KR101224593B1 - Apparatus for scheduling of packet flow and method thereof in packet switching network - Google Patents

Abstract

The present invention detects an empty queue for each packet flow in a packet switching network that supports a null packet, which is a packet without information to be delivered, and a real packet having delivery information. The present invention relates to a packet flow scheduling method and apparatus for a packet switching network for determining a packet service method according to whether an empty queue exists.

According to the present invention, by generating and processing a null packet in an empty queue, it is possible to reduce the maximum delay of the packet switching network and to ensure the quality of service (QoS) in the packet switching network. have.

Packet Switching Network, Flow Scheduling, Empty Queues, Null Packets

Description

Apparatus and method for scheduling packet flow in packet switching network {APPARATUS FOR SCHEDULING OF PACKET FLOW AND METHOD THEREOF IN PACKET SWITCHING NETWORK}

1 is a flowchart illustrating a packet flow scheduling method of a packet switching network according to a first embodiment of the present invention;

2 is a flowchart illustrating an empty queue processing step for each packet flow in the packet flow scheduling method of FIG. 1;

3 is an operation flowchart illustrating an aggregate packet flow generation step in the packet flow scheduling method of FIG. 1;

4 is a flowchart illustrating an operation of separating an aggregated packet flow in the packet flow scheduling method of FIG. 1;

FIG. 5 is a flowchart illustrating an operation of identifying a received packet flow and setting a maximum transmission allowance value in the packet flow scheduling method of FIG. 1;

6 is a flowchart illustrating an operation of performing a service for a packet flow in the packet flow scheduling method of FIG. 1;

7 is a block diagram showing a packet flow scheduling apparatus according to a first embodiment of the present invention;

8 is an exemplary view showing the operation of a system to which the packet flow scheduling apparatus and method of the present invention are applied;

9 is an exemplary view showing the structure of a switch node for explaining a packet flow scheduling method according to the present invention;

10 is an exemplary diagram illustrating a queue processing of a packet switch in a packet flow scheduling method according to the present invention;

11 is an exemplary diagram illustrating a structure of a switch node that receives an aggregation flow in a packet flow scheduling method according to the present invention;

12 is an exemplary diagram illustrating an aggregation and a aggregation model of a flow according to an embodiment of the present invention;

13 is an exemplary diagram illustrating a delay range of a flow having aggregation and aggregation according to an embodiment of the present invention;

14 is an exemplary diagram illustrating flow aggregation according to an end-to-end delay of the present invention;

15 is an exemplary diagram illustrating aggregation and aggregation at each switch node;

16 is an exemplary view showing aggregation of an aggregation flow according to the present invention;

17 is an exemplary view showing the configuration of a packet flow scheduling apparatus according to the present invention;

18A, 18B, and 18C are exemplary diagrams illustrating a packet service execution step of the packet flow scheduling method according to the present invention.

Description of the Related Art

720: receiver 730: controller

740: transmission unit 750: storage unit

The present invention relates to a method and apparatus for scheduling a packet flow in a packet switching network. More particularly, the present invention relates to a packet flow scheduling method, which detects an empty queue for each packet flow and determines a packet service method according to the existence of an empty queue. A packet flow scheduling method and apparatus for transmitting data according to a scheme.

In general, in a packet switching network that transmits and receives simple control signals or data information or transmits real-time streaming data between AV devices, a problem of ensuring quality of service (QoS) is continuously studied.

With the recent proliferation of service users and the emergence of applications such as video-on-demand and video conferencing where a large amount of data must be transmitted in real time, it is difficult to guarantee end-to-end QoS for real-time transmission. There is this.

In order to solve this problem, the present invention provides a packet flow scheduling apparatus in a packet switching network to reduce the maximum delay by improving the quality of service (QoS) by generating a null packet in an empty queue to service it; The purpose is to provide a method.

In order to achieve the above object, the packet flow scheduling method of the present invention is a packet switching that supports a service packet including a null packet not including delivery information and a real packet including delivery information. A packet flow scheduling method of a network, comprising: receiving a packet flow and processing an empty queue for each packet flow; And performing a service on the processed packet flow.

In the packet flow scheduling method of the present invention, the empty queue processing may include determining whether the empty queue exists and, as the empty queue exists, the null in the empty queue. Generating a packet.

