JP2002026970A - Path control system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a path control system that enhances the efficiency of a synthesis network. SOLUTION: The path control system used inside the synthesis network, is provided with a border path control node that is placed in an external element network and conducts packet transfer and path control between the synthesis network and an external network outside the synthesis network and with division networks that receives each assigned division extension address space formed by dividing an address space of an extension address acquired by the synthesis network and a non-acquisition extension address not acquired according to a prescribed division pattern and is obtained by logically dividing the synthesis network.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a routing control system, and is applicable to, for example, a routing control mechanism in a Core Protocol Network (CPN) proposed as an IP backbone network.

[0002]

2. Description of the Related Art The following literature 1 describes this CPN proposed by the ITU.

Reference (Reference 1): IEICE
IEICE Tech.

FIG. 2 shows a basic CPN model 10.

In FIG. 2, a CPN or platform 10 includes a core network 11, an access network 12 (12A,
12B), shell network 13, user (LAN) U1
U4.

(A-1) Core Network The core network 11 is a connectionless type wide area frame transfer network which is a central part of the platform 10. Connectionless Internet to connectionless ATM
The ATM network itself is not used for the core network because various problems due to signaling occur when overlaying the network directly.

[0007] The configuration of the core network 11 is a connectionless server (CL) that relays data using a single connectionless packet transfer protocol called a core protocol.
S).

[0008] The CLS is arranged in a region hierarchy, and accommodates each network element of the shell network 13 such as the edge node EN.

In the core network 11, packet transfer is performed using a core address. The core address is
1, 12, 13 and U1 to U4 are uniquely assigned in advance, and the conversion between the user address (IP address) and the core address is performed by the shell network 1
3 is performed.

(A-2) Access Network The access network 12 is composed of users (user LANs) U1 to U4.
Is connected to the edge node EN and accommodated in the core network 11.

The main components of the access network are an access node 15 and a user terminal 14.

(A-3) Shell Network The shell network 13 surrounds the core network 11 so as to surround the core network 11.
3 and the access network 12 and the core network 1
1 is performed.

The main components of the shell network 13 are an edge node (EN), a default forwarder (DF), and a route server (RS).

EN is a subscriber accommodating node, and its main functions are a frame format conversion process as a transfer process and an address format conversion process as a route control process.

In the address format conversion process, EN
Receives the user packet, derives the core address of the destination EN that is the destination on the core network 11 with reference to the IP address.

Specifically, the EN resolves the core address of the destination EN with reference to the destination Internet address described in the header of the Internet packet.

When the core address is derived, the corresponding EN
Performs a frame format conversion process, encapsulates the user packet in a core network frame, and transfers it to the core network 11.

DF is a relay node of the default route. That is, when the originating EN cannot derive the core address of the receiving EN, the EN transfers the core network frame to the DF.

The DF that has received the core network frame derives the core address of the destination EN by referring to the destination IP address. Thereafter, the DF notifies the originating EN of the core address of the receiving EN as route information. The notified calling side EN can resolve the core address of the called side EN.

The RS is a routing management device that collectively manages routing information in the shell network 13. That is, R
S manages routing information (correspondence between the user's IP address and core address) that the EN and DF refer to when transferring a packet, and changes the routing information without contradiction with a change in the accommodation destination EN of the user. The routine information is set in EN or DF.

Next, an address aggregation mechanism of the CPN 10 and a packet transfer device / routing management device based thereon will be described. These are both important technologies for the path control mechanism of the CPN 10.

(B-1) Address Aggregation Mechanism The address aggregation mechanism is important from the viewpoint of centralizing the routing information (reducing the routing table).

If the routing information is consolidated and reduced,
This is because it is expected that the utilization efficiency of the route control capability of the router is enhanced and the transfer capability of the router is substantially improved.

The address aggregation method of the CPN 10 divides the CPN 10 logically into a plurality of SubASs as shown in FIG. 8 (since the CPN is considered to be one AS, the divided unit is called a SubAS). ), IP address /
Sub with users with the same prefix upper part
They are collectively handled as a group called AS.

That is, for example, when the CPN becomes large in scale and acquires a plurality of IP address blocks, the CPN is logically divided into address block units.
The division unit is set to SubAS.
As a result, it is possible to reduce the route information in the CPN.
The reason is described below.

One default forwarder (SubAS-DF) is installed in each SubAS. SubAS-D
F has a function of transferring a user packet addressed to its own SubAS to an appropriate destination EN. That is, SubAS-
DF is the IP address of all users in SubAS /
It has a set of information composed of a prefix and the core address of the EN that accommodates the user. This S
By using the ubAS-DF, the route information to be held by the EN can be aggregated. That is, in each EN, SubAS
IP address / prefix representing SubA and its SubA
Only one set of information consisting of the core address of the SubAS-DF belonging to S and one
Reachability to all subnets in S can be secured. Then, even if the connection configuration such as the subnet is changed in the SubAS, it is only necessary to change the path information of the SubAS-DF, and the path information of the EN does not need to be changed.

(B-2) Packet Transfer Apparatus / Routing Management Apparatus The packet transfer apparatus / routing management apparatus has a problem in terms of the number of hops of packet transfer (the smaller the number of hops, the smaller the delay) and the complexity of managing routing information. Important from.

FIG. 3 shows a packet transfer device / routing management device of the CPN, incorporating the concept of SubAS described above.

