KR20070061126A - Packet processor in ip version 6 router and method thereof - Google Patents

Abstract

An apparatus and a method for processing a packet in an IPv6 router are provided to easily obtain protocol and port information of a source or a destination by analyzing a flow label without having to search several heads, and obtain 4-tuple information in real time by using a hardware-wise hash scheme. When a packet is received(201), an input processing unit verifies whether the received packet has an error or not(202). A flow label processing unit obtains a 4-tuple hash value from the received packet(203) and checks whether the obtained 4-tuple hash value exists in a 4-tuple hash table(204). If the 4-tuple hash value does not exist in the 4-tuple hash table, the flow label processing unit reads a QoS parameter, a policy table value and an FIB table value as service profile data from an FIB table and a QoS/policy information table(206). The flow label processing unit stores a new 4-tuple hash value and corresponding service profile data as a new entry in a hash table(207). A flow controller checks an output interface with reference to the read service profile data, and processes QoS by setting a queue and a scheduler(208). A QoS controller processes queuing and a scheduler according to the setting(209) and an output processing unit outputs a packet according to the processing result(210).

Description

Packet processor in IP version 6 router and method thereof

1 is a block diagram of an IPv6 network to which an IPv6 packet processing apparatus according to the present invention is applied.

FIG. 2 is a diagram showing a form of a detailed structural diagram of an IPv6 header generated in an IPv6 terminal.

3 is a block diagram schematically illustrating an apparatus for processing a packet in an IPv6 router according to the present invention.

4 is a flowchart illustrating an embodiment of a method for processing an IPv6 packet according to the present invention, which is performed when a device shown in FIG. 3 is inputted with a packet having an IPv6 header of the type shown in FIG.

5 is a flowchart for describing in detail a 4-tuple hash information collection routine performed by a flow label processing unit.

The present invention relates to an IP version 6 (IPv6) router. In particular, when a packet including a flow label in which transport layer information is recorded is transmitted, the flow label is analyzed to extract 4-tuple information, and the extracted 4-tuple information is extracted. The present invention relates to a packet processing apparatus and method in an IPv6 router for providing differentiation of QoS services for each flow and extending the level of fineness of differentiation services.

 Recently, the Internet has evolved from IPv4 to IPv6 due to the exhaustion of addresses. Accordingly, the heads of frames transmitted under the IPv6 addressing scheme have been modified to be simplified in the router through standardization as well as simplification of the Internet protocol head. On the other hand, in IPv6, in order to provide Internet QoS, a session is supported through a method that supports TC (Traffic Class), which bundles and processes traffic groups in the same way as the DSCP (DiffServ Code Point) method used in IPv4. There is a way to use the FlowLabel to inform the router of the appropriate QoS parameters and to distinguish them in future packets. In the case of using the flow label, a standard has been proposed to use the flow label for guaranteeing the quality of service for each flow and for the rapid processing of the data of the flow. In other words, when using IPv6 flow labels, when receiving packets with multiple IPv6 heads from the network device, each stream is distinguished through the flow label of the first head without identifying all the heads and differentiated by each session. Can be. However, in order to fully use the flow label, each time a terminal establishes a session, functions that need to classify and manage sessions through flow classification using a signaling protocol such as RSVP (Resource Reservation Protocol) and mapping of appropriate QoS parameters are provided. need. When using such a function, the router has to process too many signals, so it is not easy to be applied to the network due to overhead.

Alternatively, in some cases, when QoS parameter information is transmitted to the head information processed every hop when transmitting the initial data packet without the signaling protocol, the router processing the first packet may process the same as the QoS providing method by the signaling protocol. Some techniques are proposed, but they are still difficult to apply to networks.

The technical problem to be achieved by the present invention is to refine the flow classification and traffic classification in real time by analyzing the head of the first IPv6 packet using a hardware hash technique when an IPv6 packet having a special meaning is assigned to the flow label from the terminal. The present invention relates to a packet processing apparatus and method in an IPv6 router capable of processing QoS.