Meanwhile, in the packet flow scheduling method of the present invention, the size of the null packet may be set to a value obtained by adding a quantity value of a service to be received by one flow to a remaining maximum transmission allowance value. .

In addition, the packet flow scheduling method of the present invention may further include a packet flow attribute comparison step of comparing the attributes of the packet flows and aggregating packet flows having similar attributes among the packet flows to generate an aggregate packet flow. have.

On the other hand, the packet flow scheduling method of the present invention may further comprise the step of receiving the aggregate packet flow and the aggregate flow separation step of obtaining the packet flows by the aggregation of the aggregate packet flow.

In addition, the packet flow scheduling method of the present invention preferably includes identifying the received packet flow and setting the maximum transmission allowance value with reference to the packet flow.

On the other hand, the step of setting the maximum transmission allowance value of the packet flow scheduling method of the present invention preferably comprises the step of setting the remaining maximum transmission allowance value to the value of the service value (Quantum value) to receive one flow. .

In addition, the performing of the service on the packet flow of the packet flow scheduling method of the present invention comprises comparing the size of the service packet with the size of the remaining maximum transmission allowance value and the packet selected according to the comparison result of the size comparison step. It is preferable to include the step of executing a service.

On the other hand, the packet service execution step of the packet flow scheduling method of the present invention determines whether the service packet is the null packet as the size of the remaining maximum transmission allowance value is greater than or equal to the size of the service packet. And performing a service in a media access control layer (MAC Layer) on the null packet as the service packet is the null packet and updating the remaining maximum transmission allowance value. It is preferable.

In addition, the packet service execution step of the packet flow scheduling method of the present invention may be performed by performing a service in a media access control layer on the real packet and performing the real packet as the service packet is the real packet. Preferably, the method includes transmitting to the PHY layer and updating the remaining maximum transmission allowance value.

On the other hand, the updating of the maximum transmission allowance value of the packet flow scheduling method of the present invention includes setting a value obtained by subtracting the size of the service packet from the remaining maximum transmission allowance value to a new remaining maximum transmission allowance value. It is desirable to.

In addition, the packet service execution step of the packet flow scheduling method of the present invention preferably includes waiting for the packet as the size of the remaining maximum transmission allowance value is smaller than the size of the service packet.

Meanwhile, the packet flow scheduling apparatus of the present invention is a packet flow scheduling apparatus of a packet switching network supporting a service packet including a null packet not including delivery information and a real packet including delivery information. And a controller configured to receive a packet flow, process the empty queue for each packet flow by receiving the packet flow from the receiver, and perform a service on the processed packet flow.

In addition, the control unit of the packet flow scheduling apparatus of the present invention determines whether the empty queue exists, and as the empty queue exists, generating the null packet in the empty queue. It may include.

Meanwhile, the control unit of the packet flow scheduling apparatus of the present invention sets the size of the null packet to a value obtained by adding a quantity value of a service to be received by one flow to a maximum remaining allowable transmission value. It is preferable.

In addition, the controller of the packet flow scheduling apparatus of the present invention may compare the attributes of the packet flows and aggregate the packet flows having similar attributes among the packet flows to generate an aggregated packet flow.

Meanwhile, the receiving unit of the packet flow scheduling apparatus of the present invention may receive the aggregate packet flow, and the controller may segment the aggregate packet flow to obtain packet flows.

In addition, the control unit of the packet flow scheduling apparatus of the present invention preferably identifies the packet flow and sets the remaining maximum transmission allowance value with reference to the packet flow.

On the other hand, the control unit of the packet flow scheduling apparatus of the present invention preferably sets the remaining maximum transmission allowance (Deficit) value to the value of the service value (Quantum value) to receive one flow.

In addition, the control unit of the packet flow scheduling apparatus of the present invention compares the size of the service packet and the size of the remaining maximum transmission allowance (Deficit) value, and executes the packet service selected according to the comparison result of the size comparison step It is preferable.

On the other hand, the controller of the packet flow scheduling apparatus of the present invention determines whether the service packet is the null packet as the size of the remaining maximum transmission allowance value is greater than or equal to the size of the service packet. As the service packet is the null packet, it is preferable to perform a service in a media access control layer (MAC Layer) for the null packet and to update the remaining maximum transmission allowance value.

The control unit of the packet flow scheduling apparatus of the present invention performs a service in a media access control layer with respect to the real packet as the service packet is the real packet, and transmits the real packet to a physical layer (PHY). layer), and it is preferable to update the remaining maximum transmission allowance value.