(B-2-1) Packet Transfer Apparatus (B-2-2) Edge Node EN The EN determines the SubAS corresponding to the packet received from the user by looking at the IP header, and determines the SubAS.
Transfer to DF. If it cannot be determined, it is transferred to the default route AS-DF. Also, the output packet corresponding to the received packet from the core network is determined by referring to the IP header, and is transferred to the access network side.

(B-2-3) SubAS-DF The SubAS-DF looks up the destination IP address of the received packet, finds the corresponding destination, and transfers it to the destination.
The destination to be determined is either EN or AS-DF. That is, the packet addressed to the external AS is
-The packet addressed to the inside of the CPN to the DF will be transferred to the destination EN.

(B-2-4) AS-DF The AS-DF uniformly transfers the packet received from the SubAS-DF to the border router. The packet received from the border router checks which IP address
It is determined whether it belongs to the bAS and transmitted to the SubAS-DF.

(B-2-5) Border Router The packet received from the AS-DF is transferred to an external AS (external AS).
(AS border router). The packet received from the external AS is transmitted to the AS-DF.

(B-3) Routing Management Device (B-3-1) SubAS-RS One SubAS-RS is installed in the SubAS.

Each SubAS-RS collectively manages routing information (correspondence between the user's IP address / prefix and the user's accommodation EN core address) that the SubAS-DF should have (for example, routing according to the movement of the user, etc.). Information is changed without contradiction), SubAS-DF
Set to

(B-3-2) AS-RS One AS-RS is installed in the CPN. The routing information (correspondence between the IP address / prefix indicating the SubAS and the core address of the SubAS-DF, and the core address of the AS-DF indicating the default route) that the EN should have, and the routing information (the default route has the default route) that the AS-DF should have The core address of the border router indicated, and the IP address / prefix indicating SubAS and Su
bAS-DF core address). Then, the routing table is set in EN and AS-DF.

Further, a packet transfer route using the above-described device will be described below.

(B-4-1) User 1 (UR1: this corresponds to, for example, U1) to User 2 (UR2:
This corresponds to, for example, the packet transfer route to U2). In FIG. 3, EN1 which has received a packet from UR1
Is the destination user (U
R) to which the SubAS belongs, and determine the SubAS
To the SubAS-DF (SubAS-DF2).

SubAS-D receiving this packet
F2 is based on the destination IP address and the EN to which the destination UR belongs.
And sends it to its EN (EN2).

The EN2 determines a UR connection link from the destination IP address and transmits a packet to the link.

Each of UR1 and UR2 is a user LAN provided with a router. This is the same in the following.

(B-4-2) The packet transfer route EN1 from UR1 to UR3 in the external AS tries to determine the SubAS to which the called user belongs using the called IP address as a key, but cannot find it. Transfer to the default route (AS-DF).

The AS-DF that has received this packet,
Transfer the received packet to the border router.

Then, the border router transfers the received packet to the external AS.

(B-4-3) Packet transfer route from UR3 to UR2 The border router transmits the packet received from the external AS to the AS.
-Send to DF.

The AS-DF determines the SubAS to which the destination UR belongs by using the destination IP address as a key, and the SubAS-DF (SubAS-DF2) of the SubAS.
Send to

The SubAS-DF2 uses the destination IP address as a key to determine and transmit the EN (EN2) to which the destination UR belongs.

In EN2, a connection link of the UR is determined from the destination IP address, and a packet is transmitted to the link.

Since the above-described CPN 10 employs an overlay type mechanism, there is an advantage that the route control mechanism can be designed independently of the transfer mechanism.

[0049]

By the way, the above CPN
The ten address aggregation mechanisms and the packet transfer device / routing management device have the following problems.

(1) Address Aggregation Mechanism According to Document 1, when an address acquired by one AS called CPN 10 is a plurality of address blocks, the SubAS is divided for each address block.

However, for example, when the address space of the address block 0 handled by SubAS # 0 is significantly different from that of the address block 1 handled by SubAS1, the traffic volume handled by SubAS # 0 and SubAS # 1 is naturally greatly different. Therefore, inevitably, Su
A difference appears in the processing amount between bAS-DF1 and SubAS-DF2, resulting in a difference in packet transfer time. In this most basic service, it is inappropriate to make a difference between SubASs in the same AS.

(2) Packet Transfer Apparatus Further, Document 1 has a problem that the number of hops of a packet arriving from an external AS is large.

For example, the packet transfer route in the shell network 13 from UR3 to UR2 in FIG.
S-DF → SubAS-DF2 → EN2, which is 3 hops.

On the other hand, UR1 to UR2 between internal users
In the packet transfer route to, EN1 → SubAS-D
F2 → EN2, two hops.

That is, a packet coming from the external AS is
A redundant route is transferred for one hop.

(3) Routing Management Apparatus Further, in Reference 1, there are two types of routing management apparatuses, SubAS-RS and AS-RS, depending on which routing information is managed. However, different types of routing management devices require different logics to be set for each type, leading to division loss and complicated management, resulting in higher costs in forming a network.

(4) Regarding Packet Transfer Apparatus / Routing Management Apparatus Furthermore, Document 1 does not show how to deal with a user having a carry-in IP address.

