KR100941118B1 - Apparatus having layer 3 switching function - Google Patents

Abstract

The present invention relates to an apparatus having a three-layer switching function, the apparatus comprising: a detector for detecting a reception route of a packet; And a registration unit for registering the reception route if the reception route is different from the previously registered route. According to the present invention, it is possible to perform three-layer switching to support the mobility of the terminal in terms of operability, efficiency, economics, and business.

Forwarding, routing, mobility, forwarding table

Description

Device with three-layer switching function {apparatus having layer 3 switching function}

1A to 1E are views for explaining the concept of the present invention.

2 is a block diagram illustrating an apparatus having a three-layer switching function according to an embodiment of the present invention.

The present invention relates to a three-layer (layer 3) switching technology, and more particularly to a device having a three-layer switching function that can efficiently support the mobility of the terminal.

Layer 3 switching means forwarding (delivering) a packet from one network to the next network (i.e. next hop) on the third layer of Open Systems Interconnection (OSI) to reach its final destination. Examples of devices having a three-layer switching function include a three-layer switch, a router, and an optical network unit (ONU) of an Ethernet-Passive Optical Network (E-PON). . In the present specification, a device having a three-layer switching function will be described as a router for convenience.

In order for a router to perform three-layer switching, routing information such as information about a route reaching each network, a metric (eg, hop count, cost, etc.) of the route must be obtained in advance. As a method of acquiring such routing information, there is a static routing method directly input by an administrator and a dynamic routing method obtained through a routing protocol for exchanging routing information known between routers. Examples of routing protocols include Routing Information Protocol (RIP), Open Shortest Path First (OSPF) protocol, and Border Gateway Protocol (BGP).

The router constructs a RIB including the routing information obtained by the above-described method as an entry, and extracts an optimal entry in terms of metrics from among the entries of the RIB to construct the FIB. That is, only routes selected as valid and valid in the RIB are included in the FIB as entries. Specifically, the FIB includes entries that bind each network address (the IP address of the network), the corresponding output port of the router, and router information of the next hop.

Such FIB may be directly used for switching, but after a longest prefix matching (LPM) table is established based on the entry of the FIB for faster retrieval, switching may be performed based on the LPM table. . In the present specification, the FIB and LPM tables are collectively referred to as a forwarding table.

In general, the LPM algorithm is used for switching using the forwarding table. According to the algorithm, the router has a network address that most matches among the entries in the forwarding table in the order of the upper bits of the destination address of the received packet. Select the entry to determine the next hop.

On the other hand, when a terminal transmits a packet after moving from the first network (or first subnet) to the second network (or second subnet) connected for the first time, a situation in which the packet cannot reach a desired destination or an inefficient route The situation arises that the destination is reached.

The former situation occurs when the router of the second network (or second subnet) detects that the route of the packet received from the terminal is different from the route of the forwarding table, and discards the packet. This function is called reverse path filtering. A router sees a packet received by a route different from the registered route as a kind of denial of service attack and discards it.

The latter situation occurs when a tunneling technique is applied to support the mobility of the terminal. In the case of tunneling, the packet transmitted after the terminal movement must go through the optimized path and the tunneling path sequentially, and thus use the inefficient route because it passes through the bypass path instead of the optimal path. In addition, each router must perform additional operations such as care of address (CoA) -binding, encapsulation, and decapsulation to support such tunneling schemes. It is a burden, and it is a great economic and business burden for the operators who build the network.

Therefore, in order to solve this problem, there is a need for a three-layer switching scheme that supports the mobility of the terminal in terms of operability, efficiency, economics, and business.

An object of the present invention is to provide a device having a three-layer switching function that can efficiently support the mobility of the terminal.

In order to achieve the above technical problem, a first aspect of the present invention provides a detection unit for detecting a reception route of a packet; And a registration unit for registering the reception route when the reception route is different from the previously registered route.

In order to achieve the above technical problem, a second aspect of the present invention includes a storage unit for storing each registered route; And a forwarding unit that performs packet forwarding even if the packet receiving route is different from the route stored in the storage unit.

