For example, the switch sets one queue (Q1) to Priority Q for output port x, and the other three queues (Q2, Q3, and Q4) to weighted round robin (WRR) queues, and the weights of Q2, Q3, and Q4. (weight) can be set to w1%, w2% and w3%, respectively.

When the setting of the queue characteristic for the output port x is finished, the switch may transmit a queue status message to the controller (S560). Here, the queue status message may include an output port x, a queue list, a queue characteristic value, and the like.

The controller may generate a flow entry for QoS control after receiving the queue status message from the switch (S570), and load the generated flow entry into a flow table change message to the switch (S580).

Accordingly, the switch receiving the flow table change message may update the flow table (S590), and execute QoS control on the actual packet (S593).

Referring to FIG. 6, the controller may add a queue to a corresponding output port by sending an 'ADD' command to the switch.

For example, a case where only one Q1 is assigned to the output port x of the switch and the property of Q1 is assigned to BE (Best Effort) will be described.

In this case, all packets coming into output port x are processed equally through Q1.

The controller can add a queue with new queue characteristics to the switch according to the request of the service system or the change of the QoS policy.

For example, if a requirement that traffic to a specific destination must guarantee at least some bandwidth, then the Q1 set on an existing switch can't handle this requirement if it is requested from a service system or QoS policy. You need to add a queue.

According to FIG. 6, the controller may decide to add Q2 to the output port x according to the QoS policy, and the bandwidth according to the characteristic of Q2 may be allocated as min_rate = a and max_rate = b (S610).

That is, Q2 guarantees minimum bandwidth a bps (kbps, Mbps) for traffic (e.g., if output port x is a 1 Gbps port, the minimum bandwidth is a × 10 Mbps), while the maximum bandwidth is b bps (kbps, Mbps). For example, if the output port x is a 1 Gbps port, it may mean that the maximum bandwidth is limited to b × 10 Mbps.

The controller may transmit a queue state change message to the switch (S620). Here, the queue state change message may include parameters such as an output port x, a command to add a new queue, an id of a queue to be added, and property information on the queue to be added. Therefore, Q2 is added to output port x using a queue state change message that includes a command to add a new queue (ADD), and the minimum bandwidth (min_rate) is a% of the maximum transmission speed of output port x by the characteristic of Q2. The maximum bandwidth (max_rate) can be set to be b% of the maximum transmission rate of the output port x.

More specifically, the switch may further set Q2 to the output port x after receiving the queue state change message from the controller (S630). For example, two queues Q1 and Q2 may be set at the output port x.

After setting a new queue (Q2) for the output port x, the switch can send a queue status message to the controller (S640). Here, the queue status message may include an output port x, queue lists Q1 and Q2, and queue property values.

After the controller receives the queue status message from the switch, the controller may generate a flow entry for QoS control with reference to this (S650).

For example, the controller could create a flow entry to send all packets destined for the destination (Dest. IP) to Q2 on output port x.

After the controller generates a new flow entry, the controller may load the flow table change message to the switch (S660). Here, the flow table change message may include a parameter such as a new flow entry and a command to add it to the flow table.

The switch receiving the flow table change message may add a newly received flow entry to the flow table (S670). Therefore, the switch may transmit all packets destined for the destination (Dest. IP) to 10.1.1.0 among the packets entering the switch to Q2 of the output port x to perform QoS control according to the properties of Q2 (S680). ).

Referring to FIG. 7, the controller transmits a 'DELETE' command to a switch to delete a queue at a corresponding output port.

The output port x of the switch is currently assigned two queues, Q1 and Q2, and the property of Q1 is assigned to BE (Best Effort), and the property of Q2 is that the minimum bandwidth is equal to the maximum transmission speed of output port x. The case where a maximum bandwidth is allocated by a% (min_rate = a) and b% (max_rate = b) of the maximum transmission rate of the output port x will be described as an example.

The controller may determine to delete the queue set in the switch according to the request of the service system or the change of the QoS policy (S710). For example, you can guarantee at least some bandwidth for traffic destined for a particular server, but when the server stops service, you no longer need to guarantee bandwidth. In addition, if you guarantee a minimum amount of bandwidth for traffic to a specific customer site, and the customer terminates or changes the service, you may not need the minimum bandwidth guarantee. In this case, it is necessary to delete Q2 set at the output port x of the switch.

