JP4044226B2 - ATM relay device and router device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an ATM (Asynchronous Transfer Mode) relay device and a router device.
[0002]
[Prior art]
High-speed switching between hop-by-hop transfer that performs software processing of all IP packets by the CPU of the IP processing unit and cut-through transfer that establishes an ATM path and performs hardware processing with an ATM switch core unit and IP processing unit There is a CSR (Cell Switch Router) method as a method that realizes transfer of an IP packet.
[0003]
In such a CSR system, hop-by-hop transfer generally depends on the performance of the CPU as in the conventional IP router system. That is, in order to increase the transfer capability, it is a common method to increase the CPU speed. Cut-through transfer is used when transferring an IP packet such as FTP or WWW with a long connection time from an adjacent CSR node at high speed. In the CSR method, FANP is used as a protocol for realizing cut-through transfer.
[0004]
In the conventional IP router, as shown in FIG. 9, the path control of the packet data 12 is performed using only the network layer among the plurality of layers 11 of the communication protocol. However, as shown in FIG. This is a technique for transferring packet data at high speed using a data link layer in addition to the layer, and is a high-speed, high-performance router that does not require packet transfer by a CPU (software).
[0005]
FIGS. 11 (a) and 11 (b) are diagrams for explaining such two cell transfer methods of CSR.
[0006]
In FIG. 11A, at the hop-by-hop transfer, the IP packet from the adjacent relay device 33 is sent to the IP processing unit 31A via the ATM switch 31B of the CSR relay device 31, and the software processing is performed by the CPU. To the adjacent relay device 33 '. At this time, a virtual connection VC-1 is established between the adjacent relay device 33 and the CSR relay device 31, and a virtual connection VC-2 is established between the CSR relay device 31 and the adjacent relay device 33 '. 32 is a communication protocol called FANP, which is usually used in CSR, and is a protocol for establishing / releasing a path for transferring a packet directly in the data link layer (ATM-SW).
[0007]
That is, when establishing a session such as FTP from the packet transfer by the IP processing unit 31A and performing cut-through transfer by the ATM switch 31B, as shown in FIG. Establish a cut-through path to switch to processing. The transfer process at this time is performed by the ATM switch 31B, and the IP processing unit 31A does not perform the packet transfer process. A virtual connection VC-3 is established between the CSR relay device 31 and the adjacent relay device 33 'during cut-through transfer.
[0008]
By the way, the above-described CSR cut-through method is roughly divided into two methods. One is a method called topology driven, in which ATM cut-through paths are made to the same subnet, that is, end-to-end CSR, and high-speed transfer of packets with a large amount of data is continuously performed for a long time. Otherwise, hop-by-hop forwarding is performed in the same way as a normal router. Another method is a flow-driven method, in which the high-speed transfer is performed using adjacent CSRs (link-link CSRs).
[0009]
The flow-driven method described above is used when cut-through processing is performed for each individual communication (flow), and routes and packet transfers that require fine quality for each flow that can flexibly respond to changes in the network configuration. It is suitable for a small number of routes. On the other hand, the topology-driven method is used when cut-through is performed for the same destination at once, and is suitable for the backbone of a large-scale network, suitable for routes where a large amount of packet forwarding is expected, etc. There are features.
[0010]
FIG. 12 is a diagram showing a schematic configuration of a conventional CSR relay device, and an IP processing unit 41 is arranged at a subsequent stage of the ATM switch 42 via a link 44 with IP processing. When ATM UNI (User Network Interface) signaling processing is performed, the signaling cell is supplied to a cell / data unit assembly unit (SAR: Segmentation and Reassembly) 43 via a signaling cell path 46 and reassembled into a data unit. This data unit is processed for UNI signaling by a software processing unit 45 including a CPU via a signaling data unit path 49. The reverse is also done.
[0011]
When performing hop-by-hop transfer in IP forwarding, ATM cells are supplied to a cell / data unit assembly unit (SAR) 43 via a cell path 47 other than signaling and reassembled into data units. This data unit is supplied to the software processing unit 45 via a data unit path 48 other than signaling, and performs processing for packet transfer. After such processing, it is transferred again to the cell / data unit assembling unit 43 and assembled into a cell, and then returned to the ATM switch 42 and transferred to the next destination.