In order to achieve the above technical problem, an IPv6 router according to the present invention for processing an IPv6 packet including a protocol and a flow label field in which a source or destination port information is recorded, a source IP address, and a destination IP address information in a head region. In the packet processing apparatus is an input processing unit for receiving an IPv6 packet, the destination IP address, the source IP address, the protocol and the source or destination port information from the head of the received IPv6 packet 4 for packet classification A flow control unit for processing the QoS of the IPv6 by setting a queue and a scheduler according to the extracted 4 tuple hash value, a QoS processing unit for processing the queuing and scheduling of the IPv6 according to the setting, and the queuing and It is preferable to include an output processor for outputting an IPv6 packet according to the scheduling process result.

In order to achieve the above technical problem, an IPv6 packet processing method for processing an IPv6 packet including a protocol and a flow label field in which a source or destination port information is recorded, a source IP address, and a destination IP address information in the head region is described above. (A) receiving an IPv6 packet, collecting the protocol information, the source or destination port information, the source IP address information, and the destination IP address information from the received IPv6 packet as 4 tuple hash information ( b) setting a queue and a scheduler according to the service profile data corresponding to the collected 4-tuple hash information by referring to the 4-tuple hash table storing service profile data for setting the queue and the scheduler according to the 4-tuple hash information. Processing the queuing and scheduling of the IPv6 packet, and IPv6 according to the result of the queuing and scheduling process. And the inclusion of (d) outputting a kit is preferred.

Hereinafter, an apparatus and a method for processing a packet in an IPv6 router according to the present invention will be described with reference to the accompanying drawings.

1 is a block diagram of an IPv6 network to which an IPv6 packet processing apparatus according to the present invention is applied.

Referring to FIG. 1, a plurality of IPv6 terminals 10 and 90 using an IPv6 addressing scheme includes aggregation of a plurality of IPv6 routers 30, 40, 50 and 70 and IPv6 terminals 10 and 90. Is connected to the IPv6 network 60 through the L2 equipment 20, 80. In the IPv6 network 60, three-layer processing is performed in a plurality of routers 30, 40, 50, and 70 that support IPv6. Data communication in the IPv6 network is the same as in the existing IPv4 network, but the address system is changed, so the processing of the part is wrong.

Meanwhile, the IPv6 terminal records protocols and meanings of the source and destination port information in the flow label when generating the flow label, so that the application programs can be easily identified by the transmission network equipment to provide differentiated services between the applications. can do. In particular, it uses traffic classes for differentiating services in IPv6 networks, and divides the flow label field to include source and destination port information in order to subdivide it, so that devices such as routers in the network can find source or destination protocol and port information. By analyzing the flow label without having to go to the head, it is easy to classify it into a flow based on 4 tuple information, and generate an IPv6 header as shown in FIG.

FIG. 2 is a diagram showing a form of a detailed structural diagram of an IPv6 header generated in an IPv6 terminal.

Referring to FIG. 2, the IPv6 head includes 4 bits of version information 410, 1 byte of traffic class 420, 20 bits of flow label 500, and data payload. Length 430, concatenated next head information 440, hop limit 450 indicating the hop count through which the current packet can flow, source IPv6 address field 460, destination IPv6 address field 470. In order to make the 20-bit flow label 500 in the IPv6 terminal, the 3-bit protocol field 510, the 1-bit SD field 520, and the 16-bit source or destination port field 530 are used. It is preferable to form separately. In this case, the 3-bit protocol field 510 records a value of 0 (ie, '000') 511 if it is not intended to be used, and 1 ('for UDP, which is currently used a lot. 001 ') (512), and the value of TCP as 2 (' 010 ') (513), and 3 (' 011 ') (514) for SCTP (Stream Control Protocol) In case of DCCP (datagram congestion control protocol), it is expressed as 4 ('100') (515), and it is expressed as 5 ('101') or 6 ('110') for other protocols. If a special control meaning is given, it is set to be expressed as 7 ('111') 516. Meaning of each protocol described above is merely an example and may be variously implemented. For example, the figure shows that 1 ('001') is written for UDP and 2 ('010') is written for TCP, but 2 ('010') is used for UDP and 1 ('010') is used for TCP. '001'). This record may also apply to other protocols. In addition, the SD field 520 is used to distinguish whether the content recorded in the port field 530 is a source protocol port or a destination protocol port, and 0 (521) for the source protocol field. In the case of the destination protocol field, a value of 1 522 may be recorded, and vice versa.