The carry-in IP address means that, for example, a user assigns an IP address to an institution IR (internet registry).
y) or a case where an IP address obtained directly from a local IR is brought directly into the CPN for connection, or a certain provider (Internet connection operator: IS
P) assigns the IP address obtained from the local IR to the corporate user who has subscribed to the provider, and then the corporate user subscribed to another provider (this is the CPN 10 provider) while holding the IP address. This is a connection mode that occurs in such a case.

When a carry-in IP address is generated, CPN
10 needs to correspond to an IP address that is not included in the originally assumed address block (an unplanned IP address in Document 1), but Document 1 does not describe a corresponding method.

[0060]

In order to solve the above-mentioned problems, the present invention is based on a predetermined core protocol.
A core element network that performs packet relay processing;
Or an access element network accommodating a plurality of user networks, and the core element network and the user, which are arranged between the access element network and the core element network, using a source edge relay node and a destination edge relay node. Connect to the network,
A routing control system used inside a combined network comprising: a shell element network for performing the packet route control processing; (1) a routing control system arranged in the shell element network and an external network outside the synthesis network; A boundary route control node that performs transfer and route control of the packet with the combined network;
(2) With respect to the extended network in which both the composite network and the external network are constituent elements, the address space of the acquired extended address acquired by the composite network among the extended addresses used to transfer a predetermined extended packet; The address space of the unacquired expanded address that has not been obtained is allocated to each divided expanded address space formed by dividing the expanded address into a sequential number according to a predetermined division pattern, and the combined network is logically allocated. (3) each divided network includes at least a divided path information server that shares its own divided network, the edge relay node, and a default transfer device, and the external network When exchanging packets with Tsu di relay node directly, characterized by transmitting and receiving packets to the boundary routing node.

[0061]

BEST MODE FOR CARRYING OUT THE INVENTION (A) Embodiment An embodiment of the present invention will be described below by taking, as an example, a case where the routing system of the present invention is applied to a core protocol network CPN.

The first to fifth embodiments are different from the reference 1 in the portion of the CPN shell network, and the features for solving the above-described problems exist in the shell network.

In the shell network of each embodiment, the entire IP address space is divided without discriminating between the SubAS and the external AS,
It is characterized in that each division is regarded as SubAS.

On the other hand, in Reference 1, AS-RS and Sub
The root server having the AS-RS two-layer structure is a single layer of only the SubAS-RS, and the default forwarder is also a single layer of the SubAS-DF from the two-layer structure of AS-DF and SubAS-DF. Hierarchy and each SubA
It is characterized in that the S-DF directly exchanges packets with the border router BR.

(A-1) Configuration of the First Embodiment The external structure of the core protocol network CPN 20 of this embodiment is the same as that of the CPN 10 shown in FIG.

That is, as shown in FIG. 2, the CPN 20 of the present embodiment comprises a core network 11 and an access network 12 (12
A, 12B), and users (LANs) U1 to U4.

However, the shell network of the CPN 20 according to the present embodiment is distinguished from the shell network 13 by the reference numeral 21 because its internal structure is different.

The components of the shell network 21 are roughly the same as in Reference 1 when roughly grasped, and include an edge node EN, a route server RS, and a default forwarder DF. The RS in Reference 1 includes AS-RS and SubAS- While there are two types of RS, the RS of the present embodiment is integrated into a SubAS-RS.

As for the DF, the DF in Document 1 contains A
In contrast to the two types of S-DF and SubAS-DF, the DF according to the present embodiment is integrated into a SubAS-DF.

The functions of each of the SubAS-RS and SubAS-DF are different between this embodiment and Reference 1.

In Reference 1, SubA is used for each address block.
S is set, and one SubAS-RS corresponds to one address block.
RS also corresponds to a plurality of address blocks.

Here, the address block is usually defined by an IP address assigning agency IR (internetregistry).
Refers to a group of serial numbers assigned to a provider.

The address block includes CIDR (classl
ESS inter-domain routing) type and class type. In the case of the class type, at least 254 IP addresses are allocated, but in the case of the CIDR type, only a few IP addresses are allocated, and the address block is very small. Small.

At present, it is generally difficult to obtain a class type address block due to the possibility of exhaustion of IP addresses. AS (autonomous syst
em) is a network that is a constituent unit of the Internet and is an autonomous network operated by one organization. This organization is typically a provider. The AS manages a relatively large address block, and the user always becomes a subscriber of some AS.

The CPN 20 of this embodiment also corresponds to one AS, but the CPN 20 is an extremely large-scale AS accommodating, for example, 10 million subscribers. , U in FIG.
1 to U4.

Therefore, the CPN 20 of this embodiment is
It is expected that address blocks obtained from the R institution can also be obtained in a relatively high level class type.

FIG. 1 shows the configuration of the address aggregation mechanism of this embodiment. FIG. 1 shows each SubA in the shell network 21 as it is.
It shows the IP (IP address) space of the S-RS.

Of these, SubAS $ 0 and SubAS $ 1 are
2 shows two SubASs in the AS 21. The sizes of these IP address spaces are almost equal.

However, this equivalence need not be so strict. The purpose of equalization is to equalize the packet transfer time in each SubAS, but a slight difference in the size of the allocated address space is likely to cause a dramatic change in the packet transfer time in each SubAS. Is considered to be low.

Another SubAS, that is, SubAS $ 2 to $ 6
Is an IP address space other than the IP address space assigned to the AS 21.

As shown in FIG. 1, in the CPN 20, all the IP address spaces are configured as SubAS.