[0012]
The cell paths 46 and 47 pass through the same physical link, and the data unit paths 48 and 49 are physically the same socket.
[0013]
[Problems to be solved by the invention]
The above-described CSR method supports ATM UNI signaling processing in both PVC (Permanent Virtual Connection) and SVC (Switched Virtual Connection) connections, and can be connected to an existing ATM network. In the case of supporting SVC, an increase in the amount of signaling traffic is avoided and high-speed signaling performance for SVC control is required. These two traffics must be processed and transferred by a single physical link.
[0014]
However, the conventional ATM relay device described above cannot satisfy these requirements. Further, in the conventional ATM relay device, the processing capability of the traffic that does not perform cut-through transfer by the CSR method, that is, the transfer capability of the packet that is hop-by-hop transfer is processed by software (processing by the CPU). .
[0015]
The present invention has been made paying attention to such problems, and the object of the present invention is to achieve high-speed signaling performance for SVC control while avoiding an increase in the amount of signaling traffic particularly when supporting SVC. It is another object of the present invention to provide an ATM relay apparatus and a router apparatus that can improve the packet transfer capability during hop-by-hop transfer.
[0016]
The present invention also provides an ATM relay device capable of greatly improving the packet transfer capability at the time of hop-by-hop transfer in the network packet relay device using the CSR method, particularly in the processing of packets addressed to the own station and non-IP packets. There is to do.
[0017]
Furthermore, the present invention can reliably receive a packet addressed to the local station without performing TTL subtraction more than necessary in the processing of the packet addressed to the local station or an unknown destination IP packet. An object of the present invention is to provide an ATM relay device capable of improving the transfer distance.
[0018]
[Means for Solving the Problems]
To achieve the above object, according to a first aspect of the present invention, there is provided an ATM relay apparatus having a network layer packet transfer function using an ATM switch, which is provided in association with the ATM switch, via the first link. Signaling processing means for receiving a signaling packet from the ATM switch and performing signaling processing; and this signaling processing means is provided separately, and the IP packet is transmitted via a second link physically different from the first link. And forwarding processing means for performing IP packet transfer processing by hardware using a predetermined memory.
[0019]
Further, the second invention is characterized in that, in the first invention, at least a packet addressed to the own station and an IP packet that cannot be transferred by the forwarding processing means are physically defined as the first and second links. CPU processing means is further provided for performing CPU processing upon receiving via a different third link.
[0020]
In a third aspect based on the second aspect, the second link for signaling processing and the third link for CPU processing are physically the same link.
[0021]
According to a fourth aspect of the present invention, in the second or third aspect of the present invention, at least a packet addressed to the own station in the ATM relay apparatus and an available time of an IP packet that cannot be transferred by the forwarding processing means. And adjusting means for adjusting.
[0022]
In a fifth aspect of the present invention, in a router apparatus provided in association with an ATM switch and having an IP packet transfer function in a network layer, a signaling packet is received from the ATM switch via a first link and a signaling process is performed. The signaling processing means for performing the processing and the signaling processing means are provided separately, and receives an IP packet via a second link physically different from the first link, and performs an IP packet forwarding process. And a forwarding processing means which is performed by hardware using a memory.
[0023]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.
[0024]
FIG. 1 is a diagram showing a schematic configuration of a CSR relay device to which the ATM relay device according to the first embodiment of the present invention is applied. In the configuration of the present embodiment, the ATM switch 42 is connected to the IP processing unit 61 via the link 63 with the IP processing for signaling and is physically different from the link 63 with the IP processing for signaling. Is connected to the IP processing unit 61 via a link 44 with the IP processing. That is, a path for performing forwarding processing (here, hop-by-hop transfer) and a path for performing signaling processing are physically different.
[0025]
More specifically, the ATM switch 42 is connected to a signaling cell / data unit assembly (hereinafter referred to as a signaling SAR) 62 via a signaling cell path 46, and the signaling SAR 62 is connected via a signaling data unit path 96. Connected to a software processing unit 52 including a CPU. The signaling cell path 46, the signaling SAR 62, the signaling data unit path 96, and the software processing unit 52 constitute signaling processing means.