Hereinafter, embodiments of the present invention for solving the above technical problem are described with reference to the accompanying drawings.

The present invention uses a scheme for the router to learn the path that the download packet should go based on the source address of the upload packet transmitted from the terminal. More specifically, the present invention is a method in which medium access control (MAC) address learning in a two-layer switch (or bridge) is applied to three-layer switching of a router, and the router is a destination IP address of a received packet. And forwarding by querying the forwarding table (forwarding procedure), and forwarding where the source IP address of the received packet, the source 2 layer address, and the route determined by the received port are matched from the LPM perspective. If different from the root of the table entry, information corresponding to the source IP address of the received packet, the source layer 2 address, and the received port is registered as an entry in the forwarding table (learning procedure). Here, in case of Ethernet communication, the source layer 2 address corresponds to the source MAC address. Preferably, the router deletes the entry from the forwarding table according to a predetermined aging procedure if the entry registered in the forwarding table is not used for a predetermined period in this manner.

1A to 1E are diagrams for explaining a concept of the present invention, and FIG. 1A is a diagram for describing a packet forwarding process (general packet forwarding process) after a terminal (ie, a mobile host) is initially connected to a network (or subnet). 1B is a diagram illustrating a process of managing a forwarding table according to the movement of a terminal, and FIG. 1C is a diagram illustrating a process of correctly delivering a packet from a server to a terminal through management of the forwarding table of FIG. 1B. 1D is a diagram for describing a process of managing a forwarding table as the terminal moves again, and FIG. 1E is a diagram for explaining a process of deleting a forwarding table entry additionally registered in FIG. 1B according to an aging procedure. It is for the drawing.

First, referring to FIG. 1A is as follows.

In FIG. 1A, a network environment consisting of network A, network B, network C, and network S is shown. The network environment may be an internetwork environment consisting of a plurality of networks or a network environment consisting of a plurality of subnets. . In the latter case, networks A, B, C, and S correspond to each subnet. Although the present invention is applicable to both the former network environment and the latter network environment, it is more preferable to apply to the latter network environment in terms of security.

In FIG. 1A, each router Ra, Rb, Rc, Rn, and Rs constructs a forwarding table shown in FIG. 1A using the obtained static / dynamic routing information or RIB, and performs packet forwarding according to the forwarding table. In FIG. 1A, since the terminal is connected to the network A for the first time, the packet transmitted from the terminal to the server is transmitted to the server through the router Ra, the router Rn, and the router Rs according to each forwarding table. The packet transmitted from the server to the terminal may reach the terminal through the router Rs, the router Rn, and the router Ra according to each forwarding table. That is, as shown in FIG. 1A, the terminal before the mobile station can transmit and receive a packet with the server according to the conventional method. Since the mobile station does not move, the learning process by the router is not performed.

1B is as follows. As illustrated in FIG. 1A, a terminal under network A moves and is under network B, which corresponds to FIG. 1B.

When the mobile station transmits a packet destined for the server, the router Rb does not perform reverse path filtering, looks up the destination IP address of the received packet, looks up the forwarding table, and determines the next hop as the router Rn. , Forward the received packet. In addition, the router (Rb) is the root of the forwarding table entry (192.168.2.0/24 nexthop Rn) matching the source IP address of the received packet in LPM perspective, the source IP address of the received packet, the receiving layer 2 address, and the received Detecting the route determined as the port is different, the information corresponding to the source IP address of the received packet, the receiving layer 2 address, and the received port is additionally registered as an entry in the forwarding table. In FIG. 1B, an underlined entry among the entries of the forwarding table of the router Rb is an entry registered additionally by the above-described method.

In this process, the packet is delivered to the router Rn, and the router Rn performs the same process. That is, the router Rn looks at the destination IP address of the received packet without performing reverse path filtering, queries the forwarding table, determines the next hop as the router Rs, and then forwards the received packet. In addition, the router Rn is configured to determine that the root of the forwarding table entry (192.168.2.0/24 nexthop Ra) that matches the source IP address of the received packet in terms of LPM is the source IP address of the received packet, the receive layer 2 address, and the received packet. Detecting the route determined as the port is different, additional information corresponding to the source IP address of the received packet, the receiving layer 2 address, and the received port is further registered as an entry in the forwarding table. In FIG. 1B, an underlined entry among the entries of the forwarding table of the router Rn is an entry registered additionally by the above-described method.