Specifically, the CPN acquisition IP address space and the other (non-acquisition) IP address space (external AS)
In the above, SubAS conversion is performed so that the size of the IP address space for each SubAS is the same.

That is, the sizes of the IP spaces of SubASs # 0 and # 1 are equal, and the sizes of the IP spaces of SubASs # 2 to # 6 are equal.

The size of the IP space of one of SubAS # 2 to # 6 is usually much larger than the size of the IP space of SubAS # 0 or # 1.

Specifically, the CPN acquisition IP address space is 100.0.0.0 to 100.255.255.2.
55 and 200.0.0.0 to 209.255.255.
255, but this is not converted to SubAS as it is.

200.0.0.0 to 209.255.2
55.255 into two additional address blocks, 20
0.0.0.0 to 204.255.255.255,
205.0.0.0-209.255.255.255
Divided into 100.0.0.0 to 100.255.25
5.255 and 200.0.0.0 to 204.255.2
55.255 and one SubAS (SubAS♯0)
To form 205.0.0.0 to 209.255.25.
One SubAS (SubAS # 1) is formed at 5.255.

As a result, the size of both IP spaces becomes the same.

Also, the IP space of the external AS, that is, the non-acquired IP space (0.0.0.0 to 255.255.25)
5.255) is divided into approximately five equal parts and divided into five SubASs.

After all, the CPN 20 has two SubASs (internal SubASs) corresponding to the acquired IP space,
Since there are five SubASs (external SubASs) corresponding to the space, there are a total of seven SubASs.

Each of these seven SubASs is provided with one SubAS-RS and one SubAS-DF, and one or a plurality of ENs.

SubAS-RS sharing SubAS−0 is designated as Su
bAS-RS0 and SubAS-DF as SubA
S-DF0.

Similarly, in the following, the SubAS # 1
The AS-RS is referred to as SubAS-RS1, and the
-DF is SubAS-DF1 and SubASＡ2 is shared
SubAS-RS is referred to as SubAS-RS2, and
The AS-DF is referred to as SubAS-DF2, and the SubAS-RS sharing SubASSub3 is referred to as SubAS-RS3.
SubAS-DF is referred to as SubAS-DF3, SubAS-RS for sharing SubAS # 4 is referred to as SubAS-RS4, and SubAS-DF is referred to as SubAS-DF4.
SubAS-RS is assigned to SubAS-RS5, and SubAS-DF is assigned to SubAS-DF5.
The SubAS-RS sharing S♯6 is referred to as SubAS-RS6, and the SubAS-DF is referred to as SubAS-DF6.

As shown in FIGS. 8 and 9, the packet transfer device is referred to as EN, the SubAS-DF, the border router, and the routing management device as SubAS-RS.
N.

The EN belonging to the internal SubAS (for example, E
N1) looks up the destination IP address of the packet received from the user, finds the corresponding SubAS, and
Transfer to the bAS SubAS-DF.

Here, since all the IP spaces are converted to SubAS, the corresponding SubAS can always be determined. The packet received from the core network is transmitted to the IP
The output link is determined by looking at the header and transferred to the access side.

Further, SubA belonging to the internal SubAS
The S-DF (for example, SubAS-DF1) finds the destination by looking at the destination IP address of the received packet, and transfers the destination to the destination. The destination to be determined is any E
N or border router BR.

That is, a packet addressed to the external AS is transferred to the border router, and a packet addressed to the inside of the CPN is transferred to the connection destination EN.

The number of border routers BR provided in the shell network 21 of the CPN 20 is, in this embodiment,
It is assumed that there is one.

The border router BR has a SubAS-
While the packet received from the DF is transferred to the external AS (border router of the external AS), the packet received from the external AS refers to the destination IP address, determines the corresponding destination SubAS, and determines the SubAS-DF. Send to

The internal SubAS or the external Su
SubAS-RS belonging to bAS (for example, SubAS-R
S1) is a routing management device of the CPN 20.
The routing information (correspondence between the IP address / prefix of the user and the EN core address of the accommodating destination of the user, and the core address of the border router as a default route) which the SubAS-DF should have, and the routing information (the SubAS which the EN should have) (Correspondence between the IP address / prefix and the core address of the SubAS-DF) and set it in the SubAS-DF and EN.

The operation of the first embodiment having the above configuration will be described below.

(A-2) Operation of First Embodiment The operation of each device when a packet is transferred by the system 20 will be described with reference to FIGS.

FIG. 8 shows a packet transfer route when the carry-in IP address user UR2 has a normal connection, and FIG. 9 shows a packet transfer route when the carry-in IP address user UR2 has an illegal connection.

First, the meaning of normal / illegal connection will be described. The normal connection is when the user (UR2) of the brought-in IP address is connected to the EN (EN2) that should be accommodated.

The meaning of “should be accommodated” means “EN
2 is under the control of a routing management device (SubAS-RS2) which is the administrator of the route information (EN is to belong to one SubAS), and SubAS
-RS2 manages the IP address user of SubAS # 2. Since the IP address of UR2 belongs to SubAS # 2, UR2 should be accommodated in EN2 controlled by SubAS-RS2. "

On the other hand, the illegal connection means the carry-in IP
This is when the user of the address (UR2) is connected to the EN that should not be accommodated (EN1 in the example of FIG. 9). The meaning of "should not be contained"
"EN1 is under the control of the routing management device (SubAS-RS1) which is the administrator of the route information, and
The AS-RS1 manages an IP address user of SubAS @ 0. UR2 IP address is SubAS @ 2
UR2 is Despite SubAS-
It is accommodated in EN1 controlled by RS1. "

As shown in FIG. 9, the illegal connection state is set to UR1 already connected to CPN20.
Occurs when 2 is connected.