[0026]
The ATM switch 42 is a cell / data unit assembly (for hop-by-hop transfer) provided separately from the signaling SAR 62 via another cell path 47 physically different from the above-described signaling processing path. (Referred to as SAR) 43. The hop-by-hop transfer SAR 43 is connected to the IP forwarding processing unit 54 via the data unit path 97. The cell path 47, hop-by-hop transfer SAR 43, data unit path 97, and IP forwarding processing unit 54 described above constitute a forwarding processing means.
[0027]
Here, in the present embodiment, as a means for improving the hop-by-hop transfer capability as the forwarding process, the header address of the IP layer and the TCP / UDP layer is detected by hardware, and the associative memory and hardware cache search are performed. The packet forwarding process is performed at high speed without performing software processing by the CPU using the IP packet forwarding table in combination. Such forwarding processing is described in Japanese Patent Application No. 9-252161 by the present applicant.
[0028]
The operation of the CSR relay device shown in FIG. 1 will be described below. The cells input to the ATM switch 42 in FIG. 1 are input to the IP processing unit 61 via dedicated links according to their types. That is, if it is a signaling cell, it is transferred to the signaling SAR 62, and the other cells are transferred to the hop-by-hop transfer SAR 43 and reassembled into data units by the respective SARs.
[0029]
The data unit reassembled by the signaling SAR 62 is directly transferred to the software processing unit 52 without being taken into the IP forwarding processing unit 54, and software processing by the CPU is performed.
[0030]
On the other hand, the data unit reassembled by the hop-by-hop transfer SAR 43 is transferred to the IP forwarding processing unit 54, and a hop-by-hop transfer process is performed. That is, it is determined whether or not it matches the IP address registered in the table of associative memory, and if it matches, the address information to be sent to the next relay device is changed according to the address information, and then the hop-by. The cell is reassembled by the hop transfer SAR 43 and sent to the ATM switch 42.
[0031]
Here, it is assumed that rewriting and management of the associative memory table is performed by software processing by the CPU.
[0032]
As described above, in this embodiment, the signaling cell and other cells are transferred through physically different links and are processed separately, so that the amount of signaling traffic particularly in the case of supporting SVC. High-speed signaling performance for SVC control can be obtained while avoiding the increase. Further, the hop-by-hop transfer process is not performed by transferring to the CPU of the software processing unit 52, but is performed by hardware processing in the IP forwarding processing unit 54. The transfer capability can be greatly improved.
[0033]
FIG. 2 is a view showing a modification of the above-described embodiment. In this modification, a plurality of IP processing units 61-1, 61-2 and 61-3 are connected to the ATM switch 42. Each IP processing unit 61-1, 61-2, 61-3 is assumed to have the configuration shown in FIG. According to such a configuration, since there are a plurality of configurations of the present embodiment as shown in FIG. 1, the above-described effects can be obtained synergistically.
[0034]
According to the first embodiment described above, in the network packet relay apparatus according to the CSR method, particularly when SVC is supported, signaling performance can be obtained, and the packet transfer capability during hop-by-hop transfer is greatly increased. Can be improved.
[0035]
The second embodiment of the present invention will be described below. In the CSR relay apparatus described in the first embodiment, the IP forwarding processing unit 54 is implemented as hardware in order to improve the hop-by-hop transfer capability that has hindered the original capability of the CSR. Determines the IP packet addressed to its own station including the protocol and the packet whose hop-by-hop transfer processing method is unknown, and transfers the packet to the software processing unit 52. It is necessary to perform processing for receiving a packet and a packet that has not been processed, or a packet generated by processing this packet.
[0036]
Further, when the destination address is not entered in the IP forwarding processing unit 54, the address entry in the IP forwarding processing unit 54 is updated. In particular, some of the packets whose processing is unknown have not been entered in the forwarding address in the IP forwarding processing unit 54, and when sending the corresponding packet, the address entry of the IP forwarding processing unit 54 The number of cases where it becomes necessary to perform an update increases.
[0037]
Such processing hinders the improvement of the original hop-by-hop transfer capability of the IP forwarding processing unit 54 and makes the structure of the IP forwarding processing unit 54 complicated.
[0038]
Therefore, in the second embodiment, it is possible to provide an ATM relay device that can improve the original capability of CSR by further improving the transfer capability of hop-by-hop, and can simplify the structure of the device. Intended.