Meanwhile, although the number of bits for each field is shown in FIG. 3 as an example, the number of bits may be implemented differently for each field.

As described above, when forming the header of IPv6 in the IPv6 terminal, the desired 4-tuple information (source address information, destination address information, protocol information, source or destination) based on 5-tuple information that can divide each flow. Port information) so that the 4-tuple information can be found even if only the first header information of IPv6 is viewed. In this case, since 20-bit flow label is used, there is a problem in that it cannot include all 16-bit source or destination protocol fields, but it is possible to finely classify in network devices without special signaling protocol, so that future service diversification and service-specific If differentiation is required, it can be handled appropriately.

3 is a block diagram schematically illustrating a packet processing apparatus in an IPv6 router according to the present invention, and includes a router controller 110 and a line interface packet processor 160.

Referring to FIG. 3, the router controller 110 generally includes a management unit 111 for interworking with an administrator, a routing protocol processing unit 112 for operating as a router, a signaling protocol processing unit for processing RSVP, MPLS signaling, etc. 113), a QoS control unit 114 for controlling QoS processing, and a policy control unit 115 for managing the overall policy.

The line interface packet processor 160 classifies the packet from the input processor 120 for processing the input packet, the destination IP address, the source IP address, the protocol, and the source or destination port information from the head of the received IPv6 packet. Queue processing and scheduling according to the settings of the flow control unit 130 and the flow control unit 130 to process the QoS of the IPv6 by setting the queue and the scheduler according to the extracted 4-tuple hash value, and setting the queue and the scheduler according to the extracted 4-tuple hash value. QoS control unit 140, and an output processing unit 150 for processing the output packet. Specifically, the input processor 120 includes an L1 / L2 processor 121 for L1 / L2 line matching and an L3 IPv6 processor 122 for IPv6 routing. The flow control unit 130 stores a 4-tuple hash value consisting of a source IPv6 address, a destination IPv6 address, a protocol, and a source or destination port information, and a 4-tuple hash table 131 storing service profile data corresponding thereto. And a flow label processing unit for performing packet flow processing using the FIB table 132 storing information on an output interface, the QoS table 133 storing QoS information and policy information, and these tables 131, 132, 133. 134 is configured. The QoS controller 140 includes a queuing unit 141 and a scheduling unit 141 for transmitting a packet, and performs packet processing based on a required quality of service and a policy. The output processing unit 150 includes an L1 / L2 processing unit 151 for L1 / L2 line matching and a unit 152 for L3 header processing, and interoperates with neighboring routers through them.

FIG. 4 is a flowchart illustrating an embodiment of a packet processing method according to the present invention, which is performed when a device shown in FIG. 3 receives a packet having an IPv6 header of the type shown in FIG. 2.

3 and 4, when a packet is received (step 201), the input processor 120 verifies an error of the received packet (step 202).

After step 202, the flow label processing unit 134 performs a 4-tuple hash value extraction routine (step 300) to obtain a 4-tuple hash value from the received packet (step 203), and the obtained 4-tuple hash value is 4-tuple. It is checked whether or not it exists in the hash table 131 (step 204). A routine (step 300) of analyzing the header of the received packet and collecting the 4-tuple hash value will be described in detail with reference to FIG.