More specifically, such illegal connection may occur when a router in UR2, which is a user LAN, connects a packet transmitted from UR2 to UR1.

Since UR1 is a user having an IP address belonging to SubAS @ 0, EN1 is naturally a SubAS-
Although it is under the control of RS1, if UR2 is connected to UR1 (UR2 is a carry-in IP address and can be freely connected), UR2 is UR to EN1 which is not the original connection destination.
They will be connected through one.

Next, a packet transfer operation of a carry-in IP address (normal connection) will be described with reference to FIG.

(A-2-1) Packet Transfer Operation from UR1 to UR2 EN1 finds the SubAS to which the destination UR belongs on the core network routing table, using the destination IP address as a key, and the SubAS of the SubAS. DF (Sub
AS-DF2).

SubAS-DF2 which received the packet
Calculates the EN to which the destination UR belongs from the destination IP address and transmits it to the EN (EN2).

Then, EN2 receives U from the destination IP address.
The connection link of R is determined, and a packet is transmitted to the link.

The following is a case where the carry-in IP address user UR2 transmits a packet to the external AS in the same normal connection state.

(A-2-2) UR1 to UR3 (external A
Packet transfer operation to S) EN1 finds the SubAS to which the destination UR belongs on the core network routing table using the destination IP address as a key, and the SubAS-DF (SubA) of the SubAS
The packet is transmitted to S-DF2). In this case, all I
Since the P space is made into SubAS, a UR belonging to the external AS also belongs to some SubAS.

The SubAS-DF2 determines the destination of the packet using the destination IP address as a key. In the case of an external AS, since the destination IP address itself (the IP address of the destination user) is not registered in the routing information, the core address of the border router BR is extracted as a default route, and the packet is transmitted.

The border router BR transfers the received packet to the external AS.

The UR3 receiving this packet is not actually accommodated in the CPN 20, but is
Since the entire IP space is SubAS, the reception of the packet by the UR3 is depicted in FIG. 8 as an event occurring on the address space of SubAS # 2.

The last case is that the UR2 receives a packet from the external SubAS in the same normal connection state.

(A-2-3) Packet Transfer Operation from UR3 to UR2 The border router BR looks up the destination IP address of the packet received from the external AS, determines the SubAS to which it belongs,
The SubAS-DF of the SubAS (SubAS-DF
2) Send the packet to.

The processing of SubAS-DF2 and EN2 is as follows.
This is the same as (A-2-1).

On the other hand, the operation in the case of illegal connection is as follows.

The packet transfer operation in the case of a carry-in IP address (illegal connection) will be described with reference to FIG.

The first is a case where the UR2 which has been illegally connected receives a packet from the external AS.

(A-3-1) Packet Transfer Operation from UR3 to UR2 The border router BR looks up the destination IP address of the packet received from the external AS and determines the SubAS to which it belongs.
The SubAS-DF of the SubAS (SubAS-DF
2) Send the packet to.

The SubAS-DF2 receives the DF (Su) of the SubAS containing the destination UR from the destination IP address.
bAS-DF1) and its SubAS-DF1
Send to

Then, the SubAS-DF1 receives the destination IP address.
EN (EN1) containing destination UR from address
And sends it to EN1.

The EN1 determines a UR connection link from the destination IP address and transmits a packet to the link.

In this illegal connection state, the information of the set of SubAS-DF1 and EN1 is represented by the OSPF (Open Shortest PathFirst) protocol, for example.
It will be rewritten corresponding to the illegal connection of the UR2.

In the above description, cut-through and cache are not mentioned. However, also in the present embodiment, after receiving the route information from the SubAS-DF to which the destination EN belongs, It is also possible to perform cut-through using the path information and directly transfer the packet to the destination EN.

Of course, the route information used once can be held as a cache.

In the present embodiment, focusing only on the inside of the AS 20, as described above, two types of RS
And DF are unified into SubAS-RS and SubAS-DF, but when AS20 of the present embodiment is viewed from an external AS, as SubAS-RS behaves like a conventional RS, it is as if in AS20 It appears that there are as many independent ASs as there are SubASs.

From this point, in the present embodiment, the AS concept of the AS 20 itself can be regarded as completely different from the conventional AS concept.

(A-3) Effects of the First Embodiment As described above, according to the present embodiment, the following effects can be obtained.

(1) Effect of New Address Aggregation Mechanism By dividing / combining acquired IP address blocks and allocating the same size IP address space to SubAS, the traffic volume handled by each SubAS becomes almost the same. As a result, the packet transfer time in all the SubASs can be made uniform as compared with the conventional technology.

(2) Effect of New Address Aggregation Mechanism and New Packet Transfer Device SubAS of all IP spaces (SubAS is also performed for IP space of external AS), and border as new packet transfer device By installing the router, SubAS-DF, and EN, and not installing the AS-DF that supports the entire AS, the packet transfer route from the external AS (for example, UR3) to UR2 changes from the border router to the Su.
bAS-DF2 → EN2, so that the number of hops can be reduced and the packet transfer time can be reduced as compared with the conventional technology.