[0039]
Hereinafter, a second embodiment of the present invention will be described in detail with reference to the drawings. FIG. 3 is a diagram showing a schematic configuration of a CSR relay device to which the ATM relay device according to the second embodiment of the present invention is applied. In the configuration of this embodiment, the ATM switch 142 is connected to the IP processing unit 100 via a link 163 with IP processing for signaling, and is physically different from the link 163 with IP processing for signaling. It is connected to the IP processing unit 100 via a link 106 with the IP processing for hop transfer.
[0040]
Furthermore, the IP processing unit 100 is connected via a link 103 with IP processing for software processing (CPU processing) that is physically different from the links 163 and 106 with IP processing for signaling and forwarding. Thus, the path for performing forwarding processing (here, hop-by-hop transfer), the path for performing signaling processing, and the path for performing software processing are physically different.
[0041]
More specifically, the ATM switch 142 is connected to a signaling cell / data unit assembly (hereinafter referred to as a signaling SAR) 162 via a signaling cell path 146, and the signaling SAR 162 is connected via a signaling data unit path 246. Connected to a software processing unit 152 including a CPU. The signaling cell path 146, the signaling SAR 162, the signaling data unit path 246, and the software processing unit 152 constitute a signaling processing unit.
[0042]
On the other hand, the ATM switch 142 is a cell / data unit assembly provided separately from the signaling SAR 162 via another cell path (hop-by-hop cell path) 105 that is physically different from the above-described signaling processing path. (Referred to as a hop-by-hop transfer SAR) 107. The hop-by-hop transfer SAR 107 is connected to the IP forwarding processing unit 254 via the hop-by-hop data unit path 205. The cell path 105, the hop-by-hop transfer SAR 107, the hop-by-hop data unit path 205, and the IP forwarding processing unit 254 constitute a forwarding processing unit.
[0043]
The ATM switch 142 is connected to the CPU processing cell / data unit assembling unit 102 via a CPU processing cell path 104 physically different from the signaling cell path 146 and the hop-by-hop cell path 105 described above. The cell / data unit assembly unit 102 for CPU processing is connected to a software processing unit 152 including a CPU via a CPU processing data unit path 201.
[0044]
Here, in this embodiment, the header address of the IP layer and the TCP / UDP layer is detected by hardware as means for improving the hop-by-hop transfer capability as forwarding processing, and the associative memory and hardware cache search are used in combination. The IP packet forwarding table may be used to perform packet forwarding processing at high speed without performing software processing by the CPU. Such forwarding processing is described in Japanese Patent Application No. 9-252161.
[0045]
The operation of the CSR relay device shown in FIG. 3 will be described below. The cells input to the ATM switch 142 in FIG. 3 are input to the IP processing unit 100 via dedicated links according to their types. That is, if it is a signaling cell, it is transferred to the signaling SAR 162, and the other cells are transferred to the hop-by-hop transfer SAR 107 and reassembled into data units by the respective SARs.
[0046]
The data unit reassembled by the signaling SAR 162 is directly transferred to the software processing unit 152 without being taken into the IP forwarding processing unit 254, and software processing by the CPU is performed.
[0047]
On the other hand, the data unit reassembled by the hop-by-hop transfer SAR 107 is transferred to the IP forwarding processing unit 254, and a hop-by-hop transfer process is performed. That is, it is determined whether or not it matches the IP address registered in the table of associative memory, and if it matches, the address information to be sent to the next relay device is changed according to the address information, and then the hop-by. The cell is assembled again in the hop transfer SAR 107 and sent to the ATM switch 142.
[0048]
Here, it is assumed that rewriting and management of the associative memory table is performed by software processing by the CPU.
[0049]
Also, there is a VC table memory in which address information, specifically VC information, is registered, and the VC table memory in which the corresponding VC information is stored in the associative memory area where the IP address of the packet matches. The address is stored. Here, the virtual connection VC-11 is associated with the IP address addressed to the local station to be transferred to the software processing unit 152. Further, a data unit whose hop-by-hop transfer processing method is unknown is associated with the virtual connection VC-11 so that it is transferred to the software processing unit 152 in the same manner as a non-IP packet. The virtual connection VC-11 may be divided into a plurality of virtual connections.