On the other hand, the control unit of the packet flow scheduling apparatus of the present invention preferably sets the new maximum transmission allowance value by subtracting the size of the service packet from the remaining maximum transmission allowance value.

In addition, the controller of the packet flow scheduling apparatus of the present invention preferably waits for the packet as the size of the remaining maximum transmission allowance value is smaller than the size of the service packet.

Meanwhile, the packet flow scheduling apparatus of the present invention includes an ID (ID), a flag, a service value to be received by one flow, a remaining maximum transmission allowance value, and an initial packet pointer. The apparatus may further include a storage unit configured to store packet flow information and packet information, such as a next packet pointer.

The operation principle of the preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear. The following terms are defined in consideration of the functions of the present invention, and may be changed according to the intentions or customs of the user, the operator, and the like. Therefore, the definition should be based on the contents throughout this specification.

1 is a flowchart illustrating a packet flow scheduling method of a packet switching network according to a first embodiment of the present invention.

First, a packet including information of a service packet is received through the receiving unit 720 of the packet flow scheduling apparatus 710 according to the present invention (S110).

Here, in a packet switching network that transmits and receives data packets, a service bandwidth is reserved for each traffic unit called 'flow'. In order to guarantee this reserved service bandwidth, the packet switching network can efficiently apply the allocated bandwidth to the traffic of a specific flow using a scheduling algorithm.

In addition, the scheduling algorithm in IntServs (Integrated Services), where many flows are handled independently, selects or transmits the next packet to transmit based on given Performance Metrics, ie delay or loss. Determine when.

Meanwhile, the schemes represented by the scheduling algorithm for providing bandwidth to the flow are packet-level generalized processor sharing (PGPS) and weighted fair queuing (WFQ).

In addition, the performance and variation of this traffic scheduling algorithm can be made by packet schedule estimation of accurate schedulers based on the flow flow model of the fluidity, and provides minimum delay and maximum fairness to each flow while being provided at the promised rate. Can provide.

Meanwhile, according to the flow management, the leaky-bucket parameter is included in the traffic specification and the reservation rate is included in the request specification.

In addition, DiffServs (Differentiated services), which controls the quality of service based on the type of traffic, prioritizes traffic without securing a bandwidth when communicating, and prioritizes control in a LAN switch or router according to header or packet header information. Run

Here, DiffServs classifies flows by class and schedules them, thereby processing a single flow or a bulk flow.

Thereafter, the control unit 730 of the packet flow scheduling apparatus 710 processes an empty queue for each received packet flow (S120).

Thereafter, the controller 730 may perform a service on the processed packet flow (S130). In the case of the service for the packet flow, the size of the service packet is compared with the size of the remaining maximum transmission allowance value, and the selected packet service is executed according to the size comparison result.

2 is a flowchart illustrating an empty queue processing step for each packet flow in the packet flow scheduling method of FIG. 1.

First, the controller 730 of the packet flow scheduling apparatus 710 determines whether an empty queue exists (S210).

Here, a queue is a kind of buffer that is stored and waiting for data information to be processed. The best known behavior of queues is FIFO (First-In-First-Out) queuing.

For a FIFO queue, the first element in the queue comes first.

A queue is a natural data structure for a system that serves incoming requests. In operating systems, most process scheduling uses queues.

As a result of determining the existence of the empty queue, the control unit 730 generates a null packet in the empty queue when there is an empty queue (S220).

Herein, the size of a null packet may be set to a value obtained by adding a quantity value of a service to be received by one flow to a remaining maximum transmission allowance value.

3 is an operation flowchart illustrating an aggregate packet flow generation step in the packet flow scheduling method of FIG. 1.

First, the controller 730 of the packet flow scheduling apparatus 710 may compare the attributes of the packet flows (S310).

Subsequently, the controller 730 may aggregate packet flows having similar attributes among the packet flows (S320) and generate an aggregated packet flow (S330). Here, the packet or flows containing the packet are classified into traffic classes, such as expedited forwarding (EF), assured forwarding (AF) and best effort (BE) service classes, and have the same quality of service (QoS) requirements per class. Aggregate traffic with

4 is a flowchart illustrating an operation of separating an aggregated packet flow in the packet flow scheduling method of FIG. 1.

First, the controller 730 of the packet flow scheduling apparatus 710 may receive an aggregate packet flow (S410).