(3) Effect of New Address Aggregation Mechanism and New Routing Management Apparatus All IP spaces are converted to SubAS (SubAS is also performed for IP space of external AS), and SubAS-RS is used as a routing management apparatus. Is installed (no AS-RS is installed), there is no redundancy in RS types compared to the conventional technology, and there is no need to create two types of RSs, and as a result, a reduction in network cost can be expected. .

(4) It is possible to cope with a user having a carry-in IP address which cannot be handled by the conventional technology.

(B) Second Embodiment In this embodiment, it is not always possible to make the traffic volume the same simply by making the size of the IP space allocated to each SubAS the same, and to make the packet transfer time the same. This was done with a view to not being able to guarantee that

(B-1) Configuration and Operation of the Second Embodiment The main difference in the configuration between the first embodiment and this embodiment is that the address aggregation mechanism shown in FIG. There is no substantial difference in operation.

In the following, only points different from the configuration and operation of the first embodiment will be described.

In the present embodiment, all the IP address spaces are converted to SubAS, as in the first embodiment, but the method of allocating address blocks to each SubAS is different.

In the first embodiment, the address blocks are allocated so that the size of the IP address space of each SubAS is the same. In the present embodiment, the address blocks are allocated so that the traffic volume of each SubAS is the same. .

That is, if the traffic volume of the IP address block can be known by some measurement or prediction, the address block is divided so that the traffic volume becomes the same, and each is divided and assigned to SubAS.

As shown in FIG. 10, the CPN acquisition IP address space is 100.0.0.0 to 100.255.2.
55.255 and 200.0.0.0 to 209.255.
255.255, which is not converted to SubAS as it is. By some means, for example, 100.0.0.
0-100.255.255.255 and 200.0.
A user's traffic volume in the range of 0.0 to 208.255.255.255, and 209.0.0.0 to 209.2
If it is determined that the traffic volume of the user in the range of 55.255.255 is substantially the same, each is converted to SubAS.

An IP address other than the CPN acquisition IP address
The space is also divided into, for example, three SubASs in which the traffic volume of the user to which the user belongs is the same.

The structure of the packet transfer device / routing management device is the same as that of the first embodiment.

(B-2) Effect of Second Embodiment When the size of the IP address space of each SubAS is the same as in the first embodiment, the number of users accommodated in each SubAS is almost the same. It can be expected that the size of the routing table will be the same. That is, the search time of the routing table is considered to be the same.

The equalization of the routing table search time, which has a large effect on the packet transfer delay, is performed by using the SubAS.
This is desirable for equalizing the delay time between them.

However, even if the size of the IP address space of each SubAS is the same, if the difference in the traffic amount (packet amount) handled by each SubAS is greatly different, the difference in the queuing delay becomes large and the delay time between the SubASs becomes longer. It is possible that a difference occurs.

In such a case, as shown in the present embodiment, the queuing delay can be expected to be uniform by implementing the SubAS so that the traffic volume becomes almost the same.
As a result, the delay time between SubASs can be made uniform.

In other words, when there is a difference in the traffic volume handled by each SubAS, it is expected that the delay time between SubASs is made uniform by configuring as shown in the present embodiment.

(C) Third Embodiment This embodiment focuses on the point that the delay time of the SubAS allocated to the CPN acquisition IP address space is different from the delay time of the SubAS allocated to the IP address other than the CPN acquisition. It is a thing.

(C-1) Configuration and Operation of Third Embodiment A major difference in the configuration between the first embodiment and the present embodiment is that an address aggregation mechanism shown in FIG.
SubAS setting method), there is no substantial difference in operation.

In the following, only points different from the configuration and operation of the first embodiment will be described.

As shown in FIG. 7, in this embodiment, all the IP address spaces are converted into SubAS, as in the first and second embodiments, but the method of allocating address blocks to each SubAS is different. .

In the first embodiment, the address blocks are allocated so that the size of the IP address space of each SubAS is the same. In the second embodiment, the address blocks are allocated so that the traffic volume of each SubAS is the same. However, in each of the embodiments, the CPN acquisition IP address space and the IP address space other than the CPN acquisition are individually converted to SubAS.

However, in the present embodiment, there is no distinction between the CPN acquisition IP address space and the IP address space other than the CPN acquisition, and SubAS is performed across both IP address spaces.

As shown in FIG. 7, for example, the CPN acquisition IP
The address space is divided into three, and the IP address space other than the acquisition of the CPN is also divided into three. Then, the three divided IP address blocks are re-blocked as one SubAS.

In FIG. 7, the CPN acquisition IP address space,
In the IP address space other than the CPN acquisition, the traffic is divided into address blocks so that the amount of traffic becomes equal, and these address blocks are combined across the CPN acquisition IP address space and the IP address space other than the CPN acquisition to form a SubAS. However, as in the first embodiment, the IP address space may be divided so as to have the same size.

The structure of the packet transfer device / routing management device is the same as in the first embodiment.

(C-2) Description of Effect As in the first and second embodiments, when the CPN acquisition IP address space and the IP address space other than the CPN acquisition are individually allocated to the SubAS, each IP address is assigned. In SubAS in the space, uniform delay time can be expected. However, the SubAS assigned to the CPN acquisition IP address space and the SubAS assigned to the IP address space other than the CPN acquisition usually have a large difference in the size of the IP space as described above. The time is expected to be different. This is because the larger the size of the IP space, the longer the time required for searching the routing table and the like tend to be, and the delay time is usually longer.