[0050]
When the IP forwarding processing unit 254 receives an IP packet addressed to itself such as FANP protocol data, the packet is given address information corresponding to the virtual connection VC-11 from the IP address, and then the hop-by-hop transfer SAR 107. Then, it is assembled again into a cell and sent out to the ATM switch 142 again. The ATM switch 142 determines that the virtual connection VC-11 should be transferred to the software processing unit 152, transfers this cell to the CPU processing SAR 102 via the CPU processing cell path 104, and the CPU processing SAR 102. Then, the cells are assembled into data units, and the assembled IP packets are delivered to the software processing unit 152 by the CPU.
[0051]
Furthermore, when an IP packet or a non-IP packet data unit whose hop-by-hop transfer processing method is unknown arrives at the IP forwarding processing unit 254, address information corresponding to the virtual connection VC-11 is given as a non-IP packet. Thereafter, the cell is assembled again in the hop-by-hop transfer SAR 107 and sent out to the ATM switch 142 again.
[0052]
The ATM switch 142 determines that the virtual connection VC-11 should be transferred to the software processing unit 152, transfers this cell to the CPU processing SAR 102 via the CPU processing cell path 104, and the CPU processing SAR 102. Then, the cells are assembled into data units, and the assembled data units are transferred to the software processing unit 152 by the CPU.
[0053]
On the other hand, the IP packet originated from the software processing unit 152 is given a destination IP address by the software processing unit 152, and the CPU processing SAR 102 transfers the IP packet by the hop-by-hop transfer function of the IP forwarding unit 254. Address information corresponding to the virtual connection VC-11 is given and delivered to the IP forwarding processing unit 254 via the ATM switch 142, the cell path 105, and the SAR 107, and hop-by-hop transfer processing is performed.
[0054]
Further, the data unit of the non-IP packet whose hop-by-hop transfer processing method is unknown in the IP forwarding processing unit 254 is not transferred to the IP forwarding processing unit 254, but is transferred to the destination virtual connection VC-10. The address information corresponding to is attached and sent out.
[0055]
As described above, in the present embodiment, data transfer between the IP forwarding processing unit 254 and the software processing unit 152 is physically separated to reduce the load on the IP forwarding processing unit 254. Packet forwarding capability can be improved, and the configuration of the IP forwarding processing unit 254 can be simplified. In addition, since the conventional technology can be used as it is, the configuration of each function other than the IP forwarding processing unit 254 can be simplified.
[0056]
FIG. 4 is a diagram showing a modification of the above-described embodiment. In this modification, a link for signaling processing and a link for CPU processing are connected to the IP processing unit 100 via a link 111. That is, the IP processing unit 100 is physically connected via the same link. According to such a configuration, the configuration of FIG. 3 can be further simplified, so that the above-described effects can be obtained synergistically in a communication mode with few signaling processes.
[0057]
According to the second embodiment described above, in the packet packet relay apparatus using the CSR method, it is possible to greatly improve the packet transfer capability at the time of hop-by-hop transfer, particularly in the processing of packets addressed to the local station and the non-IP packet.
[0058]
The third embodiment of the present invention will be described below. In the ATM relay device of the above-described embodiment, every time an IP packet that passes through the IP forwarding processing unit 254 passes through the IP forwarding processing unit 254, TTL (Time To Live) is subtracted more than the actual processing time. Therefore, there is a problem that the packet addressed to the local station does not reach the software processing unit 152 as a result of the TTL value being subtracted even though it has once reached the IP processing unit 100. In addition, in the case of an IP packet whose destination is unknown, the TTL value subtraction is performed at least three times, twice in the IP forwarding processing unit 254 and once in the software processing unit 152, so that it is replaced with a network node. There is a problem that the transfer distance is shortened or packets are discarded in the ATM relay apparatus. Such a problem will be further described below.
[0059]
FIG. 5 shows an example of TTL subtraction according to the packet flow. When an IP packet addressed to the own station is transferred through the IP packet transfer path 302, first, the IP forwarding processing unit 254 performs TTL subtraction for the first time. Here, when the subtraction value at the IP forwarding processing unit 254 is 1 or the TTL value a held by the IP packet (not shown) is 1, the TTL value after subtraction is 0, so the software processing unit 152 IP packets cannot be transferred to
[0060]
On the other hand, for an IP packet whose destination is unknown, the TTL subtraction is performed for the first time by the IP forwarding processing unit 254, for the second time by the software processing unit 152, and for the third time by the IP forwarding processing unit 254 at the time of transfer after address resolution.
