WO1999034558A1 - Repeteur mta et reseau le comprenant - Google Patents
Repeteur mta et reseau le comprenant Download PDFInfo
- Publication number
- WO1999034558A1 WO1999034558A1 PCT/JP1998/005964 JP9805964W WO9934558A1 WO 1999034558 A1 WO1999034558 A1 WO 1999034558A1 JP 9805964 W JP9805964 W JP 9805964W WO 9934558 A1 WO9934558 A1 WO 9934558A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- atm
- processing
- cell
- information
- bucket
- Prior art date
Links
- 238000012545 processing Methods 0.000 claims abstract description 447
- 238000004891 communication Methods 0.000 claims abstract description 79
- 238000012546 transfer Methods 0.000 claims description 268
- 238000000034 method Methods 0.000 claims description 172
- 230000011664 signaling Effects 0.000 claims description 136
- 230000006870 function Effects 0.000 claims description 77
- 230000008569 process Effects 0.000 claims description 60
- 238000012544 monitoring process Methods 0.000 claims description 26
- 238000003860 storage Methods 0.000 claims description 15
- 238000012806 monitoring device Methods 0.000 claims description 9
- 238000001514 detection method Methods 0.000 claims description 6
- 238000005242 forging Methods 0.000 claims description 5
- 238000005562 fading Methods 0.000 claims description 2
- 230000007704 transition Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 118
- 238000006243 chemical reaction Methods 0.000 description 49
- 238000012986 modification Methods 0.000 description 22
- 230000004048 modification Effects 0.000 description 22
- 238000005516 engineering process Methods 0.000 description 12
- 230000005540 biological transmission Effects 0.000 description 11
- 238000001914 filtration Methods 0.000 description 11
- 238000009826 distribution Methods 0.000 description 10
- 230000009977 dual effect Effects 0.000 description 9
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 6
- 239000000284 extract Substances 0.000 description 6
- 238000007616 round robin method Methods 0.000 description 6
- 208000037820 vascular cognitive impairment Diseases 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 230000000873 masking effect Effects 0.000 description 5
- 101100294112 Caenorhabditis elegans nhr-47 gene Proteins 0.000 description 4
- 101100222704 Neurospora crassa (strain ATCC 24698 / 74-OR23-1A / CBS 708.71 / DSM 1257 / FGSC 987) csr-1 gene Proteins 0.000 description 4
- 230000022131 cell cycle Effects 0.000 description 4
- 238000007796 conventional method Methods 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 4
- 238000003672 processing method Methods 0.000 description 4
- 230000009467 reduction Effects 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000001193 catalytic steam reforming Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 241000255925 Diptera Species 0.000 description 2
- 241001522296 Erithacus rubecula Species 0.000 description 2
- 206010041235 Snoring Diseases 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 2
- 230000003111 delayed effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000000275 quality assurance Methods 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- HHDUMDVQUCBCEY-UHFFFAOYSA-N 4-[10,15,20-tris(4-carboxyphenyl)-21,23-dihydroporphyrin-5-yl]benzoic acid Chemical compound OC(=O)c1ccc(cc1)-c1c2ccc(n2)c(-c2ccc(cc2)C(O)=O)c2ccc([nH]2)c(-c2ccc(cc2)C(O)=O)c2ccc(n2)c(-c2ccc(cc2)C(O)=O)c2ccc1[nH]2 HHDUMDVQUCBCEY-UHFFFAOYSA-N 0.000 description 1
- KOUVXIWGXJSZHF-UHFFFAOYSA-N 9-[3-[4,6-bis(3-carbazol-9-ylphenyl)-1,3,5-triazin-2-yl]phenyl]carbazole Chemical compound C12=CC=CC=C2C2=CC=CC=C2N1C1=CC(C=2N=C(N=C(N=2)C=2C=C(C=CC=2)N2C3=CC=CC=C3C3=CC=CC=C32)C=2C=CC=C(C=2)N2C3=CC=CC=C3C3=CC=CC=C32)=CC=C1 KOUVXIWGXJSZHF-UHFFFAOYSA-N 0.000 description 1
- 101100461812 Arabidopsis thaliana NUP96 gene Proteins 0.000 description 1
- 101100148729 Caenorhabditis elegans sar-1 gene Proteins 0.000 description 1
- 208000032544 Cicatrix Diseases 0.000 description 1
- 235000008557 Ehretia anacua Nutrition 0.000 description 1
- 240000008235 Ehretia anacua Species 0.000 description 1
- 101000653469 Homo sapiens T-complex protein 1 subunit zeta Proteins 0.000 description 1
- 235000008694 Humulus lupulus Nutrition 0.000 description 1
- 102100030664 T-complex protein 1 subunit zeta Human genes 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000006727 cell loss Effects 0.000 description 1
- 230000005754 cellular signaling Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 239000013256 coordination polymer Substances 0.000 description 1
- 230000002354 daily effect Effects 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005187 foaming Methods 0.000 description 1
- 230000002431 foraging effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 229940002612 prodrug Drugs 0.000 description 1
- 239000000651 prodrug Substances 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 235000002020 sage Nutrition 0.000 description 1
- 231100000241 scar Toxicity 0.000 description 1
- 230000037387 scars Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000032258 transport Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
- H04L47/622—Queue service order
- H04L47/6225—Fixed service order, e.g. Round Robin
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
- H04L47/621—Individual queue per connection or flow, e.g. per VC
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
- H04L47/6215—Individual queue per QOS, rate or priority
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/10—Packet switching elements characterised by the switching fabric construction
- H04L49/104—Asynchronous transfer mode [ATM] switching fabrics
- H04L49/105—ATM switching elements
- H04L49/108—ATM switching elements using shared central buffer
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/25—Routing or path finding in a switch fabric
- H04L49/256—Routing or path finding in ATM switching fabrics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/30—Peripheral units, e.g. input or output ports
- H04L49/3081—ATM peripheral units, e.g. policing, insertion or extraction
- H04L49/309—Header conversion, routing tables or routing tags
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5638—Services, e.g. multimedia, GOS, QOS
- H04L2012/5646—Cell characteristics, e.g. loss, delay, jitter, sequence integrity
- H04L2012/5651—Priority, marking, classes
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5638—Services, e.g. multimedia, GOS, QOS
- H04L2012/5646—Cell characteristics, e.g. loss, delay, jitter, sequence integrity
- H04L2012/5652—Cell construction, e.g. including header, packetisation, depacketisation, assembly, reassembly
- H04L2012/5653—Cell construction, e.g. including header, packetisation, depacketisation, assembly, reassembly using the ATM adaptation layer [AAL]
- H04L2012/5658—Cell construction, e.g. including header, packetisation, depacketisation, assembly, reassembly using the ATM adaptation layer [AAL] using the AAL5
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5638—Services, e.g. multimedia, GOS, QOS
- H04L2012/5665—Interaction of ATM with other protocols
- H04L2012/5667—IP over ATM
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
Definitions
- the present invention uses the destination information represented by TCPZIP.
- the present invention relates to an ATM repeater based on the ATM (.Asynchronous Transfer Mode) 1 ⁇ method based on the IP relay function in the packet communication method and a network including the same.
- ATM Asynchronous Transfer Mode
- a packet relay device such as a bridge or a router
- logical network sub-network
- logical network sub-networks
- routing processing is performed for each bucket destination in the network layer, and transfer is performed.
- FIG. 1 is a diagram showing the configuration of such a conventional ATM relay device.
- the ATM switch 16 includes an ATM switch core 160a having a plurality of ports (# 1 to #N). 0, which is connected to the following stage, and the cell processing unit assembly (SAR: Segmentation And Reassembly) 16 1 a, and the IP processing unit 16 including the CPU
- CSR cell switch router
- the conventional IP router controls the route of the bucket data 372 using only the network layer out of the multiple layers 371 of the communication protocol.
- CSR is a technology that uses a data link layer in addition to the network layer to transfer bucket data at high speed. Depending on the CPU (software). It is a high-speed, high-performance router that does not require packet transfer.
- VPI / VCI used in the ATM relay device will be briefly described.
- FIGS. 4A and 4B are diagrams for explaining the two cell transfer methods of CSR described above.
- the IP packet from the adjacent relay device 383 is transmitted via the ATM switch 381B of the CSR relay device 381 to the IP processor.
- the data is sent to the shinging section 381 A, subjected to software processing by the CPU, and transferred to the next adjacent relay device 383 '.
- a virtual connection VC_1 is provided between the adjacent relay device 383 and the CSR relay device 381
- a virtual connection VC_1 is provided between the CSR relay device 381 and the adjacent relay device 3883 '.
- a virtual connection VC-2 is set up.
- 382 is a communication protocol called FANP, which is usually used in CSR, and de-establishes the path for transferring the bucket to the direct in the data link layer (ATM-SW). It is a protocol for
- the IP processing unit 3881A When establishing a session such as a packet transfer power, FTP, etc., and performing cut-through transfer by the ATM switch 381B, as shown in FIG. 4B. Thus, an ATM path is established, and a cut-through path for switching to ATM layer processing is established. The transfer process at this time is performed by the ATM switch 3881B, and the IP processing unit 3881A does not perform the bucket transfer process.
- Reference numeral 388 denotes a virtual connection VC-3 set between the CS R relay device 381 and the adjacent relay device 383 'during cut-through transfer.
- topology driven to the same subnet that is, ATM cut through to the end-end CSR. And reduce the amount of data
- This is a method in which high-speed transfer of many buckets is performed continuously for a long time. If not, hop-by-hop transfer is performed in the same way as a normal router.
- the other method is a one-drive method, in which adjacent CSRs (link-link CSRs) perform the above-mentioned high-speed transfer.
- the above-mentioned flow-driven method is used when performing cut-through processing for each individual communication (flow), and can flexibly respond to changes in the network configuration. However, it is characterized by the fact that it is suitable for routes that require detailed quality for each flow and for routes with a small number of buckets.
- the topology driven method is used when performing cut-through for the same destination at once, and is suitable for the backbone of large-scale networks. It is suitable for routes where a large number of buckets are expected to be transferred.
- an IP processing unit is provided at the subsequent stage of the ATM switch, and the IP processing unit controls cut-through transfer by the ATM switch.
- FIG. 5 shows the configuration of a CSR type ATM relay device.
- the IP processing unit 290 is connected to the ATM switch 290 connected to the IP processing unit 290 via one physical path. It consists of and. Then, via a cell / data unit assembly (SAR) 295 interposed in this physical path, an IP forwarding process (IP bucket, Routing information), signaling processing signals, and other control signals are collectively collected in the main control unit 291.
- Software processing or hardware processing
- FIG. 6 is a diagram showing another example of the configuration of the conventional CSR relay device.
- the IP processing unit 3 is connected via a link 334 with the IP processing. 3 1 is arranged.
- the signaling cell is sent to the cell data unit assembly (SAR: Segmentation and Reassembly) 33 33 via the signaling cell path 33 36. Supplied and reassembled into de-units.
- This data unit is processed for UNI signaling in a software processing unit 335 including a CPU via a signaling data unit route 339. The reverse is also true.
- ATM cells are transferred to the cell / data unit assembling section (SAR) via a cell route 337 other than signaling. 3) Supplied to 3 3 and reassembled into data unit.
- SAR cell / data unit assembling section
- the data is supplied to the software processing section 335 via a data transmission path 338 other than the signaling, and is processed for a socket transfer. . After such processing, the data is transferred again to the cell / data unit assembling section 3333, assembled into cells, returned to the ATM switch 3332, and transferred to the next destination.
- cell routes 33 6 and 33 7 pass through the physically same link, and the data unit routes 33 8 and 33 39 are physically the same socket.
- FIGS. 7A and 7B show an example in which the ATM I / F is used for CPU communication between the ATM switch 772 and the IP processing unit 771.
- the software of the IP processing unit 771 consists of a CSR port, 77 la, UDP (User Datagram Protocol) / IP (Internet Protocol) 1b, BSD power 1c, and ATM network. It consists of a line 1d, a serial line 1e, and a force.
- FIGS. 7C and 7D show an example in which an asynchronous port such as an Ethernet is used for the CPU communication of the ATM switch 772 and the IP processing section 771.
- the ATM driver 771d is used for connection with the ATM switch 772.
- the CPU of the IP processing unit 771 communicates with the CPU of the ATM switch 772, and the VPI (Virtual Path Identifier) of the ATM switch 772 for the cut-through. Instructs / VCI (Virtual Channel Identifier) setting and release.
- CSR uses conventional software processing. High-speed transmission using a hardware-based, high-speed packet transfer mode, that is, a hop-by-hop transfer mode and an ATM switch. Provide the packet transfer mode at the same time. When the CSR with this configuration is used, as described later in Fig.
- the operation is such that the processing of the IP processing unit 771 is separated during the communication, and the ATM switch 772 It is possible to increase the transfer speed by making a power-through with. Cut-through means that a shortcut is taken, that is, the input bucket is not always transferred to the router, that is, the IP processing unit 771, but is output to the output port only by the ATM switch 772. It will be relayed. Power through processing is not visible to the end terminal or the ATM switch of the relay. The end terminal sends an IP address to the adjacent CSR (IP router) in the same way as when using the existing IP router network. Just send the packet.
- IP router IP router
- the first packet is transferred by the hop-by-hop transfer mode, the packet is reconstructed in the CSR, and the packet Is analyzed.
- the CSR uses the bucket header information to forward the bucket to the next hop node (the one with the router is the destination host), and the one to the net to the router (router).
- Applications that are expected to have a long application duration.