Subsequently, the controller 730 may segment the aggregate packet flow (S420), and may acquire packet flows (S430).

5 is a flowchart illustrating an operation of identifying a received packet flow and setting a maximum transmission allowance value in the packet flow scheduling method of FIG. 1.

First, the controller 730 of the packet flow scheduling apparatus 710 may read and identify the packet flow information from the storage unit 750 in which the information of the packet flow received through the receiver 720 is stored (S510). ).

Subsequently, the controller 730 sets the maximum transmission allowance value to a value obtained by adding a quantity value of a service to be received by one flow to the existing maximum transmission allowance value with reference to the packet flow information (S520). The value may be stored in the storage unit 750.

6 is a flowchart illustrating an operation of performing a service for a packet flow in the packet flow scheduling method of FIG. 1.

First, the control unit 730 of the packet flow scheduling apparatus 710 when the size of the remaining maximum transmission allowance value stored in the storage unit 750 is greater than or equal to the size of the service packet (S610), the service packet is null It is determined whether the packet (S620).

If the service packet is a null packet, the controller 730 may execute a service in a medium layer on a null packet (S630).

On the other hand, if the service packet is a real packet, the controller 730 executes the service in the medium access control layer for the real packet, and delivers the real packet to the physical layer (PHY layer) for transmission. The transmission can be made through the unit 740 (S650).

The controller 730 may set the maximum transmission allowance value as a new maximum transmission allowance value by subtracting the size of the service packet from the existing maximum transmission allowance value (S640) and store it in the storage unit 750.

In addition, when the size of the remaining maximum transmission allowance value stored in the storage unit 750 is smaller than the size of the service packet, the controller 730 may queue the packet (S660).

7 is a block diagram illustrating a packet flow scheduling apparatus according to a first embodiment of the present invention.

The packet flow scheduling apparatus 710 of the present invention may include a receiver 720, a controller 730, a transmitter 740, and a storage 750.

Here, the receiver 720 may serve to receive a packet flow including information of a service packet.

In addition, the storage unit 750 may include an ID (ID) of a packet flow, a flag, a quantity of services to be received by one flow, a remaining maximum transmission allowance value, an initial packet pointer, and a next packet. It plays a role of storing packet flow information and packet information such as a next packet pointer, but is not limited thereto.

The controller 730 receives a packet flow from the receiver 720 and processes an empty queue for each packet flow, but is not limited thereto.

In addition, the controller 730 stores the information of the received service packet, which performs a service on the processed packet flow, in the storage 750, but is not limited thereto.

On the other hand, the control unit 730 is the ID (ID), the flag (Flag) of the packet flow stored in the storage unit 750, the quantity of service (Quantum value), the maximum transmission allowance value, the first packet pointer (Initial Packet) Pointer, a packet flow information such as Next Packet Pointer (Read Next), and the packet information to read, update, perform a packet service, and transmits through the transmission unit 740, but is not limited thereto.

In addition, the controller 730 compares the attributes of the packet flows, aggregates packet flows having similar attributes among the packet flows to generate an aggregate packet flow, aggregates the aggregate packet flows, and separates the aggregate packet flows. Packet flows may be generated, but are not limited thereto.

Meanwhile, the transmitter 740 plays a role of transmitting a packet, but is not limited thereto.

8 is an exemplary view showing the operation of a system to which the packet flow scheduling apparatus and method of the present invention are applied.

As shown in FIG. 8, the packet switching network may include a first switch node 810 and a second switch node 820 to a seventh switch node 870.

In addition, each switch node may control the transmission of the data packet by the switch control unit provided internally.

Here, the first switch node 810 includes a first switch control unit 880, the second switch node 820 includes a second switch control unit 890, and the seventh switch node 870 includes a seventh switch node. The switch controller 940 may be provided.

In the packet switching network of FIG. 8, the data packet to be transmitted may include a first packet flow to a fourteenth packet flow.

The first packet flow and the second packet flow may be transmitted to the seventh switch node 870 via the first switch node 810 and the second switch node 820.

In contrast, the third to eighth packet flows may be scheduled only through the first switch node 810, and the ninth and fourteenth packet flows may be scheduled only through the second switch node 820. . In addition, the sixth packet flow may be scheduled via the seventh switch node 870.

Here, since the first packet flow and the second packet flow are scheduled from the first switch node 810 to the seventh switch node 870, similar characteristics may exist in the transmission path.