[0061]
Here, when the subtraction value in the IP forwarding processing unit 254 is 1 and the subtraction value in the software processing unit 152 is 1, the TTL value after the subtraction is a-3 as illustrated.
[0062]
Here, the IP forwarding processing unit 254 is configured by hardware, and can be realized sufficiently shorter than a generally known unit time of TTL subtraction. Moreover, even if the processing times of all ATM relay apparatuses are combined, the unit time may not be reached. In this case, the TTL subtraction value in the entire ATM relay apparatus can be set to 1, but as described above, In the ATM relay device of the embodiment, the subtraction value increases.
[0063]
Therefore, the third embodiment provides an ATM relay device that does not perform TTL subtraction more than necessary, reliably receives a packet addressed to itself, and improves the packet transfer distance when replaced with a network node. With the goal. In order to achieve this object, means for adjusting the TTL subtraction for a specific IP packet is provided in the IP forwarding processing means and the software processing means via the third link of the ATM relay apparatus. Hereinafter, the third embodiment will be described in detail.
[0064]
6 is a configuration example of an ATM relay device according to the third embodiment, an example of an IP packet transfer path, and TTL subtraction. FIG. 7 is a flowchart of TTL calculation processing applied to the IP forwarding processing unit 254 (steps S401 to S405). FIG. 8 is a diagram illustrating an example of a flowchart (steps S501 to S504) of a TTL calculation process applied to the software processing unit 152.
[0065]
In the third embodiment, the ATM switch 142 is connected to the IP processing unit 100 via a link 111 with the IP processing unit 100 and via a link 106 that is physically different from the link 111 with the IP processing unit 100. Connected to the IP processing unit 100.
[0066]
That is, a path for performing forwarding processing (here, hop-by-hop transfer) is physically different from a path for performing signaling processing and software processing.
[0067]
More specifically, the ATM switch 142 is connected to the SAR 109 via the signaling cell path 146, and the SAR 109 is connected to the software processing unit 152 including the CPU via the signaling data unit path 246. The signaling cell path 246, the SAR 109, and the software processing unit 152 described above constitute a signaling processing unit.
[0068]
On the other hand, the ATM switch 142 is connected to a SAR 107 provided separately from the SAR 109 via another cell path 105 physically different from the above-described signaling processing path. The SAR 107 is connected to the IP forwarding processing unit 254 via the data unit path 205. The cell path 105, the SAR 107, and the IP forwarding processing unit 254 described above constitute a forwarding processing unit.
[0069]
Further, the ATM switch 142 is connected to the SAR 109 via another cell path 104 different from the above-described signaling and forwarding process paths. The SAR 109 is connected to a software processing unit 152 including a CPU via a data unit path 201. The cell path 104, the SAR 109, and the software processing unit 152 described above constitute a CPU processing unit for packets.
[0070]
In the configuration described above, the IP forwarding processing unit 254 determines whether or not it is addressed to the own station based on the destination of the received IP packet (step S402), or determines whether or not the destination is unknown (step S403), and determines the TTL. It is determined whether or not to perform subtraction, and TTL subtraction processing (step S404) is performed. Also, the software processing unit 152 sends the packet to the IP forwarding processing unit 254 again. The TTL value of the IP packet is added by the numerical value subtracted by the IP forwarding processing unit 254.
[0071]
Here, also in this embodiment, the forwarding processing by hardware processing described in Japanese Patent Application No. 9-252161 described in the second embodiment may be performed.
[0072]
The operation of the ATM relay apparatus shown in FIG. 6 will be described below. A cell input from the outside to the ATM switch 142 is input to the IP processing unit 100 through a dedicated link according to the type of the cell. That is, if it is a signaling cell, it is transferred to the SAR 109, and the other cells are transferred to the SAR 107 and reassembled into data units by the respective SARs.
[0073]
The data unit reassembled by the SAR 109 is directly transferred to the software processing unit 152 without being taken into the IP forwarding processing unit 254, and software processing by the CPU is performed.