- the protocol between the adjacent nodes (CSR protocol) is used. No ,.
- the cut-through transfer path is established by notifying the correspondence between the packet flow and the cut-through transfer path (ATM-VC). Cut-through transfer path The route establishment is determined (triggered) using the information of the TCPZ UDP port number or the bucket length.
- the bucket is not reconstructed in the CSR and the ATM header information VPI / VCI, etc., is not rebuilt. It is switched using. If a decrease in the bucket toe volume of the application is detected, the CSR opens the cut-through transfer path using the CSR protocol as shown in Fig. 8 (d). I do . After the cut-through transfer path is released, the mode transits to the hop-by-hop transfer mode ((a) in Fig. 8).
- the CSR protocol is a protocol valid only between adjacent nodes, and the information / status exchanged by the CSR protocol is valid only between adjacent nodes. . Therefore, it is sufficient for the information / status regarding the power-through bucket transfer to be synchronized only between adjacent nodes, and global synchronization between all nodes is required. do not need. Perform regular information / status refreshes
- the policy of establishing a cut-through transfer path is the local decision of each CS.
- Fig. 9 is a diagram showing the operating principle of the above CSR using RSVP (Resource Reservation Protocol) as a trigger.
- CSR operates as a normal IP router that performs processing by IP processing.
- the default virtual Channel VC is used.
- the RSVP request message from the data receiving side (CSR-3) is sent. When it is received, this is triggered, and CSR-2 sets CSR-3 and a new channel VC (VPI / VCI). Similarly, a new channel flows between CSR-1 and CSR-2. After exchanging each other's VPI / VCI with the protocol between CSRs, the data flows using a new channel.
- RSVP Resource Reservation Protocol
- the flow-driven cut-through path is a path generated by transferring a control packet such as RSVP to the network side.
- the method of setting up a control driven pass and the information of the control bucket are used. It is necessary to operate the path that is generated without any means, that is, the data-driven path in dual mode.
- the method of generating a network that is, the operation of only a flow-driven type, is not applicable to a large-scale network.
- topology drive a method in which the network administrator prepares a power-through path with a different QOS class in advance.
- topology driven buckets transferred to subnets having the same address can be shared by the same cut-through path, that is, merging can be performed. It is. In other words, cut thruno. Since topology generation can be performed based on network topology information, this topology driven greatly reduces the number of VCs required for packet transfer by cut-through. It can be improved.
- the amount of bandwidth provided between the nodes is the amount of traffic (data toughness). Need to be in the band required when the traffic is relatively high. In other words, it is necessary for the network administrator to prepare a single cut-through path having a different QoS class in advance, which wastes bandwidth.
- FIG. 10A shows the operation of the above-described load driven
- FIG. 10B shows the operation of the above-described topology drive
- FIG. 10C shows a combination of both modes. The dual mode is shown.
- the flow-driven is applied simultaneously to the topology-driven, it is possible to secure the bandwidth on demand for the bucket flow that requires quality assurance. .
- the bandwidth used in the topology drive can be reduced as compared with the case where the bandwidth is applied only in the topology drive.
- a plurality of applications share a communication resource, that is, a bucket flow that shares a VC, or a topology driven path, and an application. Since it is possible that there is a bucket flow that requires a communication resource, that is, a VC, or a bucket flow that requires a different communication quality class at the same time, In order to satisfy such demands in the operation of only the drive, it is necessary to provide different communication quality classes in the topology drive. A dual mode that can simplify the hardware function by using different cut-through paths for each required communication quality class must be able to operate.
- Topologically driven CSR networks can be realized either on a VP basis or on a VC basis.
- the maximum number of paths that can be set from one node is 256 (2 to the power of 8), which makes the network poorly expandable.
- the bucket input to the ATM repeater performs hop-by-hop transfer based on a condition for determining whether to establish a cut-through transfer path called a cut-through trigger. It is determined whether or not to perform a cut-through transfer.
- the cut-through trigger refers to the TCP port number of the first packet of each flow (hereinafter referred to as the trigger packet). Cut-through transfer is considered to be particularly effective in applications where the session duration is determined to be long (eg, te1net, ftp, http).
- the hop-by-hop transfer processing is performed by software, and (b) the hop-by-hop transfer processing is performed by hardware.
- a method of a conventional relay device will be described.
- FIG. 9 is a block diagram showing a schematic configuration of a conventional relay device when performing the above. In the figure, 6 7 1-:! ⁇ 6 7
- n 1 — n is the input line
- 672 is the ATM switch
- 673 is the switch Training table
- 6 7 4 1:! 6674 — n indicates output line
- 675 indicates SAR
- 676 indicates processor
- 677 indicates frame memory.
- the first packet of the flow to be routed by the repeater is the input line 6 7 1 —:! 6 67 1 — Input to ATM switch 672 from (n-1). In fact, the input to the ATM switch 672.
- the packet is converted into an ATM cell.
- One or is, a plurality of cells is input to the ATM sweep rate Tutsi 6 7 2 in Nono 0 Ke-bit basis.
- the ATM switch 672 refers to the switching table 673 to perform switching processing on the packet (actually, the ATM cell) by hardware.
- the setting is such that the hop-by-hop transfer or the push-through transfer is not determined for the flown socket, which is output to the switching table 673 and output line 674-n. Has been done.
- the bucket output to the output line 674 1 n is input to the SAR 675. Since the bucket is divided into ATM cells and input, the original bucket is assembled on the frame memory 677 so that the processor 676 can process the SAR 675. Is received. In other words, how much data is used in a layered application? Because the data length cannot be determined by looking at only one cell, each cell is collected in bucket units. In this way, the processor 6776 can be processed.
- the SAR 675 arrives at the processor 676 with a bucket requesting the establishment of the connection. This is notified together with the address of the frame memory 677 in which the bucket is stored, and the like.
- the processor 677 which has received the notification refers to the TCP port number of the bucket, and performs a hop-by-hop transfer or a power-through of the flow through which the packet flows. Determine whether to transfer.
- the processor 676 stores the switching table 673 in the switching table 673 only by the switching process using the ATM switch 672. Then, the setting is made so that the bucket flowing through the flow is switched. When transferring the flow in a hop-by-hop manner, the switching table 673 is not changed, and the socket that flows through the flow is transferred to the processor 676 after that. Process with soft software.
- a cut-sleeve packet a packet to be transferred by hop-by-hop (hereinafter referred to as a hop-by-hop packet) in the relay device? How it is processed will be described.
- Cut through the ATM switch 672 When a packet is input, an output line is determined according to the switching table 673, and cut to the determined output line 674-1-1 to 674-(n-1). A through packet is output.
- the setting for switching to the output line 674-n is made to the switching table 673, and is performed.
- the packets output to the output line 674 — n of the ATM switch 672 are transferred to the SAR 675.
- the processor A TM Sense is the original note on the frame memory 6 7 7. Stands in the knot. No ,.
- the SAR 675 informs the processor 676 that the hop-by hop bucket has arrived. Notify together with the address of the frame memory 677 where the packet is stored.
- the notified processor 676 sends the bucket header.
- the processor 676 sends a message to the SAR 675.
- Ket; ⁇ Notifies that it exists on the frame memory 677.
- the SAR 675 fetches the packet to be transmitted from the frame memory 677, divides it into ATM cells, and then inputs the packet to the ATM switch 672. To enter.
- the bucket input from the input line 671_n is output to the output line 674-;;! To 674-(n-1) determined according to the switching table 673-. Is output.
- Pro Se Tsu Sa 6 7 6 comprises ho Ppuno I ho Ppuno ⁇ 0 Ke Tsu DOO, Application Benefits Ganoderma ⁇ 0 Ke dot and Le one tee ring to prodrugs co le information Regardless of the type of signaling cell used to establish a packet or ATM connection, the SAR 67
- FIG. 3 is a block diagram showing a configuration.
- 6 8 1 in the figure ⁇ 68 1 — n is an input line
- 68 2 is an ATM switch
- 68 3 is a switching table
- 6 684 — n indicates an output line
- 686 indicates a SAR
- 686 indicates a processor
- 687 indicates a frame memory
- 688 indicates a routing processing unit.
- the difference from FIG. 11 is that the hop-by-hop bucket, which is transferred from the output line 684—n to the SAR 685 and assembled on the frame memory 687, One routing process is performed in hardware by the routing processor 688.
- the routing processing unit 688 sends the header of the packet to the frame memory 687.
- the routing process is performed based on the information (sender address, destination address, etc.).
- the SAR 685 is notified that the bucket to be transmitted exists on the frame memory 687.
- the SAR 680 divides the bucket into ATM cells and inputs the packet to the ATM switch 682 from the input line 6811-n.
- the routing processing unit 688 by configuring the routing processing unit 688 with a nodule, high-speed routing processing can be realized.
- the bucket to be processed by the processor 6886 is a trigger bucket and a route for determining whether or not to perform a cut-through transfer. No., including protocol information. Only packets destined for the relevant relay device, such as packets and signaling cells used to establish ATM connections.
- the processor 6886 and the routing processing unit 688 in FIG. 12 are replaced by a bucket similarly to the processor 6776 in FIG. Regardless of the type of packet, the packet is processed and reviewed in the order in which the bucket arrival notification was received from SAR 686.
- connection orientated communication CO: a method of confirming the connection with the other party and then performing communication like a normal telephone
- ATM switches can accommodate connectionless connection (CL: a method of suddenly communicating without confirming the connection with the other party, such as a telegram). It is.
- Figures 13 and 14 show the conventional method.
- Fig. 13 is a block diagram showing the schematic configuration of an nxn ATM switch
- Fig. 14 is a block diagram showing the schematic configuration of the incoming ATM layer processing unit in the ATM switch. It is. Note that n X n means n inputs and n outputs.
- n ports 540— :! ⁇ 5 4 0 — n is connected.
- Cells input from each line are port 540 — :! First, the signals are input to the input-side ATM layer processing section 5 4 2 — 1 to 5 4 2 — n.
- the entry-side ATM layer processing section 5 4 2 —;! To 5 4 2 — n counts the number of passing cells for each connection, and passes the counting directly to the entry-side passing cell.
- Input side conversion table 5 4 3 —:! 5 5 4 3 — n and the entry side Number of passing passages 5 5 4 4 — n does not need to be physically separated, and may be physically configured as a single table.
- Cell is input side conversion table 5 4 3 —:! 5 5 4 3 — Transferred to ATM switch core 545 together with information obtained in n.
- the desired outgoing ATM layer processing section 546-6-1 is performed in accordance with the output destination obtained by the incoming ATM layer processing section 542-2—1 to 542—n.
- 5 5 4 6 Transfer cells to n.
- Ingress ATM layer processing unit 5 4 2 1 to 5 4 2-n
- outgoing ATM layer processing unit 5 4 6 :! ⁇ 5 4 6 — n
- the number of passing passages for each connection is counted, and the count directly is stored in the outgoing passage cell number counter table 548-1-1 to 548-n.
- the input ATM layer processing section 5 4 2 —:! ⁇ 5 4 2 — n has the configuration shown in Fig. 14.
- a cell input from the line 1 is input to the input processing unit 551.
- the input processing unit 551 extracts the input side conversion table 5431-1 and the number of cells passing through the input side 54.41-1 from the cell, and extracts VPIZVCI, etc., which is a key for searching the cell.
- the VPIZVCI is transferred to the conversion table control unit 553 and the passing cell count table control unit 554, and the senor is stored in the cell buffer 552.
- the conversion table control unit 553 searches the input side conversion table 543-1 based on the VPI / VCI extracted from the cell, and outputs information such as the output destination of the cell and the connection identifier. obtain.
- the passing cell counter table controller 554 refers to the input side passing cell counter table 544-1 based on the extracted VPI / VCI, and the input cell is normally received. If so, the value obtained by adding 1 to the referenced counter value is written back to the entry-side passing cell number counter table 5 4 4 — 1.
- the output processing section 555 it is transferred from the Serno buffer 552.
- the output destination of the cell and the information such as the connection identifier transferred from the conversion table control section 553 are combined and output to the ATM switch core 545.
- processor IZF unit 556 controls access from the processor 541.
- Processor I / F from processor 5 4 1 to access side conversion table 5 4 3 1 1 or entry side transit cell count counter 5 4 4-1 This is performed via the section 556 and the conversion table control section 553 or the passing cell number counter table control section 554.
- the counter clear function described later is realized, it is implemented in the passing cell counter table controller 554.
- the processor 541 which controls the entire ATM switch, requires the number of passing passages on all the connections to be managed. :! 5 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4.
- the processor 541 determines that the connection is not in use and performs processing to release the connection.
- n is used to count the number of passing cells when reading from the processor 54 1.
- the entry-side ATM layer processing section 542-1 to 542-n has a counter-document clearing function (when the processor 541 is read, the same as the previous time). In the case of the same count value, it is cleared to 0), the number of cells passing through the counter on the incoming side 5 4 4 — 1 to 5 4 4 — n Assumes the case of polling n .
- a description will be given of a case where ATM communication is performed using the ATM relay device as described above.
- data is transmitted and received in a fixed-length bucket of 53 bytes, that is, cells, but the VPI field is a UNI (User-Network Interface) in a 5-note header. 8 bits, 12 bits are secured by NNI (Network Node Interface), and 16 bits are secured in the VPI field, respectively.
- Such VCI / VPI identifiers are assigned to each physical line unit when communication between End and End (between terminals) is started, and after the allocation is completed, the communication starts from the start point of communication. While the cell header is rewritten to the header value of the next physical link, cells are sent following the assigned identifier.