In addition, since the data packet of each packet flow is transmitted via the first switch node 810 to the seventh switch node 870, the first packet flow and the second packet flow are each switched according to the scheduling of each switch node. Latency can occur for each node.

9 illustrates a structure of a switch node for explaining a packet flow scheduling method according to the present invention.

As shown in FIG. 9, the first switch node 810 to which flows (i) and (j) are input may include a first switch controller 880 serving as a latency-rate (LR) server. have. In addition, it may include a storage unit, a receiver, and a transmitter.

Here, the storage unit may include a packet memory unit for storing data packets input from the outside and a flow information memory unit for storing information of data packets according to each flow.

Here, the controller services an activated flow among a plurality of flows according to each flow information, and does not service a deactivated flow. In addition, the controller has an active list for activated flows and non-activated flows, and may perform a flow service according to an active list based on flow information.

On the other hand, the first switch node 810 may compare whether there are flows with similar attributes among a plurality of input flows. Here, the property may include a transmission path of each flow.

In FIG. 9, when it is determined that the attributes of the flows (i) and the flows (j) are similar, the first switch control unit 880 aggregates the flows (i) and the flow (j) at the first switch node 810. The flow g can be generated. The first switch controller 880 may transmit the aggregation flow g to the packet switching network through the transmitter.

10 illustrates queue processing of a packet switch in a packet flow scheduling method according to the present invention.

In the first switch node 810, the packet switch S plays a role of a scheduler. When a new flow is input from the outside while the current flow is serviced for each sequentially input flow, a new switch is added to the flow information storage unit. Information about the flow can be stored. The active list processor may be controlled to sequentially schedule the first flow to the nth flow.

11 illustrates the structure of a switch node that receives an aggregation flow in the packet flow scheduling method according to the present invention.

The seventh switch node 870 receiving the aggregation flow g from the packet switching network may include a seventh switch control unit 940 as an LR server.

When scheduling the input aggregation flow g, the seventh switch controller 940 may segment the aggregation flow g to obtain the original flow i and the flow j.

12 is a diagram illustrating an aggregation and a aggregation model of a flow according to the present invention.

Here, the server Sn provided in the current switch node segments the aggregation flow delivered from the server S (n-1) of the previous switch node, and displays one of the segmented flows and a newly input flow. It can be aggregated.

The server S (n-1) of the previous switch node may generate output traffic and transmit it to the current switch node Sn. The current switch node may segment the output traffic delivered from the server S (n-1) of the previous switch node at a receiver into a flow different from the flow. The server Sn of the current switch node may aggregate the flow and the newly input flow to generate and transmit an aggregate flow.

Figure 13 shows the delay range of the flow with aggregation and aggregation according to the present invention.

As shown in FIG. 13, the flows transmitted from the start node include the first server S1, the second server S2, the third server S3, the fourth server S4, the fifth server S5, And the sixth server S6 to be delivered to the endpoint node. Accordingly, the first server S1 to the sixth server S6 in the packet switching network may be referred to as the passthru server S. FIG. At this time, the packet maximum delay in the network may be calculated by the delay sum of the flows belonging to the packet traversing the network.

At some point in the network as shown in FIG. 9, the flow i is aggregated by the server S to produce an aggregation flow g, and as shown in FIG. 11, the aggregation flow g is at another point. It can be separated from and generated in flow (i).

14 is a diagram illustrating flow aggregation according to an end-to-end delay of the present invention.

In the packet switching network of FIG. 14, there are first to fourteenth flows for data packets to be transmitted.

The first switch controller 880 aggregates the first flow and the second flow with respect to the first flow, the first flow, and the third to eighth flows input to the first switch node 810 to form an aggregation flow ( 1,2 may be generated and transferred to the second switch node 820 through the transmitter.

At the second switch node 820, the second switch controller 890 schedules the aggregate flows (1, 2) and the ninth through fourteenth flows to be input, and transmits the aggregate flows (1, 2) through the transmitter to the packet switching network. Can be sent to. As such, the aggregation flow 1, 2 can be delivered to the seventh switch node 870 via other switch nodes in the packet switching network.