If the controller determines to delete Q2 at the output port x of the switch, the controller may transmit a queue state change message to the switch (S720). Here, the queue state change message may include parameters such as an output port x, a command to DELETE the queue, and a queue id to be deleted.

The switch may delete Q2 at the output port x after receiving the queue state change message from the controller (S730). Therefore, output port x can delete Q2 from the two queues Q1 and Q2 and leave only Q1.

When the deletion for Q2 is completed, the switch may transmit a queue status message to the controller (S740). Here, the queue status message may include an output port x, a queue list Q1, a queue property value, and the like.

In addition, the switch may delete the flow entry related to Q2 of the output port x in the flow table (S750) and transmit a flow entry delete message to the controller (S760).

The controller may delete the flow entry after receiving the flow entry deletion message from the switch (S770).

For example, the switch could delete the flow entry to send all packets destined for the destination (Dest. IP) to Q2 on output port x. Therefore, all packets coming into the output port x of the switch may be delivered to Q1 and processed as BE (Best Effort) (S780).

Referring to FIG. 8, the controller may analyze the overall traffic situation as well as change the QoS policy during operation of the switch, and change the previously set queue property as needed.

For example, if the WRR queue is applied to the output port x, the weight of the queue can be changed flexibly according to traffic conditions.

More specifically, the controller may periodically send a queue statistics request message to the switch to request traffic statistics data for the queues (S810). Here, the queue statistics request message may include parameters such as a corresponding output port and a queue id.

The switch may transmit the queue statistics message including the traffic statistics data for the corresponding queue according to the queue statistics request message to the controller (S820). Here, the queue statistics message may include an output port, a queue id, a number of transmission packets, and a number of lost packets.

The controller may calculate a packet loss rate for the corresponding queue by receiving the queue statistics message (S830). Here, the packet loss rate may be calculated as lost packets / (transmitted packets + lost packets).

The controller may change the weights of the WRR queues by comparing the packet loss rate of the queues with a threshold preset for each queue (S830).

For example, if silver service traffic increases more than initially anticipated and the quality of service is degraded, the weight of Q2 assigned to the silver service is increased to handle more traffic.

When the queue characteristic is changed, the controller may transfer information on the changed queue characteristic to a queue state change message and transmit the information to the switch (S840). Here, the queue status change message may include parameters such as an output port x, a command to change the queue characteristics (MODIFY), information about a queue id and queue characteristics to be changed, and the like.

The switch may change the weight for each queue according to the queue state change message (S850), and transmit the result to the queue state message to the controller (S860). For example, the switch may change the weights w1%, w2%, and w3% applied to the WRR queues Q2, Q3, and Q4 to y1%, y2%, and y3%, respectively.

Referring to FIG. 9, the controller may map a queue to handle traffic destined for important servers differently from general data traffic.

First, the switch may set the characteristics of the queue and the set queue to the output port, respectively (S910). The controller can know the information on the queue set in the switch by receiving a queue status message from the switch (S920).

The controller may configure a queue mapping table for processing traffic to important servers differently from general data traffic according to the QoS policy (S930).

For example, the controller can map traffic to the SDN Controller to Q2 for processing as Silver service, and traffic to SIP Call Server to Q3 for processing to Bronze service.

After configuring the queue mapping table for the important servers, the controller may generate a flow entry so that the actual flow may be mapped to the corresponding queue (S940).

For example, suppose the SDN Controller uses 10.1.1.1 as the IP address and yyyy as the port for the TCP connection with the switch. The SIP Call Server uses 20.1.1.1 as the IP address and the TCP connection for the terminal. It is assumed that zzzz is used as the port.

The flow entry generated by the controller may be divided into a match field for distinguishing flows and an action field indicating what action to perform on packets of the flow.

For example, since the flow of traffic to the SDN Controller only needs to specify the destination IP and port, the source IP and port can be represented by '*' indicating Any, and the traffic to the SIP Call Server. The source IP and port may be marked with '*' because the destination IP only needs to specify the destination IP and port.

In addition, the action field may include an action to send the traffic flow to the SDN controller to Q2, and to send the traffic flow to the SIP Call Server to Q3.

After generating the flow entry, the controller may load the flow table change message to the switch (S950).

The switch may receive the flow table change message from the controller and update the flow table of the switch by using the flow entry included in the flow table change message (S960).