In this process, the packet is delivered to the router Rs, and the router Rs performs the same process. That is, the router Rs does not perform reverse path filtering, looks up the destination IP address of the received packet, looks up the forwarding table, determines that the next hop is the local network, and then forwards the received packet to a server in the local network. In addition, the router (Rs) is configured such that the root of the forwarding table entry (192.168.2.0/24 nexthop Rn) matching the source IP address of the received packet from the LPM point of view is the source IP address of the received packet, the receiving layer 2 address, and the received packet. Check whether the route defined as the port is different. In this case, since the two routes are the same, additional registration of the forwarding table is not made.

1C is as follows. FIG. 1C is a diagram for describing a case in which a packet is transmitted from a server to a terminal when the terminal is under the network B as shown in FIG. 1B.

When transmitting a packet from the server to the terminal, each router (Rs, Rn, Rb) is able to deliver the packet to the terminal in the optimal path through the managed forwarding table as described in Figure 1b. In FIG. 1C, the router Rs looks at the destination IP address of the received packet, determines that it matches the forwarding table entry {192.168.2.0/24 nexthop Rn}, and forwards the packet to the router Rn. The router Rn looks at the destination IP address of the received packet, determines that it matches the additionally registered forwarding table entry {192.168.2.1/32 nexthop Rb}, and then delivers the packet to the router Rb. Unlike the present invention, if not additionally registered, the router Rn incorrectly forwards the packet to the router Ra according to the LPM matching forwarding entry {192.168.2.0/nexthop Ra}. The router Rb looks at the destination IP address of the received packet, determines that it matches the additionally registered forwarding table entry {192.168.2.1/32 connected}, and then delivers the packet to the terminal in the local network.

1D is as follows. As shown in FIG. 1C, the UE under the network B moves and is under the network C, as illustrated in FIG. 1D.

When the mobile station transmits a packet destined for the server, the router Rc does not perform reverse path filtering, looks up the destination IP address of the received packet, looks up the forwarding table, and determines the next hop as the router Rn. , Forward the received packet. In addition, the router (Rc) is configured such that the root of the forwarding table entry (192.168.2.0/24 nexthop Rn) that matches the source IP address of the received packet from the LPM point of view is the source IP address of the received packet, the receiving layer 2 address, and the received packet. Detecting the route determined as the port is different, additional information corresponding to the source IP address of the received packet, the receiving layer 2 address, and the received port is further registered as an entry in the forwarding table. In FIG. 1D, an underlined entry among the entries of the forwarding table of the router Rc is an entry additionally registered by the above-described method.

In this process, the packet is delivered to the router Rn, and the router Rn performs the same process. That is, the router Rn looks at the destination IP address of the received packet without performing reverse path filtering, queries the forwarding table, determines the next hop as the router Rs, and then forwards the received packet. In addition, the router (Rn) is the root of the existing forwarding table entry (192.168.2.1/32 nexthop Rb) that matches the source IP address of the received packet in terms of LPM in the LPM perspective of the received packet. Detecting the difference between the source IP address, the receiving layer 2 address, and the route determined by the received port, the forwarding table entry (192.168.2.1/32 nexthop Rb) additionally registered in FIG. 1C is received. The information corresponding to the layer 2 address and the received port is alternately registered as an entry in the forwarding table. In FIG. 1D, an underlined entry among the entries of the forwarding table of the router Rn is an entry registered additionally by the above-described method.

In this process, the packet is delivered to the router Rs, and the router Rs performs the same process. That is, the router Rs does not perform reverse path filtering, looks up the destination IP address of the received packet, looks up the forwarding table, determines that the next hop is the local network, and then delivers the received packet to a server in the local network. In addition, the router (Rs) is configured such that the root of the forwarding table entry (192.168.2.0/24 nexthop Rn) matching the source IP address of the received packet from the LPM point of view is the source IP address of the received packet, the receiving layer 2 address, and the received packet. Check whether the route defined as the port is different. In this case, since the two routes are the same, additional registration of the forwarding table is not made.