[0074]
On the other hand, the data unit reassembled by the SAR 107 is transferred to the IP forwarding processing unit 254, and hop-by-hop transfer processing including TTL calculation is performed. That is, it is determined whether or not it matches the IP address registered in the table of associative memory. If the IP address matches, the TTL operation is performed according to the address information and the address information to be sent to the next relay device is determined. After the replacement, the cell is reassembled by the hop-by-pop transfer SAR 107 and sent to the ATM switch 142.
[0075]
Here, it is assumed that rewriting and management of the associative memory table is performed by software processing by the CPU. Also, a VC table memory in which address information, specifically, virtual connection (VC) information is registered, is stored in an associative memory area in which the packet IP address matches, and the corresponding VC information is stored in the VC. Stores the address of the table memory.
[0076]
Here, VC = VC1 is associated with the IP address of the local station as a virtual connection to be transferred to the CPU processing unit. Further, a data unit whose hop-by-hop transfer processing method is unknown is associated with VC = VC1 in the same manner as an unknown IP packet. The VC may be divided into a plurality of virtual connections.
[0077]
When an IP packet addressed to itself such as FANP protocol data arrives at the IP forwarding processing unit 254, the IP forwarding processing unit 254 does not perform TTL subtraction. In this case, the packet is given address information corresponding to VC = VC1 from its IP address, and is then reassembled into a cell by the SAR 107 and sent out to the ATM switch 142 again.
[0078]
The ATM switch 142 judges from VC = VC1 and transfers this cell to the SAR 109 via the CPU processing cell path 104. In the SAR 109, the cell is assembled into a data unit, and the assembled IP packet is sent to the software processing unit 152 by the CPU. Is passed without a TTL error.
[0079]
Further, when a data unit of an IP packet whose hop-by-hop transfer processing method is unknown arrives at the IP forwarding processing unit 254, TTL subtraction is not performed, and address information corresponding to VC = VC1 is set as an unknown IP packet. After being granted, the cell is reassembled in the hop-by-hop forwarding SAR 107 and sent out to the ATM switch 142 again.
[0080]
The ATM switch 142 transfers this cell to the SAR 109 via the CPU processing cell path 152 based on the value of VC = VC1. In the SAR 109, the cell is assembled into a data unit, and the assembled data unit is a software processing unit 152 by the CPU. Is passed on.
[0081]
The software processing unit 152 performs IP packet destination analysis. Here, the subsequent transfer process is not performed for the IP packet addressed to the own station and the IP packet whose next transfer destination is unknown, and the TTL value corresponding to the processing time in the software processing unit 152 is subtracted for the other packets (FIG. 8). In step S502), the TTL value subtracted by the IP forwarding processing unit 254 is incremented by 1 (step S503). The reason why 1 is added here is that 1 is subtracted by the IP forwarding processing unit 254.
[0082]
After the destination information of the IP packet is registered in the IP forwarding processing unit 154, the packet subjected to the TTL calculation is transferred again to the IP forwarding processing unit 254 via the ATM switch 142.
[0083]
Here, as an example, if the subtraction value of the TTL operation is 1 in FIG. 7, the addition value of the TTL operation is 1 in FIG. 8, and the TTL subtraction value of the software processing unit 152 is 1, the IP forwarding processing unit 254 in FIG. In this case, when an IP packet whose transfer destination is unknown is transferred from the outside to the IP packet transfer path 302, first, the TTL subtraction in the IP forwarding processing unit 254 is not performed, and the software processing unit 152 sets the original TTL value to 1. One subtraction and one addition.
[0084]
Further, the IP forwarding processing unit 254 performs a process of subtracting one TTL value in order to newly register a destination through a series of processes. In other words, only one TTL value is decremented from when the IP packet is input to the ATM relay apparatus of this embodiment until it is transmitted.
[0085]
According to the third embodiment described above, in the network packet relay apparatus based on the ATM relay system, in particular, in the packet destined for the local station and the IP address packet with unknown destination, packet erasure and network in the relay apparatus at the time of hop-by-hop transfer The effect of preventing a decrease in the internal transfer distance can be expected. That is, TTL subtraction is not performed more than necessary, and a packet addressed to the local station can be received reliably, and further, the packet transfer distance when replaced with a network node can be improved.