Therefore, when the packet goes to SDN Controller, the switch forwards it to Q2 of the output port x and processes it as WRR-type Silver traffic. When the packet goes to SIP Call Server, the switch forwards it to Q3 of the output port x and Bronze of WRR method. Traffic may be processed, and through this, QoS control according to a corresponding queue property may be performed (S970).

FIG. 10 is a flowchart illustrating a method of controlling QoS by mapping a queue for each network type according to an embodiment of the present invention, and FIG. 11 illustrates a case of mapping a queue for each network type according to an embodiment of the present invention. This is a conceptual diagram.

Referring to FIG. 10, the controller may perform QoS control according to the type of network in association with a switch.

First, the switch may set the characteristics of the queue and the set queue to the output port, respectively (S1010). The controller can know the information on the queue set in the switch by receiving a queue status message from the switch (S1020).

The controller may configure a queue mapping table that may set a QoS policy differently according to the network type by referring to the queue status message (S1030).

For example, the controller maps Q2 to process Silver Internet for wireless Internet traffic coming from 3G or 4G wireless access networks, and Q3 to process Bronze service for wireless Internet traffic coming from WIBRO wireless access networks. For wired Internet traffic coming from a wired access network, it can be mapped to Q4 to be treated as a Best Effort (BE) service.

In general, the cost to process the same data is highest for the 3G / 4G wireless Internet, followed by the higher WIBRO Internet and the lowest wired Internet. Therefore, the 3G / 4G wireless Internet is treated as a Silver service, the WIBRO wireless Internet is treated as a Bronze service, the QoS policy is set to treat the wired Internet as a BE (Best Effort) service, and the controller is assigned to the QoS policy for each network type. QoS can be controlled based on this.

The controller may configure a queue mapping table to differentiate traffic for each network type, and then generate a flow entry for allowing the actual flow to be mapped to the corresponding queue (S1040).

10 and 11, 3G / 4G wireless Internet traffic may enter Ingress 1 of the switch, WIBRO wireless Internet traffic may enter Ingress 2, and wired Internet traffic may enter Ingress 3.

For example, the controller sets the Action field to forward flows coming into Ingress 1 of the match field to Q2 on output port x, and the Action field sets forwarding flows to Ingress 2 of the match field to Q3 of output port x. In addition, the Action field can be configured to forward the flow into Ingress 3 of the Match field to Q4 of the output port x.

After generating the flow entry, the controller may load it on the flow table change message to the switch (S1050).

The switch may receive a flow table change message from the controller and update the flow table of the switch using the flow entry included in the flow table change message.

Therefore, the switch forwards 3G / 4G wireless Internet packets from Ingress 1 to Q2 on output port x for WRR-based Silver service, and WIBRO wireless Internet packets from Ingress 2 are forwarded to Q3 on output port x to WRR. Processed by the Bronze service of the method, and wired Internet packets from Ingress 3 are delivered to Q4 of the output port x and processed by the BE (Best Effort) service of the WRR method to perform QoS control for each network type (S1060). .

According to the embodiment of the present invention described above, QoS can be controlled through traffic discrimination processing for each service in the open flow switch.

In addition, according to an embodiment of the present invention, by applying a weighted round robin (WRR) method to an open flow switch, it is possible to differentiate and process various services according to traffic characteristics.

Furthermore, according to an embodiment of the present invention, when the queue state is changed in the switch, the controller immediately informs the controller of the change, so that the controller can effectively change the queue state of the switch according to a network operation policy or traffic condition. QoS control can be effectively performed.

Although described above with reference to a preferred embodiment of the present invention, those skilled in the art will be variously modified and changed within the scope of the invention without departing from the spirit and scope of the invention described in the claims below I can understand that you can.

Claims

In the method for the controller to control the QoS,

Determining a setting change of a queue for performing at least one of adding, deleting, and changing at least one queue set at an output port of the network device; And

Transmitting a queue state change message including information on a setting change of the determined queue to the network device to perform a setting change of the queue;

How to control QoS.
The method according to claim 1,

Determining a setting change of the queue,

Characterized by determining a setting change of the queue based on a service requirement or QoS policy change,

How to control QoS.
The method according to claim 1,

Receiving queue characteristic information for the at least one queue set at an output port of the network device;

Generating a flow entry for mapping between a service and the at least one queue based on the queue characteristic information; And

Controlling the network device to process traffic for each service according to the generated flow entry;

QoS control method.
The method according to claim 3,

Receiving the queue characteristic information,

Receiving a queue status message including the queue characteristic information from the network device;

QoS control method.
The method according to claim 3,

Generating the flow entry,

Characterized in that each of the mapping on the basis of the priority of each service to the at least one queue set to be divided into a priority queue and at least one Weighted Round Robin (WRR) queue,

QoS control method.
The method according to claim 5,

Generating the flow entry,

And to map traffic for the highest priority service to be allocated to the Priority queue and to map traffic for a next rank service to the at least one WRR queue sequentially.