Meanwhile, in FIG. 1B, the forwarding table entry 192.168.2.1/32 connected of the additionally registered router Rb is not used due to the mobile station moving to network C. As a result, the forwarding table entry 192.168.2.1/32 connected is deleted according to the aging procedure.

1E is as follows. FIG. 1E is a diagram for describing a case in which a packet is transmitted from a server to a terminal when the terminal is under network C, as shown in FIG. 1D.

In the same principle as described with reference to FIG. 1C, when a server transmits a packet to a terminal, each router (Rs, Rn, Rc) may deliver the packet to the terminal through the managed forwarding table as described in FIG. 1D in the optimal path. Will be.

Meanwhile, in FIG. 1E, the router Rs looks at the destination IP address of the received packet, determines that it matches the forwarding table entry {192.168.2.0/24 nexthop Rn}, and then delivers the packet to the router Rn. The router Rn looks at the destination IP address of the received packet, determines that it matches the additionally registered forwarding table entry {192.168.2.1/32 nexthop Rb}, and then delivers the packet to the router Rb. Unlike the present invention, if not additionally registered, the router Rn incorrectly forwards the packet to the router Ra according to the LPM matching forwarding entry {192.168.2.0/nexthop Ra}. The router Rb looks at the destination IP address of the received packet, determines that it matches the additionally registered forwarding table entry {192.168.2.1/32 connected}, and then delivers the packet to the terminal in the local network.

Meanwhile, the forwarding table entry 192.168.2.1/32 connected of the additionally registered router Rb in FIG. 1B is not used because the terminal moves to the network C, and as shown in FIG. 1E, the router Rb Deleted from the forwarding table.

FIG. 2 is a block diagram illustrating an apparatus having a three-layer switching function according to an embodiment of the present invention, wherein the ports 200 and 202, the processing unit 210, the hardware address table 220, and the forwarding table 230 are illustrated. It is made, including. Here, RIB and the like are not shown in FIG. 2 for the purpose of briefly explaining the principles of the invention. Unlike in FIG. 2, the processing unit 210 may be classified / implemented as a link protocol controller, a network protocol controller, or the like. The scope of the present invention is not limited to the embodiment shown in FIG. 2 according to the hardware implementation, and those skilled in the art can fully understand.

Packets are input / output to / from port 200 or 202.

The processor 210 determines the next hop and the output port 200 or 202 of the packet received from the corresponding port 200 or 202, and then performs hierarchical processing to transmit the determined next hop, and then determines the determined output port. Packet forwarding is performed by outputting to 200 or 202.

The hardware address table 220 is managed by the processing unit 210 in the form of an Address Resolution Protocol (ARP) cache and the like, and the hardware address table 220 includes a local network (local) connected to each of the ports 200 and 202. Layer 2 addresses (eg, MAC addresses), IP addresses, etc. of IP nodes (ie, IP hosts and routers) within the subnet are registered / stored.

The forwarding table 230 is managed by the processor 210 based on the RIB (not shown). The forwarding table 230 includes a network address (IP address) and next hop information (as shown in FIGS. 1A to 1E). For example, a route consisting of the next hop router IP address) or the like is registered / stored.

Referring to FIG. 2, the processor 210 includes a detector 212, a register 214, and a forwarding unit 216.

The detection unit 212 detects a reception route of the received packet. In one example, the receive route is determined based on the source IP address, source layer 2 address, and received port 200 or 202 of the received packet.