- IP network used in the Internet and the like
- communication between End and End is performed using a number that identifies each place, which is called an IP address.
- U As a method of assigning this IP address, the same network address is assigned to each of networks of a certain size, and the inside of one network is further increased.
- the terminals are divided into groups to a certain extent, grouped as subnets, and each terminal is assigned a position within this subnet. This makes it easier to manage the entire network, and the network exchange between End and End is performed efficiently.
- IP networks and subnets are divided by dividing the IP address by the number of bits from the most significant bits.
- the subnet mask is called a subnet mask. Clarify by assigning the bits that filter the addresses to each subnet.
- a router or a host is generally arranged, and the header of the received bucket is generally provided.
- the specified path is set according to the destination address of the destination, and it is transferred to the next destination.
- pNNI private network network interface is one of the protocol sets for setting up connections between ATM switches.
- hierarchical source routing is performed by forming a hierarchical Peer Group by combining multiple nodes.
- routing and connection settings are performed independently, and the connection depends on the request of the host. Set by the end.
- there are other services such as Classical IP over II and NHRP (Next Ho Resolution Protocol), and Classical IP over ATM LIS (Logical
- An ARP (Address Resolution Protocol) server is placed for each IP Subnet, and when a connection is set up, the server is accessed to obtain address information.
- ARP Address Resolution Protocol
- NHRP NHRP server
- an NHRP server is set for each LIS, and this NHRP server exchanges information between servers.
- the ATM repeater and the network including the ATM repeater have been described above.
- the conventional ATM repeater described above has various problems as described below.
- a first object of the present invention is to disperse the load of the IP forwarding unit so as to increase the operation speed and processing capacity of the IP forwarding function without increasing the speed and processing speed.
- ATM relay device capable of relaying buckets with large capacity
- the signaling processing load on the IP processing unit that sets the signaling processing of the ATM switch increases.
- the communication path between the IP processing unit and the ATM switch becomes more occupied by the signaling processing ATM cells, the bandwidth for hop-by-hop transfer of the original IP bucket is increased. I will make it thin.
- a protocol for managing and controlling the state of the ATM relay device such as SNMP (Simple Network Management Protocol) is indispensable for the communication device. Therefore, the processing load for managing and controlling the ATM switch and the IP processing unit also increases. In other words, the data for management and control is also mapped to the ATM switch and transferred between the ATM switch and the IP processing unit, so that the transfer bandwidth of the IP bucket is increased. I will make it thin.
- SNMP Simple Network Management Protocol
- the IP switch, the signaling processing, and the management / control communication are performed by concentrating the ATM switch and the IP processing unit with one physical path.
- the software processing load on the relay device is concentrated, the transfer bandwidth of the IP packet itself is narrowed, and the efficiency of the IP packet relay process cannot be improved.
- the bucket traffic will increase, so the traffic of hop-by-hop transfer will increase and the communication traffic will increase. Is expected to be a torneck You.
- a second object of the present invention is to reduce software processing of each function of IP forwarding processing, signaling processing, management, and control. 2, IP ,.
- IP IP
- the efficiency of the packet forwarding, signaling, and management and control processes can be improved, and the operating speed and processing speed (CPU performance) can be reduced.
- An object of the present invention is to provide an ATM repeater capable of performing high-speed and large-capacity data relay processing without raising the cost.
- the conventional ATM repeater described above could not meet these requirements. Further, in the conventional ATM relay device, the traffic that does not perform the cut-through transfer by the CSR method, that is, the transfer capability of the packet that is transferred by hop-by-hop is controlled by the software. A) The processing capacity is considerably reduced because the processing is performed by (processing by the CPU).
- the third object of the present invention is to It is possible to obtain high-speed signaling performance for SVC control while avoiding an increase in signaling traffic when supporting C, and to provide hop-by-hop performance.
- An object of the present invention is to provide an ATM relay device and a router device capable of improving the packet transfer capability at the time of transfer. Also, in the network bucket relay device using the CSR method, especially for the own station. Kets, non-IP nodes. In comparison with the packet processing, the transfer capability of the bucket during hop-by-hop transfer is greatly improved. Furthermore, in the above-mentioned packet addressed to the own station and the unknown destination IP packet processing, the packet addressed to the own station can be reliably received without performing TTL subtraction more than necessary. Improves the transfer distance of the bucket when replaced with a node.
- reference numeral 421 in the figure denotes a public line, for example, a public line such as an NTT net.
- 4 2 2, 4 23 is a GW (Gateway), which interconnects systems / networks of different protocols.
- two GWs 4 2 4 and 4 2 5 are installed, and the backbone network (backbone network) 4 2 4 and 4 2 5 are connected to them.
- IP — Subnet (subnet) 4 26 to 429 constitutes one network peak.
- GWs 424 and 425 correspond to the gateway of this network, and the networks are connected by a dedicated line service of the carrier.
- VCC Virtual Channel Connection
- V C C logical path
- V P I Z V C I logical path
- VPIZVCI in a logical path is normally allocated to one between physical links, but is simplified in the figure because it is not essential in the present invention.
- VP virtual path
- each IP—S as shown in Figure 16
- the VCC power used in ubnets 4 26 to 4 29 can also be allocated to SB ackbonenets 4 2 4 and 4 2 5 at the same time.
- IPs are allocated by the number of terminals in the subnets 4 26 to 4 29, so that the backbone networks 4 2 4 and 4 2 5 and the GW 4 2 2 and 4 2 3 Many VCCs will be concentrated.
- the first packet passes first, and the next bucket Control is required, and the data of one packet is long, or if the cells that make up the bucket are widely separated, the first The problem was that it took a long time to pass through the packet, and the second packet could not be transferred without enough power.
- a fourth object of the present invention is to combine a plurality of VCCs into one, and particularly to provide a VPC in a backbone network.
- ATM switch that reduces the amount of IVCI resources and implements a VCC merging method that does not hinder packet transfer. It is to provide Tsuchi.
- the connection is made by polling all the connections by the processor using the number of cells passing through each connection. Releasing the session was being performed.
- connection is released in this way, the time elapsed since the change to the current counter value until the processor polled. Since the time is not known, there is a problem in that the release of the connection is delayed due to the difference between the cell passing timing and the polling timing.
- the processor determines that the connection is in use because the counter value is not 0 in the next polling for the passing cell number counter, and the connection is used. Does not perform the shot release processing. In other words, the connection can be released only after the next poll.
- the processor controls not only the polling and connection release processing of the counter for the number of passing cells, but also the entire ATM switch.Therefore, when the number of connections to be managed is large. In this case, the polling period for the number of cells passing through a specific connection becomes longer (the interval becomes longer so that the count value of the number of cells passing through cannot be checked. ). For this reason, although the connection is not actually used, the number of cells passing through the specific connection is longer than the number of cells passing through the counter in proportion to the polling period of the counter. Unable to release the connection. In other words, a problem occurs in which the connection resources cannot be used effectively.
- a fifth object of the present invention is to enable early release of unnecessary connections without requiring complicated control.
- Another object of the present invention is to provide an ATM switch that can effectively utilize connection resources.
- the ATM switch 672 and SAR 675 are connected via one input / output line of the ATM switch 672.
- packets to be processed by the processor 6776 are transferred to the line in a state of being converted into ATM cells.
- the ATM switch 672 and the SAR 675 have input lines
- the cells are output in almost the same order as that input to the ATM switch 672. Also, if the state where cells destined for the SAR 675 are input to the ATM switch 672 beyond the bandwidth of the line connecting the ATM switch 672 and the SAR 675 (congestion) continues. It is conceivable that cell discarding occurs inside the TM switch 672.
- each type of cell is included in the frame memory 677.
- Each queue type has a buffer queue.
- a trigger bucket is given a higher priority than other buckets, and the processor is processed with priority.
- cell discarding occurs regardless of the type of cell in the ATM switch 672, and therefore, when viewed as a whole device, a trigger packet is necessarily generated. It cannot be said that the treatment is given priority.
- a trigger packet a bucket containing routing protocol information, a signaling cell, etc. Buckets with high importance could be discarded depending on the cell input status.
- a sixth object of the present invention is to realize priority control according to the type of packet at the preceding stage of the ATM switch, and to give priority to packets with high priority.
- An object of the present invention is to provide a relay device capable of transmitting data for efficient data communication.
- the above-mentioned VC merging method has the following problems. That is, when the cells (A1 to A4) and (B1 to B4) along the two flows are simply merged using the VC merging function as shown in FIG. After switching by the switching function 3, the cells flow in a mixed manner as shown in the figure. The beginning cell of the packet, the BOC (Begin of Cell), arrived twice, and as a result, the PDU (Protocol Data Unit) could not be properly threaded. .
- a seventh object of the present invention is to selectively use a power-through path for each class of communication quality by merging cells in units of PDUs according to the communication quality.
- An object of the present invention is to provide an ATM repeater having a dual mode capable of performing the following.
- the basis for determining whether or not to perform cut-through transfer is the cut-through trigger information. According to the current specifications, this cut-through is required. Must be set. However, when the administrator sets the cut-through trigger information according to the user's request, the request truly reduces the load on the router and is effective for the network. It is difficult to judge that something is wrong.
- cut-through trigger information After setting the cut-through trigger information, confirming whether it is working effectively can be a heavy burden for the administrator. For example, if cut-through is set for transferring a small number of packets, the delay caused by the process of moving to cut-through mode may be larger. So
- an eighth object of the present invention is to enable the setting of an effective cut-through path by monitoring the state of bucket transfer at a router, thereby improving the efficiency. It is an object of the present invention to provide a bucket transfer state monitoring device capable of realizing an efficient network operation.
- a router or host using the above-mentioned conventional label switch technology can be used as a router.
- a bucket with a QOS request that does not allow the transfer time delay because the specified path is set or not set regardless of the QOS (Quality of Service) information of the packet. Buckets that are not Nevertheless, there was a problem that the packet did not arrive within a predetermined time or the bucket was discarded.
- An object of the present invention is to provide a method of setting a designated path that can guarantee packet QoS information and realize packet transfer in consideration of a difference in service quality.
- the connection between the switches can be made using pNNI.
- information such as link capacity, link hand propagation delay, usable bandwidth, guaranteed cell loss rate, number of connections passed, and link cost used is used. Notification is required to perform resource management. However, the quality of service (QOS) is assured at the level of the conventional LAN. Only ATM resource management is required, and large capacity ATM lines and switching speeds are required. In some cases, the routing table has a lot of unnecessary information.
- pNNI performs source routing using a link state algorithm between each Peer Group and each Peer Grou, and the switch uses each topology data for each switch. You must create a data-intensive DTL (routing list).
- a tenth object of the present invention is to provide a connection setting method in an ATM network that can simultaneously realize high-speed connection setting and reduction of the number of connections. To do that. Disclosure of the invention
- the ATM relay device distributes an input ATM cell to a corresponding output port among a plurality of output ports based on the destination information.
- ATM switch cores each of which is provided corresponding to each output port of this ATM switch core, each of which has its own forging processing function, and is connected to each output port through each output port.
- Destination information providing means for providing destination information, and ATM cells of a packet to which the same destination information has been added by the destination information providing means are stored by a predetermined amount, and the predetermined amount is stored.
- Output means for outputting the ATM cells in a stored form.
- the ATM relay device provides at least first, second, and third different ports for the input ATM cell based on the destination information.
- An ATM switch distributed to one of the ATM switches, and destination information of the ATM cell is forwarded through a first path connected to a first port of the ATM switch.
- Receiving an ATM cell indicating that the IP forwarding means for performing a packet forwarding process, and a second path physically different from the first path connected to the second port of the ATM switch Signaling processing means for receiving, via a path, an ATM cell indicating that the destination information of the ATM cell is for signaling processing, and performing signaling processing of an IP packet.
- destination information of the ATM cell is managed and controlled via a third path physically different from the first and second paths connected to the third port of the ATM switch.
- Management and control means for receiving and managing and controlling the ATM cell indicating that the ATM cell is used.
- the present invention relates to an ATM repeater having a packet transfer function of a network layer using an ATM switch.
- the ring processing means and the signal processing means are provided separately, and are connected via a second link physically different from the first link.
- a forcing processing means for receiving the packet and performing IP packet transfer processing in hardware using predetermined memory.
- the present invention relates to a router device provided in connection with an ATM switch and having an IP bucket transfer function in a network layer, wherein the ATM switch is connected via a first link.
- Signaling processing is performed by receiving a signaling bucket from a switch.
- Signaling processing means is provided separately from the signaling processing means, and the first link is physically connected to the first link.
- a forwarding processing means for receiving an IP packet via a different second link and performing IP packet transfer processing in hardware using predetermined memory. Have.
- the ATM switch searches a routing table based on header information of a cell input from an incoming line, and searches the routing table for the input cell.
- Mask means for masking a part of the header information of the input cell in an ATM switch for rewriting the header information of the input cell into information on the outgoing side and exchanging it for a predetermined output line;
- a table search means for searching for the routing table based on the obtained header information; and header information of the input cell specified in the routing table based on a search result by the table search means.
- information rewriting means for rewriting the information on the outgoing side.
- the ATM switch searches for a routing table based on the VPI / VCI described in the header information of the cell input from the incoming line, and searches for the header of the input cell.
- D Rewrite the information with the information on the outgoing side and exchange it for a predetermined output line.
- a mask bit In the TM switch, a mask bit.
- Mask pattern storage means for storing turns, and mask bit numbers stored in the mask pattern storage means.
- the mask means for masking a part of the VPI / VCI bit example described in the header information of the input cell using the turn, and the mask is performed by the mask means.