Therefore, in the present embodiment, the CPN acquisition I
If the SubAS is implemented across the P address space and the IP address space other than the CPN acquisition, a packet transfer service with a uniform delay time can be provided regardless of acquisition or non-acquisition.

On the contrary, it is conceivable to positively utilize the difference in delay time between acquisition and non-acquisition.

For example, packet transfer is performed with a short delay time to a user who has been assigned an IP address in the acquired IP address space, and is performed with a relatively long delay time to a user having an IP address in the unacquired IP address space. By performing packet transfer, it is also possible to guide users to increase the number of users to which IP addresses in the acquired IP address space are assigned.

(D) Fourth Embodiment In this embodiment, the SubAS-DF belonging to the carry-in IP address user UR in the illegally connected state directly knows the accommodation destination EN of the UR and transfers the packet there. By doing so, the number of hops in the illegal connection state is reduced.

(D-1) Configuration and Operation of the Fourth Embodiment A major difference in operation between the first embodiment and the present embodiment is that the packet transfer in the illegally connected state shown in FIG. Related to each part.

Hereinafter, only points different from the configuration and operation of the first embodiment will be described.

Carry-in I in the first, second, and third embodiments
In the case of the illegal connection of the P address, as shown in FIG. 9, the DF (SubAS-D) of the SubAS to which the UR2 belongs.
F2) does not know the accommodation destination EN (EN1) of UR2, and the DF of the SubAS (SubAS-
DF1). Then, the SubAS-DF1 that has received the packet transmits the UR
2 knows the destination EN (EN1) and forwards the packet to EN1.

However, in the present embodiment, as shown in FIG. 5, the SubAS-DF2 belonging to UR2 (SubAS-DF2 in the example of FIG. 5) directly connects the accommodation destination EN (EN1 in the example of FIG. 5) of UR2. Know and try to forward packets there.

The address aggregation mechanism may use any of the first to third embodiments.

The operation is different only in the case of illegal connection, and this will be described with reference to FIG. In FIG. 5, since UR2 is in an illegal connection, a packet transfer operation from UR3 to UR2 is described below.

(D-1-1) Packet Transfer Operation from UR3 to UR2 The border router BR looks up the destination IP address of the packet received from the external AS, determines the SubAS to which it belongs,
The SubAS-DF of the SubAS (SubAS-DF
2) Send the packet to.

The SubAS-DF2 calculates the destination (EN1) to which the destination UR belongs from the destination IP address,
Perform packet transfer to N1.

The EN1 determines a UR connection link from the destination IP address and transmits a packet to the link.

(D-2) Effects of the Fourth Embodiment According to the present embodiment, the DF of the SubAS to which the brought-in IP address user UR of the illegal connection belongs belongs to the UR2.
Directly knows the accommodation destination EN of the SubAS-DF
Can transfer packets to the accommodation destination EN, so that the number of hops for packet transfer to the illegally connected user can be smaller than in the first, second, and third embodiments.

(E) Fifth Embodiment In this embodiment, the border router BR directly knows the accommodation destination EN of the brought-in IP address user of the illegal connection, and transfers the packet to the EN, so that the border router BR sends the packet to the illegal connection user. It is characterized in that the number of hops in packet transfer is reduced.

(E-1) Configuration and Operation of Fifth Embodiment In the configuration of the first embodiment and the present embodiment, the main difference in operation is that the packet transfer in the illegal connection state shown in FIG. Related to each part.

In the following, only points different from the configuration and operation of the first embodiment will be described.

In the case of the illegal connection of the carry-in IP address in the first, second, third, and fourth embodiments, the configuration of the packet transfer device / routing management device (the structure in the case of the carry-in IP address of the illegal connection) is as shown in FIG. , And FIG.

However, in the present embodiment, as shown in FIG. 6, the border router BR connects the UR2 to the accommodation destination EN (EN
Know 1) directly and forward the packet there.

As the address aggregation mechanism, any of the address aggregation mechanisms in the first to third embodiments may be used.

A substantial difference in operation between the first, second, third, and fourth embodiments and this embodiment is only in the case of illegal connection. Therefore, the operation will be described with reference to FIG.

Since the UR2 is in an illegal connection in FIG. 6, the packet transfer operation from the UR3 to the UR2 is described below.

(E-1-1) Packet Transfer Operation from EN3 to UR2 The border router BR looks at the destination IP address of the packet received from the external AS, and receives the destination EN (EN
1) is determined, and the packet is transmitted there.

Upon receiving this packet, EN1 determines a UR connection link from the destination IP address and transmits the packet to that link.

(E-2) Effects of the Fifth Embodiment According to the present embodiment, the border router directly knows the accommodation destination EN of the brought-in IP address user UR of the illegal connection, and forwards the packet to the EN. Therefore, the number of hops for packet transfer to the illegally connected user can be reduced as compared with the first, second, third and fourth embodiments.

(F) Other Embodiments In the description of each of the above embodiments, the term EN (edge node) is used for convenience for a device accommodating a subscriber. Nodes are usually multifunctional and sophisticated, and are often devices with relatively large hardware.

However, the function required for an EN accommodating a subscriber is to accommodate only a user having an IP address belonging to the same SubAS, and not to accommodate users of other SubASs. It is not necessary, nor does it need to be so sophisticated.