QoS control method.
The method according to claim 6,

Generating the flow entry,

Process the traffic for the highest-ranked service and map the traffic for the next-ranked services sequentially based on a weight applied to each of the at least one WRR queue in the remaining traffic range;

QoS control method.
The method according to claim 5,

Generating the flow entry,

Characterized by mapping traffic allocated to at least one critical server to different WRR queues,

QoS control method.
The method according to claim 3,

Controlling the network device,

Sending a flow table change message including the generated flow entry to the network device;

Updating the flow table of the network device using the flow table change message; And

And controlling the network device to process traffic for each service based on the updated flow table.

QoS control method.
In the method for the controller to control the QoS,

Receiving a queue setting information message from a network device, the queue setting information message including information on a queue that can be supported by the output port of the network device;

Mapping between a service and at least one queue based on the queue establishment information message and a QoS policy; And

Sending a queue state change message to the network device to establish the at least one queue based on a mapping relationship between the service and the at least one queue,

QoS control method using traffic processing according to service.
The method according to claim 10,

Mapping between the service and at least one queue,

Setting each of the at least one queue to a Priority queue or at least one Weighted Round Robin (WRR) queue; And

Mapping services to the at least one queue based on the priority of each service.

QoS control method using traffic processing according to service.
The method according to claim 11,

The mapping of the services to the at least one queue based on the priority of each service may include:

And to map traffic for the highest priority service to be allocated to the Priority queue and to map traffic for a next rank service to the at least one WRR queue sequentially.

QoS control method using traffic processing according to service.
The method according to claim 12,

The mapping of the services to the at least one queue based on the priority of each service may include:

Process the traffic for the highest-ranked service and map the traffic for the next-ranked services sequentially based on a weight applied to each of the at least one WRR queue in the remaining traffic range;

QoS control method using traffic processing according to service.
The method according to claim 10,

The queue state change message is

And a command for adding or deleting a queue to an output port of the network device.

QoS control method using traffic processing according to service.
The method according to claim 11,

Calculating a packet loss rate by receiving information on a lost packet from the network device; And

Changing a weight applied to each of the at least one WRR queue based on the packet loss rate;

QoS control method using traffic processing according to service.
The method according to claim 15,

Changing the weight applied to each of the at least one WRR queue,

And performing a command for changing the weight in the queue state change message and transmitting the same to the network device.

QoS control method using traffic processing according to service.
The method according to claim 10,

Receiving a queue status message including queue characteristic information for the at least one queue set based on the queue status change message;

Generating a flow entry for mapping between the service and the at least one queue based on the queue characteristic information; And

And controlling the network device to process traffic for each service according to the generated flow entry.

QoS control method using traffic processing according to service.
In the method for the network device to control the QoS,

Transmitting a queue setting information message including information on a queue that can be supported by an output port of the network device to a controller;

Receiving a queue state change message generated by the controller based on the queue setting information message and a QoS policy; And

Setting at least one queue to the output port according to the queue state change message;

QoS control method using traffic processing according to service.
The method according to claim 18,

The queue state change message is

Characterized by the controller by mapping between the service and the at least one queue based on the queue setting information message and the QoS policy,

QoS control method using traffic processing according to service.
The method according to claim 18,

The queue state change message is

Characterized by the controller by mapping between the network and the at least one queue according to the network type priority based on the queue configuration information message and the QoS policy,

QoS control method using traffic processing according to service.
The method according to claim 18,

Setting the output port,

Characterized in that the traffic for the highest priority service is set to be assigned to the Priority queue and the traffic for the next highest service is sequentially assigned to the at least one WRR queue.

QoS control method using traffic processing according to service.
The method according to claim 21,

Setting the output port,

Characterized in that the traffic for the highest-ranked service is processed and traffic for the next-ranked services is sequentially assigned based on a weight applied to each of the at least one WRR queue in the remaining traffic range.

QoS control method using traffic processing according to service.