- a table search means for searching the routing table based on VPI / VCI, and the header information of the input cell is determined in the routing table based on a search result by the table search means.
- information rewriting means for rewriting the information on the exit side.
- the ATM switch comprises a cell non-conduction time detecting means for periodically detecting a non-conduction time for each connection.
- a storage means for storing the non-conduction time of the cell detected by the cell non-conduction time detection means; and a non-conduction time of the cell stored in the storage means at a predetermined polling period.
- a connection releasing means for performing a connection releasing process based on a polling means to be polled and a non-conduction time of a cell polled by the polling means. Means.
- the relay device uses an ATM switch in a network layer.
- the repeater that performs the routing processing and the switching processing in the data link layer, before performing the above-described routing processing and the switching processing, the input data is handled in buckets, and the packet is processed.
- Priority control means is provided for discriminating a packet type and transmitting each packet to the ATM switch according to a priority according to the packet type.
- the ATM relay device includes a high-speed bucket transfer mode based on hardware using an ATM switch, and the same transfer.
- an ATM repeater that transfers the same path by sharing the same path, and a packet having the same destination address may be used for the communication quality of each bucket.
- the ATM relay device has a high-speed bucket transfer mode using hardware using an ATM switch, and the same transfer destination address is used for buckets having the same transfer destination address.
- An ATM repeater that shares and forwards packets, disassembling means for decomposing each bucket having the same destination address for each cell, and disassembling the decomposed cells according to the communication quality. Assembling means for assembling the packets in units of PDUs, and transfer means for assembling the assembled PDU-unit buckets in order of communication quality level and transferring the packets using a single path. Is provided.
- the packet transfer status monitoring device includes a monitoring unit that monitors the transfer status of a bucket transferred to a destination by performing a routing process in a network layer, and a monitoring unit that monitors the packet transfer status.
- Prediction means for predicting the tendency of packet transfer based on the monitoring results by the means, and hardware processing at the data link layer and below based on the results predicted by the prediction means.
- a judgment means for judging conditions for shifting to.
- This packet transfer state monitoring device includes monitoring means for monitoring the transfer status of a packet transferred by hardware processing below the data link layer, and monitoring means for the monitoring means. Determining means for determining whether or not the above-described hardware processing is being performed effectively based on the monitoring result obtained by the above-mentioned hardware processing; and And release resetting means for releasing or resetting the trigger information for instructing.
- a method for setting a designated path is a method for setting a specified path, which includes a network configured by a label switch and a network configured by another device.
- a network consisting of a router or label switch host or a label switch located at the boundary with the work
- the same label switch network and another label switch network can be used within a network composed of label switches.
- the method of setting this designated path may be a router or a label switch located at the boundary between a network composed of label switches and a network composed of other devices. Even in a network consisting of a switch host or a label switch, the method of setting a designated path in a router located at the boundary between the same label switch network and another label switch network When the specified path is already set for a specific route, the process of referring to the QOS information or similar information in the header of the data bucket to be transferred, and the referenced QOS information or similar information Deciding whether to set a new designated path or to use the already designated path based on Comprising a.
- a connection setting method in an ATM network is provided for each node of the ATM network.
- the ATM switch has an IP over ATM function, and when setting up a connection with such an ATM switch, a route is established between adjacent ATM switches and ATM terminals.
- a step of performing routing by a protocol is provided for each node of the ATM network.
- connection setting method in an ATM network generates a routing table using the routing protocol operating on IP in IP over ATM. Process and this routing table And a step of setting a connection based on the entry.
- FIG. 1 is a diagram illustrating a configuration of a conventional ATM relay device.
- FIG. 2 is a diagram for explaining bucket data route control using a network layer.
- FIG. 3 is a diagram for explaining high-speed transfer of bucket data using a data link layer.
- FIGS. 4A and 4B are diagrams for explaining the CSR cell transfer method.
- FIG. 5 is a diagram showing a configuration of a conventional ATM relay device.
- FIG. 6 is a diagram showing another configuration of the conventional CSR relay apparatus.
- 7A to 7D are diagrams showing an example of the configuration of the CSR.
- FIG. 8 is a diagram for explaining the basic operation of the CSR in detail.
- FIG. 9 is a diagram for explaining the operation of CSR with RSVP as a trigger.
- FIG. 10A, FIG. 10B, and FIG. 10C are explanatory diagrams of the topology driven and the flow driven of the CSR and the dual mode that supports both.
- Fig. 11 shows a schematic diagram of a conventional repeater when an ATM switch is used as the switch and hop-by-hop transfer processing is performed by software.
- FIG. 11 shows a schematic diagram of a conventional repeater when an ATM switch is used as the switch and hop-by-hop transfer processing is performed by software.
- FIG. 9 is a block diagram showing a schematic configuration of a conventional relay device when a transfer process is performed by hardware.
- FIG. 13 is a block diagram showing a schematic configuration of an n ⁇ n ATM switch.
- FIG. 14 is a block diagram showing a schematic configuration of an incoming ATM layer processing unit in the ATM switch.
- FIG. 15 is a diagram illustrating an example of a network configuration.
- FIG. 16 is a diagram showing an example of a configuration when a network is constructed by ATM.
- FIG. 17 is a diagram for explaining a problem of the conventional VC merging method.
- FIG. 18A is a diagram showing the configuration on the input side of the ATM repeater according to the first embodiment of the present invention
- FIG. 18B is a diagram showing the configuration on the output side of the ATM repeater. .
- Figure 19 shows the IP ,. It is a figure which shows the state of distributed processing of a packet.
- FIG. 20 is a diagram for explaining the problem of mixing PDU data by the distributed processing according to the present embodiment.
- FIG. 21 is a diagram for explaining a method for avoiding the problem of PDU mixing due to the distributed processing according to the present embodiment.
- FIG. 22 is a block diagram showing the configuration of the VC merge unit shown in FIG.
- FIG. 23 is a diagram illustrating an example of a method of determining that cell data input to the Per VC queuing unit has formed a PDU.
- FIG. 24 is a diagram showing the configuration of the ATM relay device according to the second embodiment of the present invention.
- Figure 25 is a diagram showing the distribution at the ATM cell level.
- FIG. 26 is a diagram illustrating a modification of the ATM relay device illustrated in FIG. 24.
- FIG. 27A and FIG. 27B are diagrams for explaining the basic principle of the Senoless Switcher (CSR).
- CSR Senoless Switcher
- FIG. 28 is a diagram for comparing and explaining the use of the bandwidth of the physical path by the conventional IP relay apparatus and the use of the bandwidth in the present embodiment.
- FIG. 29 is a diagram illustrating a schematic configuration of a CSR relay device to which the ATM relay device according to the third embodiment of the present invention is applied.
- FIG. 30 is a diagram illustrating a modification of the CSR relay apparatus illustrated in FIG. 29.
- FIG. 31 is a diagram illustrating a schematic configuration of a CSR relay device to which an ATM relay device according to a modification of the third embodiment of the present invention is applied.
- FIG. 32 is a diagram showing a modification of the ATM relay device shown in FIG. 31.
- FIG. 33 is a diagram illustrating a packet flow and a TTL operation according to the schematic configuration of the ATM relay device illustrated in FIG. 32.
- FIG. 34 is a diagram showing a flowchart of the TTL arithmetic processing applied to the IP forwarding section of the present embodiment.
- FIG. 35 shows the TT applied to the software processing unit of the present invention.
- FIG. 3 is a diagram showing a flowchart of an L operation process.
- FIG. 36 is a diagram showing a bucket flow and a TTL calculation according to the schematic configuration of the ATM relay device of the present embodiment.
- FIG. 37 is a block diagram showing a configuration of an ATM switch according to the fourth embodiment of the present invention.
- FIG. 38 is a flowchart showing a cell exchange algorithm in the fourth embodiment.
- FIG. 39 is a diagram for describing a cell multiplexing method according to the fourth embodiment.
- FIG. 40A is a diagram showing a configuration of a routing table in a table configuration according to a modification of the fourth embodiment of the present invention.
- FIG. 40B is a diagram showing a modification of the fourth embodiment of the present invention. According to the table configuration in the mask no.
- FIG. 3 is a diagram illustrating a configuration of a turn table.
- FIG. 41 is a flowchart showing a cell exchange algorithm in a modification of the fourth embodiment.
- FIG. 42 is a diagram illustrating a configuration example of an ATM network to which the method of the present invention is applied.
- FIG. 43 is a block diagram showing a schematic configuration of the nXnATM switch according to the fifth embodiment of the present invention.
- FIG. 44 is a block diagram showing a schematic configuration of an incoming ATM layer processing unit provided in the ATM switch of FIG. 43.
- FIGS. 45A and 45B are diagrams for explaining errors in the elapsed time of cells when the method of the present invention is applied.
- FIG. 46 is a schematic configuration of a relay device according to the sixth embodiment of the present invention.
- FIG. 3 is a block diagram showing a general configuration when an ATM switch is used as a switch and hop-by-hop transfer processing is performed by software.
- FIG. 47 is a block diagram showing a schematic configuration of the priority control unit used in the present relay device.
- FIG. 48 is a block diagram showing a schematic configuration when the queue management unit in the priority control unit is realized by the fixed priority method.
- FIG. 49 is a block diagram showing a schematic configuration when the queue management unit in the priority control unit is realized by a round robin method.
- FIG. 50 is a diagram showing a schematic configuration of a relay device according to a sixth embodiment of the present invention, in which an ATM switch is used as a switch, and hop-by-hop transfer processing is performed in a hardware manner.
- FIG. 4 is a block diagram showing a schematic configuration in the case of performing your operation.
- FIG. 51 is a block diagram illustrating a schematic configuration of a routing processing unit in the relay device.
- FIG. 52 is a view for explaining the basic concept of the seventh embodiment of the present invention.
- FIG. 53 is a diagram illustrating the configuration of the seventh embodiment.
- FIGS. 54A to 54E are views for explaining the operation of the seventh embodiment.
- FIG. 55A to FIG. 55E are views for explaining the operation of the modification of the seventh embodiment.
- FIG. 56 is a diagram showing a configuration of a modified example of the seventh embodiment.
- FIG. 57A to FIG. 57C are diagrams for explaining the operation of the present modification.
- FIGS. 58A to 58C are views for explaining the operation of another modification of the seventh embodiment.
- FIGS. 59A and 59B are diagrams for explaining the outline of the packet transfer state monitoring device according to the eighth embodiment of the present invention.
- FIG. 60 is a diagram illustrating an example of a network incorporating a CSR.
- Figure 61 shows the IP node during hop-to-hop transfer. It is a figure which shows the flow of a packet.
- FIG. 62 is a diagram showing the movement of CSR when a cut snoring trigger is applied.
- FIG. 63 is a diagram illustrating the movement of the CSR during the cut-through operation.
- FIG. 64 is a diagram illustrating a movement operation from the cut-through transfer force to the hop-no-hop transfer.
- FIG. 65 is a diagram illustrating an example of a result of monitoring the transfer state of the IP bucket in the CSR.
- FIG. 66 is a diagram showing a system configuration to which a method of setting a designated path according to the ninth embodiment of the present invention is applied.
- FIG. 67 is a diagram showing various examples of the QoS requirement.
- FIG. 68 is a diagram showing a state where data is not flowing through the line.
- FIG. 69 is a diagram showing a state when a less severely designated path is set for the QoS request in the state of FIG.
- FIG. 70 is a diagram showing a state when a designated path with strict QOS requirements is set in the state of FIG.
- Fig. 71 shows the case where a strictly QOS-requested packet is received while a specified path with a strictly QOS-request is set.
- FIG. 9 is a diagram showing a state in which a strict designated path is set.
- FIG. 72 shows the case where a designated path with a strict QOS request is received and a bucket with a strict QOS request is received while the designated path with a strict QOS request is set.
- FIG. 10 is a diagram showing a state where an unspecified path is set.
- FIG. 73 is a diagram illustrating a state in which a bucket with strict QoS requirements is received in a state with strict QoS requirements.
- FIG. 11 is a diagram showing a state in which two scars S and one designated path for which QoS requirements are not strict are set.
- FIG. 75 is a diagram illustrating a state in which a designated path for each QoS is set when the system starts up.
- FIG. 76 is a diagram illustrating an example of notifying the setting information of the designated path for each QoS from the user terminal.
- FIG. 77 is a diagram illustrating an example in which designated path setting information for each piece of QoS information is input from the maintenance terminal.
- FIG. 78 is a diagram showing a state in which a designated path for a request condition considered necessary at the time of startup is set, and then a new designated path is generated when a data transmission request occurs.
- FIG. 79 is a diagram showing a field in which QoS information is inserted in the header adopted in IPv4.
- FIG. 10 is a diagram showing fields into which information is inserted.
- FIG. 81 is a diagram showing a configuration of an ATM network to which the tenth embodiment of the present invention is applied.
- FIG. 82 is a diagram showing a protocol stack of the ATM switch and a flow of routing information in the present embodiment.
- FIG. 83 is a diagram showing a flow of signaling information at the time of setting SVC in the present embodiment.
- FIG. 84 is a diagram illustrating an example of a connection set in the present embodiment.
- FIG. 85 is a diagram showing a state of the VC merge in the present embodiment. BEST MODE FOR CARRYING OUT THE INVENTION
- FIG. 18A is a diagram showing the configuration on the input side of the ATM relay device according to the first embodiment of the present invention. It is assumed that this ATM relay device has a mechanism for searching for a destination IP address in hardware as an IP forwarding function.