[0086]
【The invention's effect】
According to the invention described in claim 1 or claim 5, it is possible to obtain high-speed signaling performance for SVC control while avoiding an increase in the amount of signaling traffic particularly when SVC is supported, and hopping Packet transfer capability during bi-hop transfer can be improved.
[0087]
In addition, according to the second aspect of the present invention, in the network packet relay apparatus according to the CSR method, the packet transfer capability at the time of hop-by-hop transfer is greatly improved, particularly in the packet addressed to the local station and non-IP packet processing Will be able to.
[0088]
According to the invention described in claim 3, in addition to the effect described in claim 2, the configuration can be further simplified.
[0089]
In addition, according to the invention described in claim 4, it is possible to reliably receive a packet addressed to the local station without performing TTL subtraction more than necessary, and further, the transfer distance of the packet when replaced with a network node is obtained. Can be improved.
[Brief description of the drawings]
FIG. 1 is a diagram showing a schematic configuration of a CSR relay device to which an ATM relay device according to a first embodiment of the present invention is applied.
FIG. 2 is a diagram illustrating a modification of the CSR relay device illustrated in FIG. 1;
FIG. 3 is a diagram showing a schematic configuration of a CSR relay device to which an ATM relay device according to a second embodiment of the present invention is applied.
4 is a diagram showing a modification of the ATM relay device shown in FIG. 3;
FIG. 5 is a diagram showing a packet flow and a TTL calculation according to the schematic configuration of the ATM relay apparatus shown in FIG. 4;
FIG. 6 is a diagram illustrating a flowchart of TTL calculation processing applied to the IP forwarding processing unit of the present embodiment.
FIG. 7 is a diagram showing a flowchart of TTL calculation processing applied to the software processing unit of the present invention.
FIG. 8 is a diagram showing a packet flow and a TTL calculation according to a schematic configuration of the ATM relay apparatus of the present embodiment.
FIG. 9 is a diagram for explaining route control of packet data using a network layer.
FIG. 10 is a diagram for explaining high-speed transfer of packet data using a data link layer.
FIG. 11 is a diagram for explaining a cell transfer method of CSR.
FIG. 12 is a diagram illustrating a schematic configuration of a conventional CSR relay device.
[Explanation of symbols]
42 ... ATM switch,
43 ... cell / data unit assembly (SAR for hop-by-hop transfer),
44 ... Link with IP processing (for hop-by-hop forwarding),
46 ... signaling cell path,
Cell path other than 47 ... 46,
52 ... Software processing unit,
54 ... IP forwarding processing unit,
61 ... IP processing part,
62 ... Cell / data unit assembly (SAR for signaling),
63 ... Link with IP processing (for signaling)
96 ... signaling data unit path,
Data unit path other than 97 ... 96.

Claims (5)

In an ATM relay device having a packet transfer function of a network layer using an ATM switch,
Signaling processing means provided in association with an ATM switch, for receiving a signaling packet from the ATM switch via a first link and performing signaling processing;
It is provided separately from the signaling processing means, receives an IP packet via a second link physically different from the first link, and performs IP packet transfer processing in hardware using a predetermined memory. Forwarding processing means to be performed,
An ATM relay device comprising: At least a packet addressed to the own station and an IP packet that cannot be transferred by the forwarding processing means are received via a third link physically different from the first and second links, and CPU processing is performed. 2. The ATM relay apparatus according to claim 1, further comprising CPU processing means for performing the processing. 3. The ATM relay apparatus according to claim 2, wherein the second link for signaling processing and the third link for CPU processing are physically the same link. 3. An adjusting means for adjusting at least a packet addressed to the local station and an available time of an IP packet that cannot be transferred by the forwarding processing means in the ATM relay apparatus. 3. The ATM relay device according to 3. In a router device provided in association with an ATM switch and having an IP packet transfer function in a network layer,
Signaling processing means for receiving a signaling packet from the ATM switch via a first link and performing signaling processing;
It is provided separately from the signaling processing means, receives an IP packet via a second link physically different from the first link, and performs IP packet transfer processing in hardware using a predetermined memory. Forwarding processing means to be performed,
A router device comprising:

Images