That is, the first flow and the second flow, which are the aggregation flows, may be transmitted to the seventh switch node 870 via the first switch node 810 and the second switch node 820. In contrast, the third to eighth flows may be scheduled via only the first switch node 810, and the ninth to fourteenth flows may be scheduled via only the second switch node 820. In addition, the sixth flow may be input to the seventh switch node 870, and the seventh switch controller 940 may schedule it.

15 is a diagram illustrating aggregation and aggregation at each switch node.

As shown in FIG. 15, the first switch node 810 and the second switch node 820 to the seventh switch node 870 constituting the packet switching network are configured to perform aggregation and aggregation according to the present invention. Can be done.

In FIG. 15, when the first flow and the seventh flow are input, the first switch node 810 aggregates the first flow and the seventh flow when the two flows have similar attributes. To generate an aggregation flow g. In addition, the first switch controller 880 of the first switch node 810 may transmit the aggregation flow g to the next second switch node 820.

The second switch node 820 may generate the separated first flow and the seventh flow by segmenting the input aggregation flow g. In addition, when a new second flow is input to the second switch node 820, the second switch controller 890 of the second switch node 820 is agglomerated when the second flow and the first flow have similar attributes. Can be aggregated to generate an aggregation flow f. In addition, the second switch controller 890 may transmit the aggregation flow f to the packet switch network through the transmitter.

When the aggregation flow f transmitted from the second switch node 820 is transferred directly to the seventh switch node via other switch nodes, the seventh switch node 870 segments the aggregation flow f. Create a first flow and a second flow. In this case, if there is no input for a new flow to the seventh switch node 870, the first flow is transmitted as it is, and when the seventh flow is newly input, the first flow and the seventh flow are aggregated when the attributes are similar. To generate an aggregation flow.

16 is a view showing the aggregation of the aggregation flow according to the present invention.

In FIG. 16, when the first flow and the second flow are input to the first switch node 810, the first switch controller 880 aggregates the first flow and the second flow to generate an aggregate flow to generate a packet switching network. Can be controlled to be sent.

The aggregation flow transmitted from the first switch node 810 may be transferred to the seventh switch node 870 via the second switch node 820 and other switch nodes.

The seventh switch node 870 may segment the input aggregation flow to obtain a first flow and a second flow, and may transmit them through different transmitters.

Accordingly, the delay while the data packet is transmitted from the first switch node 810 to the seventh switch node 870 is as follows.

In FIG. 16, if the first packet flow and the second packet flow are aggregated at the first switch node 810 and aggregated at the seventh switch node 870, the maximum packet size is 266 bytes and 10. In case of having the Mbps capacity, the maximum delay value during the transmission from the first switch node 810 to the seventh switch node 870 may be about 6 ms.

Meanwhile, the packet flow scheduling method of the present invention can reduce the maximum delay value to about 4 ms.

17 shows a configuration of a packet flow scheduling apparatus according to the present invention.

In FIG. 17, a packet flow scheduling apparatus according to the present invention includes a storage unit 750 including a packet memory 960 and a flow information storage unit 980, and an active list processor 980. The controller 730 and the transmitter 740 may be included.

The packet memory 960 stores a data packet 970 input from the outside. The packet memory 960 may use a large amount of SRAM, for example. As illustrated in FIG. 17, one data packet stored in the packet memory 960 includes a next packet flag, a next packet size, a next packet pointer, and a current packet. The packet size may include a packet size, a packet, and the like.

If a new flow is inputted from the outside while the service flows backlog after the current flow of the first flow, the controller 730 stores the information about the new flow in the next service order in the flow information storage unit 990 and the first flow. When the current service is terminated, the active list processor 980 may control the new flow to be serviced. When the service of the new flow is terminated, the controller 730 may service the second flow according to the scheduling, and then service the next flow such as the third flow, the fourth flow, ...

Here, when the control unit 730 services the first flow, the current packet pointer, the current flow pointer, and the amount of services to be received by one flow from the flow information storage unit 990. The first flow information such as a value, a maximum transmission allowance (Deficit) value, and a current packet size may be read, configured as a data packet to be serviced, and transmitted to the transmission unit 740.

The transmission unit 740 may buffer the data packet to be delivered received from the control unit 730 in a packet buffer and then serve the destination.

The active list processing unit 980 preferably services an activated flow among a plurality of flows according to each flow information, and does not service a deactivated flow. In addition, the active list processing unit 980 has an active list for activated flows and non-activated flows, and may perform a flow service operation according to an active list based on flow information.