- the configuration of FIG. 18A basically includes an ATM switch 110 and an IP forwarding unit 120 connected at a subsequent stage.
- the ATM switch 110 has a plurality of input ports, and an ATM switch core 11 for distributing cells to an output queue for each ATM cell destination (VPIZVCI), and a VC (node). It has queues # 1 to #N, and includes a Per VC queuing unit 112 that stores data to the queue corresponding to each VC at each output port. .
- the IP forwarding section 120 includes a plurality of IP forwarding sections # 1 (1200—1) to #N connected corresponding to the VC queues # 1 to #N. (120-N).
- Each IP forwarding unit for example, an IP forwarding unit # 1 (12.0-1) includes SAR121-1-1 for dividing and assembling an ATM cell, and an IP node. It is composed of an IP processing unit 122-1 that performs packet forwarding processing.
- FIG. 18B is a diagram showing the configuration of the output side of the ATM relay device.
- the ATM switch 110 is stored in each queue for the same destination, with the ATM switch core 111 and the PerVC queuing unit 112 being described above.
- the cell is read out while scheduling cells in PDU units, and has a VC merge section 113 for merging.
- the VC merge section 113 is connected to the output ports # 1 to #N of the ATM switch 110 and is connected.
- the packet is distributed by the ATM switch core 111 according to the destination information VPI / VCI.
- IP load corresponding to each VC is set in advance as load distribution processing.
- the packet is connected to each IP forwarding section.
- IP packet is reconstructed, and the IP address of the destination of the IP packet is searched in the IP processing unit 122. This search may be performed by a conventional software search or a table-lookup search using hardware.
- FIG. 19 shows the IP, described above. It is a figure which shows the state of distributed processing of a packet.
- the data for which the forwarding information has been obtained by the search in the IP processing sections 1 2 1-1 and 1 2-2-2 is output to the next relay device.
- mapping to the ATM sensor is performed again by the SARs 121-1-1 and 121-2-2, and then to the ATM switch 110. Entered.
- the IP forwarding processing is performed in a distributed manner by a plurality of IP forwarding units 120 — 1 to 120 — N. Therefore, the next output destination (NEXT HOP) of each IP bucket may have the same destination IP address.
- different PDUs protocol data units
- FIG. 20 is a diagram for explaining the problem of PDU data mixing due to distributed processing.
- NKT HOP next output destination of packet B
- IP sockets A and IP with VPI / VCI having the same destination # 110.
- Each of the packets B is converted into an ATM cell, and when input to the ATM switch 110, the PDU is mixed.
- the PDU—A, PDU—B cannot be restored properly.
- FIG. 21 is a diagram for explaining a method for avoiding the problem of PDU mixing by distributed processing according to the present embodiment.
- the IP sockets of the IP forwarding are treated as PDU-A and PDU-B, respectively, and are mapped.
- the above-mentioned "temporary VPI / VCI” is assigned to the ATM cell in which the destination information # 111 and # 112 are assembled, respectively.
- the next output destination (NEXT HOP) is # 119.
- the ATM cell input to the ATM switch 111 is switched to any output port according to its destination.
- the ATM cell is queued at the Per Vc queuing section 112 for each of these "temporary VPI / VCI" # 111 and # 112, so that In addition, PDU data (IP bucket data) can be independently formed and stored.
- the VC management unit 113 following the PerVc queuing unit 112 monitors the state of PerVC queuing. Encourage output of VC queue power.
- (1) means to read PDU-A
- (2) means to read PDU-B.
- FIG. 22 is a block diagram showing the configuration of this VC merge unit 113.
- the cell data input to the VC queuing unit 12 notifies the PDU determining unit 133 that the PDU has been formed.
- this determination is recognized as, for example, an AAL (ATM ADAPTATION LAYER) message by the end flag 150 of the PDU and notified.
- ATM ADAPTATION LAYER ATM ADAPTATION LAYER
- the ATM cell is read out by the scheduler 13 2 in units of PDUs from the VC queue in which the PDU is recognized to be formed by the PDU determination unit 13 3, and the data is transferred to the buffer 13 1. And stored.
- the scheduler 1332 controls the order of the PDU units, and performs the VC merge.
- the data in which the sequence of PDUs is retained by the buffer 13 1 is replaced by the VPI / VCI conversion section 13 4 with the VPI / VCI header corresponding to the next output destination (NEXT HOP). Will be transferred.
- a plurality of IP forwarding units are arranged at the subsequent stage of the ATM switch so as to perform distributed processing. Without increasing the number of dress table entries, it is possible to support faster and larger-capacity IP relay processing. Also, in combination with the ATM switch, load distribution can be realized by queuing for each PDU, that is, for each ATM connection, so that an increase in cost can be reduced.
- the operation speed and the processing capacity are not high. It is possible to support faster and larger-capacity IP relay processing without increasing the speed. Also, in combination with the ATM switch, load distribution can be realized by queuing for each PDU, that is, for each ATM connection, thus reducing the cost increase. And are possible.
- FIG. 24 is a diagram showing a configuration of a TTM relay apparatus according to the second embodiment of the present invention.
- the ATM relay device according to the present embodiment searches for a destination IP address, thereby obtaining an IP address. It has an IP forwarding function to transfer the packet hop-by-hop.
- an IP processing unit 200 is connected to the subsequent stage of the ATM switch 250.
- the IP processing unit 200 has IP forwarding processing function 202 for performing IP packet forwarding processing such as hop-by-hop transfer, and communication path setting.
- Signaling processing function 203 to perform the management and control of ATM switch 250.
- the SARs 205, 206, and 207 connected to each of the signaling processing function 203 and the management / control function 204 for assembling cells and data units are described above. It consists of a CPU 201 and a CPU that performs software processing of the data extracted by each processing function 202, 203, 204.
- the CPU 201 is shared between the processing functions 202, 203 and 204 by each bus. You.
- SAR 205 is the IP Forwarding No.
- SAR 206 is connected to port 1 of ATM switch 250 via switch A, and SAR 206 is connected to port 2 of ATM switch 250 via switch B.
- SAR 207 is connected to port 3 of ATM switch 250 via management path C.
- the above-described IP forwarding path A, signaling processing path B, and management and control path C are physically different paths.
- the ATM cells input to the ATM switch 250 at the input port are distributed to the output port by the ATM switch 220 for each destination (VPI / VCI). .
- the packet is delivered to the IP processing unit 200, the next address NEXTHOP is searched, and the output port is again passed through the ATM switch 20 according to the searched next address.
- the IP packet mapped to the ATM cell is sent to the IP processing unit 200 by the ATM switch 250 by the VPI / VCI which is the destination information. Be distributed. The distribution is made via the three paths A, B and C described above.
- the IP forwarding processing function 202 through the IP forwarding path A is used. Perform processing.
- the signal is processed by the signaling processing function 203 through the signaling processing path B. Performing the tuning process.
- management / control is performed by the management / control function 204 through the management / control path C.
- the IP forwarding processing function 202 corresponding to each VC (virtual connection) and the signaling processing function in advance are performed.
- 203 management.
- the control function 204 is determined independently, and the IP and the IP are controlled by routing for each VC. It distributes packet data, call setup data, and management and control data.
- FIG. 25 is a diagram showing such a distribution at the ATM cell level.
- ATM cells for signaling processing: VPIZVCI # 2
- ATM cells for management and control: VPI / VCI # 3
- IP forwarding processing function 202 As described above, in the present embodiment, by using a physically different communication path to connect the ATM switch 250 and the IP processing unit 200, the IP forwarding is achieved.
- IP forwarding processing function 202, signaling processing function 203, management and control function 204 provided separately for signaling processing, signaling processing, and management and control processing
- the above-described processes can be performed efficiently.
- the resources of CPU 201 are used independently, an improvement in processing performance due to the dispersion effect can be expected.
- each of the CPUs 201 is connected to the bus. Although shared, as shown in Figure 26, the CPU 201 is assigned to each of the IP forwarding dating function 202, the signaling processing function 203, and the management and control function 204. It is also possible to set them up as CPUs (CPU 210-1-1, 210-1-2, 210-1-3).
- the communication paths A, B, and C connecting the ATM switch 250 and the IP processing section 200 are not limited to ATM, and may use other communication methods.
- CSR cell switch router
- IP processing unit One physical node between the ATM switch and the IP processing unit. If you concentrate on the network, you can use the IP address forwarding, signaling process, management and control communication for the IP notch even if you use the CSR method. The load of the air processing is concentrated.
- FIGS. 27A and B show the basic principle of the cell switch router (CSR) in order to emphasize the effect of the present embodiment, as described in the background art described above.
- Figure 27A shows the normal IP forwarding operation. Hop-by-hop transfer in which the IP address is searched, the next address NEXTHOP is searched, and the transfer is performed It is. That is, the IP packet input from the CSR 1 (281) input is routed from the ATM cell to the IP processing unit 260 by the ATM switch 27 0 of the CSR 2 (280). Is performed. Then, in the IP processing unit 260, the IP address is set.
- next address NEXT ⁇ ⁇ ⁇ ⁇ is searched from the P address, and transferred to the next transfer destination (here, CSR 3) by the A ⁇ . ⁇ switch 270.
- next transfer destination here, CSR 3
- hop-by-hop transfer which is normal IP routing, is performed.
- transfer at the layer 2 level that is, cut-through transfer at the ATM switch, is performed.
- FIG. 27B is a diagram illustrating such a cut-through transfer.
- the IP processing unit 260 of CSR 2 (280) uses a cut-through between CSR 1 (281) and CSR 3 (282) according to the CSR protocol.
- the path is set to ATM switch 270.
- a "call" for setting a cut-through communication path to the IP processing unit 260 and the ATM switch 270 is generated.
- the load of IP forwarding is reduced, but on the other hand, a problem occurs in the signaling processing and the SIP processing unit 260, which causes a problem in the ATM switch 270. Both management and control are a burden on traffic.
- the signaling processing function since the processing is distributed for each support function, the signaling processing function does not affect the IP relay processing.
- FIG. 28 are diagrams for comparing and explaining the use of the bandwidth of the physical path by the conventional IP relay apparatus and the use of the bandwidth in the present embodiment.
- the IP address,. Packet forwarding, signaling processing, management and control are performed on a physical path concentrated in the communication between the IP processing unit 240 and the ATM switch 200.
- the transfer bandwidth of the packet was reduced, and the IP transfer bandwidth was reduced.
- the efficiency of the Ket relay process cannot be improved.
- the network speeds up and the capacity increases in the future no. Efficiency in bop-by-hop transfer becomes important again because of increased KET traffic. This kind of concentration.
- the configuration based on the network increases the speed and capacity of the network, and at the same time, the bandwidth of the physical path becomes insufficient, resulting in a low-expansion and low-relay system.
- the communication path is physically separated and the processing is performed for each function, so that the bandwidth can be spared and the physical Expansion of the configuration can be achieved by adding additional paths.
- the communication paths for the IP forwarding processing function, the signaling processing function, and the management / control function after the ATM switch are physically arranged separately. Since each processing is performed, the load of each function is distributed, so that it is possible to reduce software and software processing. Also, IP no.
- IP no The independence of the packet transfer bandwidth itself allows more bandwidth to be used, thus allowing for expandability, thereby increasing the IP bandwidth.
- the efficiency of the packet forwarding processing, signaling processing, management and control processing can be improved. In this way, the operating speed and processing speed (CP u no ,. For example, high-speed, large-capacity data relay processing can be performed without increasing the performance.
- load balancing can be realized by queuing for each ATM connection. It can also respond to IP relay processing when the speed is increased and the capacity is increased.
- the software processing of each function of the IP forwarding processing, the signaling processing, and the management and control can be reduced. . The efficiency of the forging, signaling, management and control processing of the packet can be improved, thereby increasing the operating speed and processing speed.
- FIG. 1 is a diagram showing a schematic configuration of a CSR relay device to which an ATM relay device according to a third embodiment of the present invention is applied.
- the ATM switch 342 is connected to the IP processing unit 361 via a link 363 with the IP processing for signaling, and is also connected to the signal processing unit 361. It is connected to the IP processing unit 361 via the link 344 with the IP processing for forming, which is physically different from the link 363 with the IP processing for ring. ing. In other words, foreground processing (here, hop-by-hop transfer) is performed. And the path that performs the signaling process are physically different.
- foreground processing here, hop-by-hop transfer
- the ATM switch 342 is connected via a signaling cell path 346 to a cell data unit assembly for signaling (hereinafter referred to as a SAR for signaling).
- the SAR 362 for signaling is connected to a software processing section 352 including a CPU via a signaling data route 396.
- the signaling cell path 346, the SAR for signaling 362, the signaling data path 396, and the software processing section 352 form the signaling processing means described above. I do.
- the ATM switch 342 is provided separately from the signaling SAR 362 via another cell path 347 that is physically different from the signaling processing path described above. Connected to the cell / data unit assembly (called hop-by-hop transfer SAR).
- the SAR 343 for hop-by-hop transfer is connected to an IP forwarding section 354 via a data unit path 397.
- the cell path 347, the hop-by-hop transfer SAR 343, the data unit path 379, and the IP forwarding section 354 constitute a forwarding processing means.
- a node address of an IP layer or a TCP / UDP layer is used as a node.
- ⁇ Detected by air, search for associative memory and no- IP no.
- high-speed bucket forwarding processing is performed without performing software processing by the CPU. Such a fording process is described in the specification of Japanese Patent Application No. 9-252121 of the present applicant.
- the operation of the CSR relay apparatus shown in FIG. 29 will be described below.