If the detected reception route and the route registered in the forwarding table 230 are different from each other, the registration unit 214 determines that the terminal that transmits the packet (hereinafter, referred to as a transmission terminal) has moved between networks and registers the reception route. In one example, the register 214 may include a forwarding table in which a source IP address, a source two-layer address, and a route defined by the received port 200 or 202 (that is, a detected route) are matched in terms of the source IP address and the LPM. If it is different from the root of the entry of 230, it is determined that the terminal has moved, and the information corresponding to the source IP address, the source layer 2 address, and the received port 200 or 202 of the received packet is forwarded as the entry table 230. To reflect). For example, if the detected reception route and the pre-registered route are different, the registration unit 214 further registers the reception route if the IP address of the pre-registered route matching the source IP address of the packet is different from the source IP address, and the source of the packet. If the IP address of the registered route matching the IP address is the same as the source IP address, the receiving route is registered instead of the matching registered route.

Meanwhile, the registration unit 214 deletes the registered reception route when the registered reception route is not used for a predetermined period for efficient operation of the forwarding table 230.

The forwarding unit 216 finds a forwarding table 230 entry matching the source IP address of the received packet, determines the next hop and the output port 200 or 202, and then processes the packet to be forwarded to the determined next hop. To output a packet through the determined output port 200 or 202. As a more specific example, when the matching forwarding table 230 entry is connected shown in FIGS. 1A to 1E, the forwarding unit 216 inquires the hardware address table 220 with the destination IP address of the packet, and then 2 A layer 2 header containing the layer address is added to the packet and provided to the corresponding output port 200 or 202. In another specific example, when the matching forwarding table 230 entry is 192.168.2.0/24 nexthop Rn illustrated in FIGS. 1A to 1E, the forwarding unit 216 may provide information about the next hop (the IP address of Rn). After obtaining the second layer address corresponding to the second layer address from the hardware address table 220, a second layer header including the second layer address is added to the packet and provided to the corresponding output port 200 or 202.

Preferably, the forwarding unit 236, even if the route determined by the source IP address, source layer 2 address, and the received port of the received packet is different from the route of the forwarding table 230 entry that matches the source IP address in terms of LPM The packet forwarding is performed without discarding the received packet.

On the other hand, unlike shown in Figure 2, the device with a three-layer switching function according to the present invention is provided with a separate storage unit (not shown), to manage the forwarding table 230 according to the existing operation method, but In the embodiment, an embodiment of managing the storage unit may be configured to additionally register and substitute a route (that is, an entry) obtained in the learning process of the present invention. According to this embodiment, the registration unit 214, when the movement between the network of the transmitting terminal is detected, stores the information corresponding to the source IP address, source layer 2 address, and the received port of the received packet in the storage unit, the forwarding unit In operation 216, the storage unit is inquired prior to the forwarding table 230, and packet forwarding is performed.

Such a method and apparatus of the present invention have been described with reference to the embodiments shown in the drawings for clarity, but these are merely exemplary, and various modifications and equivalent other embodiments are possible to those skilled in the art. Will understand. Therefore, the true technical protection scope of the present invention will be defined by the appended claims.

According to the present invention, a device having a three-layer switching function can efficiently support mobility of a terminal. Although the tunneling method is difficult to implement in hardware due to its complexity, the method according to the present invention is simple enough to be implemented in an ASIC, and excellent packet processing performance can be obtained. In addition, it is possible to lower the cost of the router supporting the mobility of the terminal, and the efficiency of the network can be improved because the path is optimized.

Claims (8)

A forwarding unit, among the entries of a forwarding table, to forward the received packet to a next hop corresponding to an entry matching a longest prefix with a destination IP address of the received packet; A detection unit that detects a reception route of the received packet, which is determined by a source IP address, a reception layer 2 address, and a reception port of the packet; And Detecting an entry in the forwarding table that matches the longest prefix with the source IP address of the packet, If the route defined as the detected entry is different from the detected receiving route and the IP address included in the detected entry is different from the source IP address of the packet, an entry corresponding to the receiving route is additionally registered in the forwarding table. , If the route determined by the detected entry is different from the detected receiving route, and the IP address included in the detected entry is the same as the source IP address of the packet, the detected entry is replaced with an entry corresponding to the receiving route. And a registration unit configured to alternately register in the forwarding table. delete delete delete The method of claim 1, wherein the registration unit And when the additionally registered or alternately registered entry is not used for a preset period of time, deleting the additionally or alternately registered entry. delete delete delete

Images