In step 204, if the 4-tuple hash value obtained through the routine of step 300 exists in the 4-tuple hash table 131, the flow label processor 134 may process 4 to process the current input packet in the same manner as the existing service profile. With reference to the tuple hash table 131, service profile data for the corresponding 4-tuple hash value is read and mapped to a queue (step 205). Here, the 4-tuple hash value extracted in step 300 is in the 4-tuple hash table 131, and the packet has the same condition as that of the corresponding packet, and the service profile data is obtained from the 4-tuple hash table. Get the QoS parameters. On the other hand, if the 4-tuple hash value obtained from step 300 does not exist in the 4-tuple hash table 131, the flow label processing unit 134 may transmit the packet corresponding to the packet from the FIB table 132 and the QoS / policy information table 133. The QoS parameters, policy table values, and FIB table values are read as service profile data (step 206). After operation 206, the flow label processing unit 134 stores the new 4-tuple hash value obtained in operation 300 and service profile data thereof as a new entry in the hash table (operation 207). After operation 207, the service interface data read in operation 206 is checked to set an output interface and set to process QoS through setting of a queue and a scheduler (operation 208).

After step 205 or step 208, the QoS control unit 140 processes the queuing and scheduler according to the setting of step 205 or step 208 (step 209), and the output processing unit 150 of the QoS control unit 140 The packet is output according to the processing result (step 210).

FIG. 5 is a flowchart for describing in detail a four-tuple hash information collection routine (step 300) performed by the flow label processing unit 134.

2 and 5, the flow label processing unit 134 first checks the source IP address 460 and the destination IP address 470 in the IPv6 head of the input packet shown in FIG. 2 (operation 301). ).

After step 301, the flow label processing unit 134 analyzes the 20-bit flow label 500 to obtain tuple information used to subdivide and classify the flow, such as a protocol and port information of a transmission destination or a destination (302). step). Specifically, the flow label 500 obtains protocol information of the input packet according to the value of the 3-bit protocol field 510 (step 303). After operation 303, it is determined whether the content recorded in the port field 530 is a source protocol port or a destination protocol port in the flow label 500 according to the value of the 1-bit SD field 520. (Step 304). After step 304, a value of the protocol port is obtained from the port field 530 (step 306).

After step 306, a 4-tuple value is generated using the source IP address, the destination IP address, the protocol, and the source or destination port information obtained in steps 301 and 302 (step 306).

After operation 306, an index of the hash table is generated using a key value of the 4-tuple information generated in operation 306 (step 307).

As described above, the IPv6 terminal transmits an IPv6 packet in which information such as a protocol and information on a destination and a port of a destination are recorded in a flow label to an IPv6 head. Then, the transmitting network equipment checks the source IP address, the destination IP address, and the flow label value, extracts the protocol and the source or destination port information based on the information included in the flow label, and uses the source IP address. You can get four-tuple hash information that consists of address, destination IP address, protocol, source or destination port information. In other words, by analyzing the flow label without having to go to various heads to find the source or destination protocol and port information in a device such as a router of the network, it is easily classified into the corresponding flow based on 4 tuple information, and the level of differentiation of differentiation service is fine. It can be done. In addition, four-tuple information can be obtained in real time since the process of obtaining four-tuple information using a hardware hash technique is used to quickly process many flows.

The invention can also be embodied as computer readable code on a computer readable recording medium. The computer-readable recording medium includes all kinds of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like, which are also implemented in the form of a carrier wave (for example, transmission over the Internet). It also includes. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. And functional programs, codes and code segments for implementing the present invention can be easily inferred by programmers in the art to which the present invention belongs.

Although the preferred embodiments of the present invention have been shown and described above, the present invention is not limited to the above-described specific preferred embodiments, and the present invention belongs to the present invention without departing from the gist of the present invention as claimed in the claims. Various modifications can be made by those skilled in the art, and such changes are within the scope of the claims.