Therefore, the term “edge device (ED)” may be used instead of “edge node (EN)”. The hardware of the ED (or EN) may be, for example, a single board for accommodating subscribers.

The combination of the address aggregation mechanism and the packet transfer device / routing management device shown in each embodiment may be changed.

In each of the above embodiments, the non-acquired IP spaces are almost equally divided and have almost the same size. However, these sizes may be greatly different. Also,
In the first embodiment, the non-acquired IP space is divided into five, but these may be divided into more or less.

In the first embodiment, the seven Sus
In each bAS, one SubAS-RS and one SubAS-DF are arranged, and one or a plurality of ENs are arranged. However, an EN dedicated to an external SubAS may not be arranged.

Although the EN and UR corresponding to the external SubAS do not originally exist in the CPN 20, the external SubA
Since the distinction between S and the internal SubAS is logical, for example, when the UR of the internal SubAS # 1 uses a carry-in IP address, the EN in the internal SubAS # 1
This is because it can function as an EN in the external SubAS to which the P address belongs.

As described above, the CPN 20 of each of the embodiments is used.
Is primarily intended for very large AS applications, for example, accommodating as many as 10 million subscribers,
Of course, the CPN 20 may be applied to a smaller AS.

That is, the present invention provides a core element network for performing packet relay processing based on a predetermined core protocol, an access element network accommodating one or a plurality of user networks, an access element network and the core element network. An outer shell element network that is arranged in the middle of a network, connects the core element network and the user network using a source edge relay node and a destination edge relay node, and performs a path control process of the packet. The present invention can be widely applied to a route control system used inside a synthetic network.

[0197]

As described above, according to the present invention, it is possible to provide a route control system capable of improving the reliability and performance of a combined network.

[Brief description of the drawings]

FIG. 1 is a schematic diagram illustrating an address aggregation mechanism according to an embodiment.

FIG. 2 is a schematic diagram showing a CPN according to the related art and an embodiment.

FIG. 3 is a diagram illustrating the operation of a conventional CPN.

FIG. 4 is a schematic diagram showing a conventional address aggregation mechanism.

FIG. 5 is an operation explanatory diagram of the embodiment.

FIG. 6 is an operation explanatory diagram of the embodiment.

FIG. 7 is a schematic diagram illustrating an address aggregation mechanism according to the embodiment;

FIG. 8 is a diagram illustrating the operation of the embodiment.

FIG. 9 is an operation explanatory diagram of the embodiment.

FIG. 10 is a schematic diagram illustrating an address aggregation mechanism according to the embodiment.

[Explanation of symbols]

10, 20: CPN, 11: Core network, 12, 12A, 1
2B: access network, 13, 21: shell network, EN: edge node, BR: border router, RS: route server, DF: default forwarder.

Claims (6)

[Claims] 1. A core element network for performing a packet relay process based on a predetermined core protocol; an access element network accommodating one or a plurality of user networks; and an intermediate between the access element network and the core element network. Inside the combined network, comprising: a core element network connected to the user network using the originating edge relay node and the destination edge relay node, and an outer shell element network that performs a route control process of the packet. A boundary route control node disposed in the outer shell element network and configured to transfer and route the packet between the external network outside the composite network and the composite network; The synthetic network and outside Regarding the extended network in which the network is a constituent element, of the extended addresses used for transferring a predetermined extended packet, the address space of the acquired extended address acquired by the combined network and the non-acquired extended address not acquired The address space is allocated to each of the divided extended address spaces formed by dividing the extended addresses according to a predetermined division pattern so that the extended addresses are serial numbers, and a plurality of logical addresses obtained by logically dividing the composite network are obtained. Each of the divided networks includes at least a divided path information server sharing the own divided network, the edge relay node, and a default transfer device. When exchanging a packet with the external network, the divided The relay node is directly connected to the boundary route Routing system, characterized by sending and receiving packets to your node. 2. The route control system according to claim 1, wherein the predetermined division pattern is such that at least each of the obtained expanded addresses is divided into equal sizes to form each divided expanded address space. A divided network for allocating an acquired divided address space obtained by dividing the address space of the acquired expanded address, and a non-acquired divided address space obtained by dividing the address space of the non-acquired extended address. A route control system, wherein the divided network is at least a logically different divided network. 3. The route control system according to claim 1, wherein the predetermined division pattern is set such that at least with respect to division of the acquired expanded address space, the traffic amount of each divided network is the same. And path control system. 4. The route control system according to claim 1, wherein the predetermined division pattern is obtained by dividing the address space of the acquired expanded address and the address space of the non-acquired expanded address so as to be mixed. A routing control system, wherein the routing control system is assigned to two divided networks. 5. The route control system according to claim 1, wherein when a connection is made from the user network using an extended address that is not assigned to the divided network, the extended address of the default transfer device is used. A default transfer device of the assigned divided network includes first accommodation destination recognizing means for recognizing an edge relay node of the divided network to which the connection has been made, and a default transfer device of the divided network to which the extended address is assigned. A route connection system, wherein a default transfer device transfers a packet directly to the edge relay node. 6. The route control system according to claim 1, wherein when a connection is made from the user network using an extended address not assigned to the divided network, the boundary route control node determines that the connection is not established. A second accommodating destination recognizing means for recognizing an edge relay node of the divided network being performed, wherein the boundary route control node directly transfers a packet to the edge relay node. Route connection system.