- the cells input to the ATM switch 324 of FIG. 29 are input to the IP processing section 361 via dedicated links according to their types.
- the signal is a signaling cell, it is transferred to the signaling SAR 362, and the other cells are transferred to the hop-by-hop transfer SAR 343, where they are transmitted to the respective SARs. It can be reassembled into more data cuts.
- the data reassembled by the signaling SAR 356 is not taken into the IP forwarding unit 354 but is sent to the software processing unit 352. Direct transfer is performed and software processing is performed by the CPU.
- the data unit reassembled by the SAR 343 for hop-by-hop transfer is transferred to the IP forwarding unit 354 to perform hop-by-hop transfer processing.
- the cell is reassembled by the SAR for hop-by-hop transfer 343 and transmitted to the ATM switch 342.
- the rewriting and management of the associative memory table is C It shall be performed by software processing by PU.
- the Sidana ring cell and the other cells are transferred using physically different links, and the respective processes are performed separately. Therefore, it is possible to obtain high-speed signaling performance for SVC control while avoiding an increase in signaling traffic in the case of supporting the SVC.
- the hop-by-hop transfer processing is not performed by transferring the data to the CPU of the software processing unit 352, but is performed by the IP forwarding unit 354. Since the transfer is performed by the hardware processing, the packet transfer capability at the time of hop-by-hop transfer can be greatly improved.
- FIG. 30 is a diagram showing a modification of the above embodiment.
- a plurality of IP processing units 361-1-3, 361-2, and 361-1-3 are connected to the ATM switch 342 and are connected.
- Each of the IP processing sections 361-1, 361-2, and 361-3 has the configuration shown in Fig. 29. According to such a configuration, since there are a plurality of configurations of the present embodiment as shown in FIG. 29, the above-described effects can be obtained synergistically.
- a signaling performance can be obtained in a network bucket relay device using the CSR method, particularly when the SVC is supported, and The packet transfer capability during up-by-hop transfer can be greatly improved.
- the SR repeater uses the inherent capabilities of CSR.
- IP off O Wadi ring section 3 5 4 and Nono over de Eai spoon Les, Ru force s IP of this off O Wadi ring section 3 5 4 Determines the IP packet including the protocol and addressed to its own station and the packet whose hop-by-hop transfer processing method is unknown, and sends the packet to the software processing unit 352.
- the software processing unit 352 has received the IP packet from the local station and processed the packet or the packet whose processing was unknown. Occurred, no ,. You must process to receive the packet.
- the address entry of the IP forwarding section 354 is updated. Is done. In particular, the above processing was unknown. In some packets, the IP address of the transfer destination is not entered in the IP forwarding section 354. When transmitting a packet, it is necessary to update the addressing of the IP forwarding unit 354 in many cases.
- the original capability of CSR can be improved by further improving the hop-by-hop transfer capability, and the structure of the device can be simplified. It is intended to provide an ATM repeater.
- FIG. 3 is a diagram showing a schematic configuration of a CSR relay device to which an ATM relay device according to a modification of the present invention is applied.
- the ATM switch 314 is connected to the IP processing unit 300 via a link 313 with the IP processing for signaling, and is also connected to the signal processing unit 314.
- Link to IP processing for ring processing 313 physically different from link 313 Connect to IP processing unit 300 via link 306 for hop-by-hop transfer IP processing It has been.
- IP processing for signaling and forwarding 313 and 306 IP processing for software processing (CPU processing) that is physically different from IP processing It is connected to the IP processing unit 300 via the link 303 of the system. In this way, the path for performing the forwarding process (here, the hop-by-hop transfer) and the signaling process are performed. This is physically different from the path for software and software processing.
- the ATM switch 314 is connected to a cell signaling assembly (hereinafter referred to as a signaling SAR) 311 via a signaling cell path 312.
- the signaling SAR 311 is connected to a software processing unit 310 including a CPU via a signaling data unit path 3222.
- the signaling cell path 312 described above, the SAR 311 for signaling, the signaling data unit path 3222, and the software processing section 310 constitute signal processing means. I do.
- the ATM switch 314 passes through another cell path (hop-by hop-cell path) 305 physically different from the above-mentioned path for the signaling processing, and the SAR for signaling 314 is used.
- a hop-by-hop transfer SAR cell Z data unit assembling section (referred to as a hop-by-hop transfer SAR) 307 which is provided separately from 1.
- the SAR 307 for hop-by-hop transfer is connected to the IP forwarding section 320 via the hop-by-hop data route 321.
- the above-described cell path 305, SAR for hop-by-hop transfer ⁇ 30 ⁇ , hop-by-hop data unit path 321 and IP forwarding section 320 serve as a forwarding processing means.
- the ATM switch 314 is used for CPU processing via a CPU processing cell path 304 that is physically different from the signaling cell path 310 and the hop-by hop cell path 305 described above.
- Cell Z Data Cut Assembly Section 302 is connected.
- the CPU data processing cell assembling section 302 for CPU processing is connected to a software processing section 310 including a CPU via a CPU processing data path 323.
- a header address of an IP layer or a TCPZUDP layer is used as a hardware. High-speed processing without software processing by the CPU using the forwarding table of the IP bucket that combines the associative memory and the hardware cache search.
- the bucket may be subjected to a foaming process. Such a foredin The processing is described in the specification of Japanese Patent Application No. Hei 9-252521.
- the operation of the CSR relay apparatus shown in FIG. 31 will be described below.
- the cells input to the ATM switch 314 in FIG. 31 are input to the IP processing unit 300 via dedicated links according to their types.
- the signal is a signaling cell, it is transferred to the signaling SAR 311 and the other cells are transferred to the hop-by-hop transfer SAR 307, where each SAR It can be reassembled into a data unit.
- the data reassembled by the signaling SAR 311 is transferred directly to the software processing unit 310 without being taken into the IP forwarding unit 320. Then, the software processing is performed by the CPU.
- the data unit reassembled by the hop-by-hop transfer SAR 307 is transferred to the IP forwarding section 320 to perform hop-by-hop transfer processing. That is, it is determined whether or not the address matches the IP address registered in the associative memory table, and if the address matches, the address to be transmitted to the next relay device according to the address information. After changing the response information, the cell is reassembled by the SAR for hop-by-hop transfer 107 and transmitted to the ATM switch 314.
- the rewriting and management of the associative memory table shall be performed by software processing by the CPU.
- the IP address addressed to the own station and the virtual connection VC-11 are associated so that the IP address is transferred to the software processing unit 310.
- data units for which the method of hop-by-hop transfer processing is unknown are non-IP-based.
- the virtual connection VC-11 is associated with the software so that the virtual connection VC-11 is transferred to the software processing unit 152.
- the virtual connection VC_11 may be divided into a plurality of virtual connections.
- the IP forwarding section 320 the IP address addressed to the own station, such as FANP protocol data.
- ⁇ the IP address addressed to the own station
- the packet arrives, ⁇ ,.
- the packet is given the address information corresponding to the virtual connection VC-11 from its IP address, and then assembled into the cell again by the SAR 307 for Hop-by-Hop transfer. And sent back to ATM switch 314.
- the ATM switch 314 determines that the virtual connection VC—11 should be transferred to the software processing unit 310 and transfers this cell to the CPU processing cell.
- the data is transferred to the SAR 302 for the CPU processing, where the cells are assembled into a data unit in the SAR 302 for the CPU processing.
- the packet is transferred to the software processing section 310 by the CPU.
- IP forwarding section 320 when an IP bucket / non-IP bucket data unit whose hop-no-hop transfer processing method is unknown arrives at the IP forwarding section 320, the non-IP packet is received. After the address information corresponding to the virtual connection VC — 11 is given as a set, The cell is reassembled in the SAR 107 for hop-by-hop transfer and sent out again to the ATM switch 314.
- the ATM switch 314 determines that the virtual connection VC-11 should be transferred to the software processing unit 310, and transfers this cell to the CPU processing cell.
- the data is transferred to the CPU processing SAR 302 via the path 304, and the cells are assembled into a data unit in the CPU processing SAR 302, and the assembled data unit is processed by the CPU. Delivered to the air processing unit 310.
- the self-generated IP from the software processing unit 310.
- the packet is given a destination IP address by the software processing unit 310, and the IP bucket is provided by the hop-by-hop transfer function of the IP forwarding unit 320.
- the address information corresponding to the virtual connection VC-11 is added by the CPU processing SAR 302, and the ATM switch 314 and the cell route 305, SAR
- the data is delivered to the IP forwarding unit 320 via 307, and is subjected to hop-by-hop transfer processing.
- non-IP bucket data units whose IP forwarding unit 320 did not know how to perform hop-no-hop forwarding are transferred to the IP forwarding unit 320. It is sent with the address information corresponding to the virtual connection VC-10 to be forwarded to the destination without being performed.
- the data transfer between the IP forwarding section 320 and the software processing section 310 is physically separated, and the load on the IP forwarding section 320 is reduced. Reduce As a result, the packet transfer capability during hop-by-hop transfer can be improved, and the configuration of the IP forwarding section 320 can be simplified. Also, since the conventional technology can be used as it is, the configuration of each function other than the IP forwarding unit 320 can be simplified.
- FIG. 32 is a diagram showing a modification of the above embodiment.
- it is connected to the IP processing unit 300 via a link for signaling processing and a link for SCU 315 for CPU processing. That is, it is physically connected to the IP processing unit 300 via the same link.
- the above-described effects can be synergistically obtained in a communication mode with less signaling processing. .
- a network based on the CSR method In a packet relay device, in particular, the address addressed to the own station. Ket, non-IP ,. In the packet processing, it is possible to greatly improve the packet transfer capability during the hop-by-hop transfer.
- FIG. 33 is a diagram showing a bucket flow and a TTL operation according to the schematic configuration of the ATM relay device shown in FIG.
- the IP packet passing through the IP forwarding unit 320 receives a TTL (T imeT o Live) is subtracted and the packet addressed to the local station reaches the IP processing unit 300 every time.
- TTL T imeT o Live
- the TTL value is subtracted and the result does not reach the software processing unit 310.
- the TTL value is twice in the IP forwarding section 320 and once in the software processing section 310, and at least three times. Due to the execution of the subtraction, if the transfer distance is shortened when the network is replaced with a single node, or the bucket is discarded in the ATM repeater. Yes There is a problem. The following describes such problems in more detail.
- Fig. 33 shows an example of TTL subtraction by bucket flow.
- the IP forwarding section 320 first performs the first TTL subtraction.
- the subtraction value in the IP forwarding section 320 is 1 or the TTL value a power S1 of the IP packet (not shown) is 1, the TTL value after the subtraction becomes 0.
- the packet cannot be transferred.
- IP unknown with unknown address.
- the IP forwarding section 320 performs the first time
- the software processing section 310 performs the second time
- the third TTL subtraction is performed at 320.
- the IP forwarding unit 320 is composed of hardware, and can be implemented sufficiently shorter than a generally known unit time of TTL subtraction.
- the processing time of all ATM repeaters may be less than the unit time in some cases.In this case, the TTL subtraction value for the entire ATM repeater can be set to 1, as described above. Therefore, the subtraction value increases in the ATM relay device of the above embodiment.
- the TTL subtraction is not performed more than necessary, and the bucket destined for the own station is reliably received, and the transfer distance of the bucket when replaced with a network node is improved.
- the purpose is to provide an ATM repeater.
- the means for adjusting the TTL subtraction for a particular IP bucket to the IP forwarding processing means of the ATM repeater and the software processing means via the third link. Be provided. Hereinafter, this will be described in detail.
- Fig. 34 shows an example of the configuration of an ATM repeater, the transfer route of an IP bucket, and an example of TTL subtraction.
- Fig. 35 shows a TTL calculation process applied to the IP forwarding unit 320. (Steps S401 to S405) are shown in Fig. 36.
- Fig. 36 is a flowchart (steps) of TTL calculation processing applied to the software processing unit 310. It is a figure which shows the example of 3501-3504).
- the ATM switch 314 is connected to the IP processing unit 300 via a link 315 with the IP processing unit 300, and is connected to the IP processing unit 300.
- the link 315 is connected to the IP processing unit 300 via a link 306 physically different from the link 315.
- the forwarding process here, hop-by-hop
- the path to perform the signaling processing and the path to perform the software processing are physically different from the path that performs the transmission.
- the ATM switch 314 is connected to the SAR 309 via a signaling cell path 321 and the SAR 309 is connected via a signaling data unit path 322.
- the software processing unit 310 including the CPU.
- the above-mentioned signaling cell path 32 2, SAR 309, and software processing section 3 10 constitute signaling processing means.
- the ATM switch 314 is provided separately from the SAR 309 via another cell path 305 physically different from the signal processing path described above. Connected to 7.
- the SAR 307 is connected to the IP forwarding section 320 via the data unit path 321.
- the above-described cell path 305, SAR 307, and IP forwarding unit 320 constitute a forwarding processing means.
- the ATM switch 314 is connected to the SAR 309 via another cell path 304 which is different from the above-described path for signaling and fording processing. Te, ru.
- This SAR 309 is connected to a software processing unit 310 including a CPU via a data unit path 323.
- the cell path 304, the SAR 309, and the software processing unit 310 constitute the bucket CPU processing means.
- the IP forwarding section 320 determines whether or not the address is to be sent to its own station based on the destination of the received IP bucket (step S402) or whether or not the destination is unknown. Judgment (Step S 403) is performed, and the TTL subtraction processing (Step S 404) is performed by determining whether or not the TTL subtraction is performed. Also, the software processing unit 310 sends the packet again to the IP forwarding unit 320. IP no. The TTL value of the packet is added by the value subtracted by the IP forwarding section 320.