As described above, according to the packet processing apparatus and method in the IPv6 router according to the present invention, if an IPv6 packet in which information such as a protocol and information on a destination or a port of a destination is recorded in a flow label is received at an IPv6 head, Check the source IP address, the destination IP address and the flow label value from the IPv6 header, and classify the flow using the 4-tuple hash information obtained by extracting the protocol and the source or destination port information based on the information contained in the flow label. And the degree of quality assurance can be processed in detail. In addition, devices such as routers in the network can easily obtain the information by analyzing the flow label without having to go to various heads to find the source or destination protocol and port information. Since the process uses a hardware hash technique, four tuple information can be obtained in real time.

Claims (6)

A packet processing apparatus in an IPv6 router for processing an IPv6 packet including a flow label field in which a protocol and a source or a destination port information are recorded in a head area, a source IP address, and a destination IP address information, An input processing unit to receive the IPv6 packet; Extracts the destination IP address, the source IP address, the protocol and the source or destination port information from the head of the received IPv6 packet as a 4-tuple hash value for packet classification, and extracts the extracted 4-tuple hash value. A flow controller configured to process the QoS of the IPv6 by setting a queue and a scheduler according to the service; A QoS processing unit for processing the queuing and scheduling of the IPv6 according to the setting; And And an output processing unit for outputting an IPv6 packet according to a result of the queuing and scheduling processing. The flow label processing unit of claim 1, wherein the flow label processing unit A 4-tuple hash table storing a 4-tuple hash value consisting of the source IPv6 address, the destination IPv6 address, the protocol and the source or destination port information, and corresponding service profile data; An FIB table storing information about an output interface; A QoS table storing QoS information and policy information; And If the 4-tuple hash value is present in the 4-tuple hash table, a queuing and scheduler is set according to the corresponding service profile data, and if the 4-tuple hash value is not present in the 4-tuple hash table, the FIB table and the And a flow label processing unit for obtaining service profile data for a corresponding packet by referring to a QoS table and setting a queuing and a scheduler according to the obtained service profile data. The flow label processing unit of claim 2, wherein the flow label processing unit And if the 4-tuple hash value does not exist in the 4-tuple hash table, storing the service profile data corresponding to the 4-tuple hash value in the 4-tuple hash table. An IPv6 packet processing method for processing an IPv6 packet including a flow label field, a source IP address, and a destination IP address information in which a protocol and source or destination port information are recorded in a head area, (a) receiving the IPv6 packet; (b) collecting the protocol information, the source or destination port information, the source IP address information, and the destination IP address information from the received IPv6 packet as four tuple hash information; (c) referring to the 4-tuple hash table storing service profile data for setting queues and schedulers according to 4-tuple hash information, and setting the queue and scheduler according to the service profile data corresponding to the collected 4-tuple hash information. Processing queuing and scheduling of the IPv6 packet by doing so; And (d) outputting an IPv6 packet according to the result of the queuing and scheduling process. The method of claim 4, wherein (c1) If the 4-tuple hash value does not exist in the 4-tuple hash table, an output interface for the packet is referred to by referring to an FIB table storing information on an output interface and a QoS table storing QoS information and policy information. Obtaining QoS information and policy information as service profile data; (c2) newly storing the 4-tuple hash value of the step (c1) and service profile data thereof in the 4-tuple hash table; And (c3) processing the queuing and scheduling of the IPv6 packet by setting a queue and a scheduler according to the obtained service profile data. The method of claim 4, wherein step (b) (b1) obtaining source IP address information and destination IP address information at the head of the received IPv6 packet; (b2) analyzing the flow label field to obtain the protocol information and port information of the source or destination; (b3) generating 4-tuple information using source IP address information, destination IP address information, protocol information, and source or destination port information obtained in steps (b1) and (b2); And (b4) generating the 4-tuple hash value that is an index of the 4-tuple hash table by using a key value of the 4-tuple information.

Images