The flow information storage unit 990 may store information of a data packet according to each flow. The flow information storage unit 990 may store a plurality of flow information, such as information about the first flow 1000 and information about the 1010 flow 1010. As shown in FIG. 17, information on each flow includes a flow ID, a flag, a value of a quantity to be received by one flow, a maximum remaining allowable value, and an initial packet pointer. Pointer, Next Flow Pointer, etc. may be included.

18A, 18B, and 18C illustrate an example of a packet service execution step of the packet flow scheduling method according to the present invention.

In FIG. 18A, packet service execution preferably serves data packets waiting in a queue for each flow in a first-in first-out (FIFO) manner. It is preferable to store one scheduling information for a head of line (HOL) in each queue. Subsequently, when the packet arrives, it may be checked whether the packet is a HOL packet, and if the packet is a HOL packet, scheduling information may be stored in a queue. The queue is assigned a quantity of service value to be received by one flow, and the scheduling information for data packets smaller than the counter value considering the quantity of service to be received by one flow is stored in the queue currently being serviced. If it is larger than the packet counter value, it can be stored in another queue. It is desirable to continue to service until the queue being served is completely empty, and to serve another scheduling queue when there is no backlog scheduling information in that queue. At this point a new round may begin. Each flow can be serviced in the same process in the new round.

18A and 18B, a packet 1020 of size 100 is serviced as a first service. Here, it is preferable that the maximum remaining allowable value (Deficit) value 1100 is 100 minus 100, which is the size of the serviced packet 1020, from the quantity value 1090 that one flow receives.

Since the 400 size packet 1030 of the second queue is larger than the service value 200 that one flow receives, it is preferable not to service the packet.

In addition, the 50th size packet 1040 of the third queue is serviced, and the remaining maximum transmission allowance value (Deficit) value 1120 is subtracted from the service value value received by one flow, which is the serviced packet size of 50. It is preferable to be 150.

Since the 340 size packet 1050 of the fourth queue is larger than 200, which is a value of a quantum value of one flow, it is preferable not to service it.

In FIG. 18B, since the third queue is confirmed to be empty, it is preferable to generate and service a null packet 1070. The size of a null packet is preferably set to 350, which is a value obtained by adding a maximum value of a residual transmission value 1120 to a quantity value of a service to be received by one flow.

As a next step, in Figs. 18B and 18C, 600 packets 1060 of the first queue are preferably not serviced because they are larger than the maximum transmission allowance value 1140 still remaining.

Then, the 400 size packet 1030 of the second queue is serviced, and the maximum transmission allowance value 1150 remaining is obtained by subtracting the service packet size 400 from the service value value one flow will receive. It is preferred to be zero.

In addition, the 350 size null packet 1070 of the third queue is serviced, and the maximum transmission allowance value 1160 remaining is the size of the serviced packet at the quantity value received by one flow. It is preferable to become 0 which subtracted 350.

In addition, the 340 size packet 1050 of the fourth queue is serviced, and the remaining maximum transmission allowance value 1170 is subtracted from the service packet value 340, which is the size of the service packet, to receive one flow. It is preferable to become 60.

In the next step, the third queue is empty, so it is desirable to generate and service null packets. The size of the null packet 1080 is preferably set to 200, which is a value obtained by adding a maximum value of the remaining allowable transmission value 1160 to a quantity value of a service to be received by one flow.

As described above, according to the present invention, a packet flow scheduling method and apparatus of a packet switching network for detecting an empty queue in a packet flow and determining a packet service method according to whether an empty queue exists. Can be realized.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention.

Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments.

The protection scope of the present invention should be interpreted by the following claims, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the present invention.

As described above, according to the present invention, by generating and processing a null packet in an empty queue, the maximum delay of the packet switching network is reduced and the quality of service (QoS) in the packet switching network is reduced. ) Is effective to ensure.