- Cells externally input to the ATM switch 314 are input to the IP processing section 300 via dedicated links according to their types. That is, if it is a signaling cell, it is transferred to SAR309, and the other cells are SAR3
- the data unit reassembled by the SAR 309 is transferred directly to the software processing unit 310 without being taken into the IP forwarding unit 320, and is sent to the CPU. Software processing is performed.
- TTL operation Transfers to hop-by-hop transfer processing including TTL operation. That is, it is determined whether or not the IP address registered in the table of the associative memory matches the IP address. If the IP address matches, the address is determined according to the address information. After performing the TTL operation and changing the address information to be sent to the next repeater, the cell is reassembled with the SAR 307 for Hop-by-Pop transfer, and the ATM switch is used. It is sent to 3 1 4.
- the rewriting and management of the associative memory table are performed by software processing by the CPU. It also has a VC table memory in which address information, specifically, virtual connection (VC) information, is registered, and an associative memory in which the IP address of the bucket matches.
- VC virtual connection
- the address of the VC table memory that stores the corresponding VC information is stored in the memory area.
- V C1 V C1 is associated as a virtual connection to be transferred to the IP address of the own station and the CPU processing unit.
- VC may be divided into multiple virtual connections.
- the assembled data unit is transferred to the software processing unit 310 by the CPU.
- the IP address perform destination analysis of the packet.
- the IP packet addressed to the local station and the IP packet for which the next transfer destination is unknown are not subjected to the subsequent transfer processing, and the other packets are transferred to the software packet.
- the TTL value corresponding to the processing time in the processing section 310 is subtracted (step S502 in FIG. 36), and the TTL value subtracted in the IP forwarding section 320 Is incremented by 1 (step S503).
- the reason why the calculation is performed here is that 1 is subtracted in the IP forwarding section 320.
- the bucket on which the TTL operation has been performed is re-transmitted. And transferred to the IP forwarding section 320 via the ATM switch 314.
- the calculated value of the TTL operation is 1 in Fig. 35
- the added value of the TTL operation is 1 in Fig. 36
- the TTL subtraction value of the software processing unit 310 is shown. Assuming that 1 is set to 1, in FIG. 34, in the IP forwarding section 320, when an IP packet whose transfer destination is unknown is transferred from outside to the transfer path 302 of the IP packet, first, The TTL subtraction in the IP forwarding unit 320 is not performed, and the software processing unit 310 subtracts and adds one to the original TTL value.
- a process of subtracting one TTL value is performed because a destination is newly registered in a series of processes. That is, only one TTL value is subtracted from the time when the IP packet is input to the ATM relay device of the present embodiment to the time when the IP packet is transmitted.
- the packet relay device in particular, packets addressed to the own station, and destination unknown IP node.
- the packet processing the effect of preventing bucket erasure in the relay device and reduction of the transfer distance in the network during hop-by-hop transfer can be expected.
- the above-described third embodiment it is possible to avoid an increase in signaling traffic particularly in the case of supporting the SVC.
- high-speed signaling performance for SVc control can be obtained, and the transfer capability of a bucket during hop-by-hop transfer can be improved. .
- FIG. 37 is a diagram showing a schematic configuration of an ATM switch according to the first embodiment of the present invention. Note that, in this drawing, components that are not directly related to the present invention are omitted.
- the ATM switch 411 of the present invention comprises an ATM switch core 412 and a plurality of cells to which the cells exchanged by the module 412 are input / output. It consists of an incoming line 4 13 and an outgoing line 4 14, and a routing table 4 15. In the example of FIG. 37, only one incoming line 4 13 is shown for simplicity. Also, the routing table 415 is shown as being located on each incoming line 413, but the placement of the routing table 415 only depends on the contents of the cell exchange information. And this is the present invention Although not directly related, it is shown as an example.
- information sent from a transmitting terminal to a receiving terminal is divided into 48 bytes, and 5 bytes of header information are added to the information, which has a fixed length of 53 bytes in total.
- Information is transmitted in units called cells. Cells sent to the ATM network are rapidly switched (exchanged) on the hardware based on the header information.
- the ATM switch 411 in FIG. 37 is used as an exchange for switching at this time, and is installed in the relay device of the ATM network. You.
- the ATM switch 411 When the ATM switch 411 receives a cell input from the incoming line 413, the ATM switch 411 reads the value of VPI / VCI described in the header of the cell, and reads the routing table 411. With reference to 5, the outgoing line 4 1 4 to output the cell is determined.
- the routing table 4 15 contains the VPI / VCI information of the input cell, the output header information to be converted when output to the output line 4 14, and the output line 4 1 5
- output exchange information which is information as to whether or not to be exchanged, and in this embodiment, a mask bit pattern unique to each VPI / VCI is stored and read.
- the routing table 415 is searched based on the VPI / VCI value described in the header of the input cell. Then, the header information of the input cell is masked by the mask bit pattern specified in the routing tape register 415, and the header VPI / VCI field bit is used. And the VPI / VCI fields in the routing table 15 are checked. To search.
- the input cells are exchanged by the ATM switch core 412 according to the output exchange information of the column hit at this time, and are output to the predetermined output line 414.
- the cell header value is rewritten in the exit header information of the routing table 415, but the VPI / VCI field is masked by the mask bit. Only the masked bit is rewritten.
- FIG 38 is a flowchart showing the cell exchange algorithm.
- the ATM switch 4 11 1 extracts the VPI / VCI value from the input cell's header power and uses it as a route. Inspection is performed for each entry specified in the operating table 415 (Step A12).
- the mask bit number of the routing table 415 is used.
- the turn is ANDed with the value of the VPI / VCI of the input senor header, and this is matched with the VPI / VCI field in the routing table 415 and the turn 0 .
- the ATM switch 411 Rewrites the header information of the corresponding cell based on the output header information specified in the routing table 415. (Step A14). The method of rewriting the information at this time will be described later with reference to FIG.
- the cell After rewriting the header information, the cell is switched through the ATM switch core 412 according to the output switching information (output line) specified in the routing table 415. (to exchange
- Step A15 sends out to the predetermined outgoing line 414 (Step A16).
- the ATM switch 411 determines that the cell is an unnecessary cell and discards it (step A17). .
- the header information (input cell header) of the input cell is determined using a predetermined mask bit notation.
- VPIZVCI VPIZVCI
- the header information is rewritten into output header information, and the cell is exchanged according to the output exchange information.
- the outgoing exchange information of each cell input through each incoming line 413 can be shared, and a plurality of VCCs can be integrated into one.
- FIG. 39 is a view for explaining a cell multiplexing method according to the fourth embodiment.
- the VPI is “0”
- the VCI power S is 4 It is assumed that the cells of the loop are input from each of the plurality of incoming lines 413. That is, it is assumed that four VCCs (communication paths) are drawn on the input side of the ATM switch 411.
- the value obtained by setting all the data of the VPI / VCI to all 1's and the mask bit pattern are EXORed, and the value is calculated.
- the VPI field is reserved for 8 bits for UNI and 12 bits for NNI, and the VPI field is 16 bits for each VPI field. Is maintained. Therefore, this 24 bit (UNI) level has a value of 28 bits (NNI) of VPI / VCI bits all set to 1 and a mask bit of 1 bit of 18 bits. . Take the EXOR with the turn and AND that value with the actual VPIZVCI value. Then, the value obtained in this way is ORed with the outgoing header information.
- the VCI numbers of the headers of the cells input from the incoming line 413 are “1 288”, “1 229”, and “1”, respectively. 30 "and” 1 3 1 ".
- a plurality of VCCs are exchanged by one VCI.
- the mask bit notation is easily separated by changing the full 24 bits (UN) into 28 bits (NNI). can do.
- the mask bit value has been described as “18”, but the setting of this mask bit value is arbitrary. For example, if the mask bit value is set to a large value, the number of input VCCs can be reduced. Also, the area to be masked is It is not based on the distinction between VPI and VCI. The point is to apply a pattern that can mask the VPI / VCI values of the input cells so that multiple VCCs can be combined into one.
- a unique mask bit pattern is set for each VPI / VCI.
- the turn does not require much in comparison to the number of VCC. Therefore, in the present modified example, the mask bit number,. Only the turn is prepared as a separate table, and the mask processing of the cell header is performed sequentially for each specified mask bit pattern, but the routing table search uses the conventional method as it is.
- FIG. 40 shows a table configuration of this modification.
- the difference from the fourth embodiment is the mask bit. Only the turn is a separate table. That is, in the present modification, the ALM switch 411 is individually provided with a notifying table 415a and a Hi-Matic turntable 415b.
- the routing table 415a contains VPI / VCI information of the input cell, egress header information to be converted when output to the egress line 414, Output exchange information, which is information on which outgoing line 4 14 is to be exchanged, is stored.
- the mask number 0 table 4 15 b ⁇ This is, for example, pattern 1, no. Turn 2, no.
- a plurality of different mask bit patterns for masking the VPI / VCI direct of the input cell are stored, such as Turn 3.
- Figure 41 shows the cell exchange algorithm in this modification.
- the end condition of the inspection is not the last entry of the routing table 415a but the last entry of the mask pattern table 415b.
- the ATM switch 411 extracts the VPI / VCI value from the header power of the input cell. And according to the mask. Each mask bit number from the turn table 4 15 b. Turns are taken one by one, and the routing table 415b is searched while masking the VPI / VCI field of the input sensor (step B12).
- the search hits that is, the VPI / VCI values masked by a certain mask bit pattern are displayed in the routing table 41.
- the ATM switch 411 sends the header information of the corresponding cell to the routing table.
- the cell is rewritten based on the output header information specified in 15a (Step B14), and the cell is rewritten according to the output switching information (output line) specified in Table 415a.
- ATM switch core 4 1
- the ATM switch 411 takes out the next mask bit pattern from the mask no. 0 turntable 415b and performs the same search as above (step N of the step B17). As a result, mask no. turn Last mask bit in table 4i5. If the search power S hit is not obtained even using the turn (Yes in step B17), the ATM switch 411 judges the cell as an unnecessary cell and discards it. (Step B 18).
- the routing table search can be realized by the associative memory, so that the degree of freedom of the mask bit pattern is slightly smaller than that in the fourth embodiment.
- search logic can be implemented with simple hardware.
- a plurality of mask bits are used when the routing table is searched. Since the turns must be taken out one by one and applied to the input cell, the mask pattern is uniquely determined for the cell group as in the fourth embodiment. It is preferable to use a network for starting up the network.
- Figure 42 shows an example where the method of the present invention is applied to an actual network.
- two GWs 4 2 4 and 4 2 5 are set up for Pub 1 ic 1 ine (public line) 4 21 ackbonenet (backbone network) 4 24, 4 25, and power connected to them S IP — Subnet (subnet) 4 26 to 4 29, one network
- a method of searching after masking the header information is used to combine multiple VCCs into one. And can be. This can reduce the amount of VPI / VCI resources, especially in mission-critical networks, and can reduce the amount of data Even if the flow of the buckets merges, those buckets will be transferred to another location. Streaming can be performed efficiently without waiting by the packet transfer.
- FIG. 43 is a block diagram showing a schematic configuration of an n ⁇ n ATM switch according to the fifth embodiment of the present invention.
- n X n means n inputs and n outputs.
- the ATM switch according to the present embodiment is an exchange used for an ATM network, and has a port processor 511 for controlling the entire ATM switch, and n switches connected to each line. Ports 1 0 — 1 to 10 — n, these ports 1 0 —:! It consists mainly of an ATM switch core 516 that controls ⁇ 10-n.
- Each port 1 0 —:! ⁇ 10-n are provided with an input-side processing unit and an output-side processing unit.
- the input side processing unit is the input side ATM layer processing unit 5 1 2 —:! 5 5 1 2 — n, input side conversion table 5 1 3-1 5 5 1 3 — n , input side passing cell count counter table 5 1 4 1 1 to 5 1 4 1 n, cell non-conduction time table 5 1 5 _:! ⁇ 5 1 5 — n forces.
- the egress processing unit includes the egress ATM layer processing unit 5 17 — 1 to 5 17 — n, the egress conversion table 5 18 — 1 to 5 18 — n, and the number of output passing cells.
- the winter table consists of 5 19-1 to 5 19 1 n.
- Incoming ATM layer processing section 5 1 2 —:! Under the control of the processor 511, 5 512 _n performs the transfer processing of the cell input from the corresponding line. Specifically, the input side conversion table 5 1 3 — :! to 5 13-n is searched from the VPI / VCI of the input cell, and the information such as the output destination of the cell and the connection identifier is retrieved. Information is obtained, the number of passing cells for each connection is counted, and the value is counted as an entry-side passing cell counter table 5 14 _ :! 5 5 1 4 — Performs processing such as storing in n. Further, in this embodiment, the cell non-conduction time table 5 15 — :! to 5 15 —n is updated.
- Input side conversion table 5 1 3 _;! 5 5 13 _n is a table that stores information necessary for input side transfer processing, such as VPI / VCI information of input cells.
- the entry-side passing cell count table 5 1 4 — 1 to 5 1 4 — n is a table for managing the number of input cells (the number of entry-side passing cells).
- the cell non-conducting time tables 5 15 _ 1 to 5 15 — n are tables for managing the time when cells are not conducting (non-conducting time).