Claims (24)

A packet flow scheduling method of a packet switching network supporting a service packet including a null packet not including delivery information and a real packet including delivery information, Receiving a packet flow; Processing an empty queue for each packet flow; And Performing a service on the processed packet flow; The size of the null packet is, A method of scheduling a packet flow in a packet switching network, characterized in that the maximum transmission allowance (Deficit) value is added to a value of a service value (Quantum value) to be received by one flow. The method of claim 1, Processing the empty queue, Determining whether the empty queue exists; And Generating the null packet in the empty queue if the empty queue exists; Packet flow scheduling method of a packet switching network comprising a. delete The method of claim 1, A packet flow attribute comparing step of comparing the attributes of the packet flows; And Aggregating packet flows having similar attributes among the packet flows to generate an aggregate packet flow; The packet flow scheduling method of a packet switching network further comprising. 5. The method of claim 4, Receiving the aggregate packet flow; And Segmenting the aggregate packet flow to obtain packet flows; The packet flow scheduling method of a packet switching network further comprising. The method of claim 1, Identifying the received packet flow; And Setting a maximum transmission allowance value with reference to the packet flow; The packet flow scheduling method of a packet switching network further comprising. The method of claim 6, The setting of the maximum transmission allowance value may include: Setting a service value (Quantum value) to be received by one flow to a remaining maximum transmission allowance value;       Packet flow scheduling method of a packet switching network comprising a. The method of claim 1, Performing a service for the packet flow, Comparing the size of the service packet with the size of the remaining maximum transmission allowance value; And Executing the selected packet service according to the comparison result of the comparing of the sizes; Packet flow scheduling method of a packet switching network comprising a. 9. The method of claim 8, The step of executing the packet service, Determining whether the service packet is the null packet as the size of the remaining maximum transmission allowance value is greater than or equal to the size of the service packet; Executing a service in a media access control layer (MAC Layer) on the null packet if the service packet is the null packet; And       Updating the maximum transmission allowance value;       Packet flow scheduling method of a packet switching network comprising a. The method of claim 9, The step of executing the packet service, If the service packet is the real packet, executing a service in a media access control layer on the real packet; Transmitting the real packet to a physical layer; And Updating the maximum transmission allowance value; Packet flow scheduling method of a packet switching network comprising a. The method of claim 9, The updating of the maximum transmission allowance value may include: setting a value obtained by subtracting the size of the service packet from the remaining maximum transmission allowance value as a new maximum transmission allowance value; Packet flow scheduling method of a packet switching network comprising a. The method of claim 9, The step of executing the packet service,       Waiting for the packet as the size of the remaining maximum transmission allowance value is smaller than the size of the service packet;       Packet flow scheduling method of a packet switching network comprising a. A packet flow scheduling apparatus of a packet switching network that supports a service packet including a null packet not including delivery information and a real packet including delivery information, A receiver for receiving a packet flow; And And a controller configured to receive the packet flow from the receiving unit, process the empty queue for each packet flow, and perform a service on the processed packet flow. The control unit, And setting the size of the null packet to a value obtained by adding a quantum value of a single flow to a remaining maximum transmission allowance (Deficit) value. The method of claim 13, The controller determines whether the empty queue exists and, when the empty queue exists, generates the null packet in the empty queue. Device. delete The method of claim 13, The controller compares the attributes of the packet flows and aggregates the packet flows having similar attributes among the packet flows to generate an aggregate packet flow. 17. The method of claim 16, And the control unit segments the aggregate packet flow and separates the aggregate packet flow to generate packet flows. The method of claim 13, And the control unit identifies the received packet flow and sets the maximum transmission allowance value with reference to the packet flow. The method of claim 18, The control unit compares the size of the service packet with the size of the remaining maximum transmission allowance value, and executes the packet service selected according to the result of the comparison of the size, packet flow scheduling apparatus of a packet switching network. 20. The method of claim 19, The controller determines whether the service packet is the null packet as the size of the remaining maximum transmission allowance value is greater than or equal to the size of the service packet, and the service packet is the null packet. In this case, the packet flow scheduling apparatus of the packet switching network, characterized in that to perform a service in the MAC layer for the null packet, and to update the maximum transmission allowance value. 20. The method of claim 19, When the service packet is the real packet, the controller performs a service in a media access control layer on the real packet, transmits the real packet to a physical layer, and transmits the maximum packet. Packet flow scheduling apparatus of a packet switching network, characterized in that for updating the allowance value. 20. The method of claim 19, And the control unit sets a value obtained by subtracting the size of the service packet from the remaining maximum transmission allowance value to a new maximum transmission allowance value. 20. The method of claim 19, And the control unit waits for the packet as the size of the remaining maximum transmission allowance value is smaller than the size of the service packet. The method of claim 13, A storage unit for storing at least one of a quantity value and a maximum transmission allowance value to be received by the one flow; Packet flow scheduling apparatus of a packet switching network further comprising.

Images