- the input side conversion table 5 1 3 _:! Number of excess tapes 5 1 4 —; ⁇ 5 1 4-n, and cell non-conducting time table 5 1 5-:! ⁇ 5 1 5-n need not be physically separated, and are physically configured as a single table You can do it.
- the output ATM layer processing section 5 1 7 —:! 5 5 1 7 — n performs a process of transferring cells output to a predetermined line under the control of the processor 5 11.
- the outgoing-side conversion table 518 1: !! is determined by the connection identifier transmitted together with the cell.
- 5 5 1 8 —n is searched to obtain a new VPI / VCI, or the new VPI / VCI is replaced with a header to output a cell to a predetermined line.
- Outgoing conversion table 5 1 8 —:! 5 5 18 _n is a table that stores information necessary for outbound transfer processing, such as outbound header information to be converted when output to the outbound line.
- the counter table 519- :! to 5191-1n is used to manage the number of cells output to a predetermined line by rewriting the header information (number of outgoing cells). It is a table.
- Outgoing conversion table 5 1 8 —:! 5 5 18 _ n and the outgoing side passing cell count table 5 19 _;! 5 5 19 _ n need not be physically separated, but It is also possible to configure it. Also, since the number of passing cells is counted on the ingress side, this is the number of passing cells on the egress side. The value of n is not particularly necessary, but if a large number of cells are input, the cells sent to the egress side may be lost. It is important.
- FIG. 44 is a block diagram showing a schematic configuration of an incoming ATM layer processing unit in the embodiment.
- Incoming ATM layer processing unit 5 1 2 — 1 to 5 12 _n includes input processing unit 5 21, cell buffer 52 2, conversion table control unit 5 2 3, passing cell counter It comprises a table controller 524, a cell non-conducting time table controller 525, an output processor 526, and a processor I / F 527.
- the input processing unit 521 extracts VPI / VCI, etc., from the input cell car, converts the conversion table control unit 523, the passing cell count counter control unit 524, and the cell non-conduction time. The data is transferred to the table controller 525. In addition, the cell itself is transferred to the cell buffer 522.
- the cell buffer 52 2 is a memory for temporarily storing cells that serve as 53 knots.
- the conversion table control unit 5 2 3 The input side conversion table 513-1 is searched based on the input cell VPI / VCI, etc., to obtain information such as the output destination of the cell and the connection identifier.
- the number of passing cells counter table controller 5 2 4 Refer to the counter table 5 1 4 _ 1, and if the input cell has been received normally, write the value obtained by adding 1 to the referenced counter value in the counter table 5 1 4 1 1. return.
- the cell non-conducting time table control unit 5 25 writes information indicating that the cell has passed into the cell non-conducting time table 5 15. In this embodiment, 0 is written.
- the cell non-conducting time table control section 525 periodically enters the cell non-conducting time table 515 once every cell cycle except when a cell is input. It has a function to update the file.
- the processor I / F section 527 controls access from the processor 511.
- Outgoing ATM layer processing section 5 1 7 —:! 5 to 7 — n have the same configuration as described above, except for the cell non-conduction time table controller 5 25.
- 5 5 1 4 Stored in n.
- the cell is transferred to the ATM switch core 516 along with the information obtained in the input side conversion table 5 13 -1 to 5 13 _n.
- the output destination obtained from ⁇ 5 1 3 —n indicates the output destination for the ATM switch core 516.
- ⁇ 5 1 7 Transfer cell to n.
- the egress conversion table is based on the connection identifier transferred together with the cell via the ATM switch core 516.
- 5 1 8 —:! Search for 5 5 1 8 — n to obtain a new VPIZVCI. Further, the new VPIZVCI is attached to a header, and the cell is output to a predetermined line.
- the outbound ATM layer processing unit 5 1 7 —:! The number of passing cells for each connection is counted, and the count value is counted as the number of outgoing passing cells. 55 1 9 1 Store in n.
- the cell input via the line 1 is the incoming ATM layer processing unit.
- 5 1 2 Input to the input processing section 5 2 1 in 1.
- the input processing unit 5 2 1 searches the cell for the input side conversion table 5 1 3 — 1, the input side passing cell counter table 5 1 4 — 1, and the cell non-conduction time table 5 1 5 — 1
- VPI / VCI, etc. which are the keys of the system, are extracted and converted into a conversion table controller 523, a passing cell count controller 524, and a cell non-conduction time table controller. Transfer to 5 2 5.
- the cell itself is transferred to the Serno buffer 522.
- the conversion table controller 523 searches the input side conversion table 513-1 based on the VPI / VCI extracted from the cell, and outputs the output destination of the cell and the connection identifier. get information.
- the passing cell number counter table controller 524 refers to the input side passing cell counter table 514-1 based on the same VPI / VCI, and the input cell is normally received. If it is, the value obtained by adding 1 to the referenced counter value is written back to the entry side passing cell number counter table 5 14 _ 1.
- “0” is written in the cell non-conducting time table 5 15 -1 as information indicating that the cell has passed.
- the cell non-conducting time table control unit 525 has a function of periodically updating the entry of the cell non-conducting time table once every cell cycle except when a cell is input. .
- the table entry is updated by changing the target entry for each cell cycle, for all the entries supported by the hardware, and reading the entry. This is done by adding 1 to the value of and writing it back.
- the cell non-conducting time table 5 15 5-1 has an entry of 0 if the cell passes periodically, but if the cell does not pass for a long time, The entry in the cell non-conduction time table 15 1 monotonically increases in proportion to the time.
- the entry time of the entry is reduced. Updating can be performed by incrementing the SRAM address by one for each cell cycle, so it takes approximately 90 ms (2.7 usx) to update the entries for all connections. 32 K) Required.
- the entry for a particular connection is updated approximately every 90 ms, so if the value of the entry is 1, 90 ⁇ 90 ms before The last cell passed the connection at the end of the connection, and if the value of the entry is 2, the last cell passed 180 ms 90 ms before .
- the entry S0 will occur when the cell passes, and the entry will be restarted.
- the trie becomes 1 at the next regular update about 90 ms later, and the entry becomes 2 about 180 ms later.
- the ingress ATM layer processing unit 5 1 2-1 the time during which no cell input via line 1 is flowing (cell non-conduction time) is periodically detected. Then, the information is stored in the cell non-conduction time table 5 1 5 -1. In the output processing section 526, the cell transferred from the senor notifier 522 and the information transferred from the conversion table control section 523 are combined and output to the ATM switch core 516.
- the processor I / F section 527 controls the access of the processor 511. Access from the processor 5 11 1 to the input side conversion table 5 1 3 — 1, the input side passing cell counter table 5 1 4-1 or the cell non-conduction time table 5 1 5 — 1 In this case, the processor IF section 527, the conversion table control section 523, and the passing cell count counter table control section 524 are provided through the cell non-conduction time table control section 525. Done.
- Processor 511 is a non-conducting time table for all connections to be managed. Polls the entries of ⁇ 5_15_n periodically, and determines whether or not to release the connection based on the read values.
- the entry of ⁇ 5 15 -n is zero if the senor is conductive, and if the cell is not conductive, the value is proportional to the time. Therefore, by comparing with the connection non-conduction time set for each connection uniformly or for each connection, the connection that has passed the reference cell non-conduction time has elapsed. It can be determined that the session is not currently being used, and connection release processing is performed immediately.
- connection release processing can be performed based on the cell non-conduction time since the cell non-conduction time increases at the next polling. As a result, unnecessary connections can be released early, and connection resources can be used effectively.
- the non-conducting time of the cell is managed on the input side (input side ATM layer processing unit) of the port connected to each line.
- input side input side ATM layer processing unit
- a similar time table and control unit are used.
- the outgoing side of the port may be provided so that both the incoming side and the outgoing side manage the non-conducting time of the cell.
- the processing of the ATM switch core 516 cannot catch up, and cells sent to the egress side may be lost.
- the reliability of the outgoing side is lower than that of the incoming side. Therefore, it is preferable to manage the non-conducting time for each connection at least on the input side.
- connection non-conduction time is managed and the connection is released based on the cell non-conduction time, so it is unnecessary regardless of the processor polling cycle. Connections can be released early. As a result, the connection resources can be used effectively.
- the relay apparatus of the present invention is used as an ATM network switch, and performs routing processing in a network layer by software or hardware (hop-by-hop).
- a bypass path (cut-through transfer) that is established according to the protocol higher than the trans- port layer that is exchanged between adjacent relay devices. (Switching) in the data link layer using only the hardware without software. By doing so, bucket routing is performed at high speed.
- the configuration of the relay device is as follows: (a) the hop-by-hop transfer process is performed by software; and (b) the hop-by-hop transfer process is performed by hardware. The explanation is made separately for and.
- Fig. 46 is a block diagram showing the schematic configuration of the relay device of the present invention when an ATM switch is used as a switch and hop-by-hop transfer processing is performed by software. is there.
- 6 1 1 —:! ⁇ 6 1 1 1 1 n is the input line
- 6 1 2 — 1 to 6 1 2 — (n — 1) is the priority control section
- 6 13 is the ATM switch
- 6 14 is the switching table
- 6 1 5 —:! 6 6 15 — n indicates the output line
- 6 16 3 ⁇ 4 SAR 6 17 indicates the processor
- 6 18 indicates the frame memory.
- Fig. 46 the difference from the conventional configuration is that the priority control unit 6 1 2-:! ⁇ 6 1 2 — (n-1) is provided.
- This priority control section 6 1 2 —:! 6 6 1 2 — (n — 1) is provided before the ATM switch 613, handles input data in bucket units, and performs priority control according to the packet type. is there.
- Packets input from input lines 6 1 1 — 1 to 6 1 1 — (n — 1) are sent to priority control section 6 1 2 —:! ⁇
- the queue keys provided for each packet type 6 2 3 —:! 6 6 23 -m are assigned priorities with respect to the order of output to the ATM switch 613.
- the priority control section 6 1 2 — 1 to 6 1 2 — (n-1) is a buffer queue with this priority.
- the packet When a bucket is input to the ATM switch 613 in the form of an ATM cell, the packet is referred to by referring to the switching table 614.
- the packet (actually an ATM cell) is switched by hardware.
- information such as which packet is output to which line is set in advance.
- a bucket flowing through a flow for which hop-by-hop transfer or cut-through transfer has not been determined is sent to the switching table 614 to the output line 615_n. The settings to output have been made.
- the bucket output to output line 6 15 — n is input to SAR 616. Since the bucket is divided into ATM cells and input, the original data is stored on the frame memory 618 so that the processor 17 can process the SAR 616. Assembled into a kit. In other words, it is not possible to determine just how much data is used in a layered application, or how long it is, just by looking at a single cell. By collecting the files, the processor 617 is processed.
- the SAR 616 stores a packet indicating that the trigger packet of the flow has arrived at the processor 617 and stores the packet. Notify with the address of the frame memory 6 18.
- the processor 617 compares the trigger condition with the trigger condition and determines whether or not to transfer the flow of the bucket through a single cut.
- the processor 617 stores the ATM switch 613 in the switching table 614 only by switching processing of the ATM switch 613. Make settings so that buckets flowing through the row will be routed. Hopba When transferring data in a hop-by-hop manner, the switching table is not changed, and the packets flowing through the flow are hop-by-hop transferred by the processor 617 using software. Transfer processing.
- a cut-through packet (a cut-through packet, a packet) and a hop-no hop.
- the processing of a packet packet transferred by hop-by-hop
- Attach Switch 6 13 to the cutout.
- an output line is determined according to the switching table, and the cut-out louver is applied to the determined output line 6 15 — 1 to 6 15 — (n ⁇ 1). The test is output.
- the setting for switching to the output line 6 15 — n is made in the switching table 6 14, and the level is set.
- the bucket output to the output line 6 15 —n of the switch 6 13 is transferred to the SAR 6 16.
- the ATM sensor is restored on the frame memory 618 so that the processor 617 can process it. Assemble into a kit.
- SAR 616 informs processor 617 that a hop-by hop bucket has arrived. This is notified together with the address of the frame memory 618 in which the packet is stored.
- the processor 617 that has received the notification performs the routing process based on the header information (the source address, the destination address, etc.) of the bucket. When the notifying process is completed, the processor 617 should transmit to the SAR 616. Signals that the packet is on frame memory 6 18. SAR 6 16 The transmission bucket is extracted from the frame memory 618, divided into ATM cells, and input to the input line 6111 of the ATM switch 613. Input line 6 1 1 — n input from input. The packet is output to an output line 6 15 — 1 to 6 15 — n determined according to the switching table 6 14.
- FIG. 47 shows the priority control section 6 1 2 —:! 6 is a block diagram showing a schematic configuration of 6 1 2 —n.
- Priority control section 6 1 2 —:! ⁇ 6 1 2-(n-1) is for each input line 6 1 1-:! 6 6 1 1 — (n-1) are provided in front of the ATM switch 6 13, and as shown in FIG. Ket assembling section 62, filtering section 621, filtering table 622, knockout section 623-1-62-3m, key Menu management section 624 and selector 625.
- the bucket assembler 620 assembles each of the divided cells input in the form of ⁇ TM cells into a bucket format. In this case, information indicating the type of the bucket is written in a predetermined cell in each cell, and the socket assembling section 6220 is based on this information. Assemble each cell in buckets.
- the filtering part 6 2 1 is no.
- the type of the bucket obtained by the bucket assembling section 62 is determined, and the bucket is adapted to the type by referring to the filtering table 62.
- VPI / VCI is used as a method of determining the type of the socket. There are methods to use.
- VPI Virtual Path Identifier
- VCI Virtual Channel Identifier: virtual channel identifier
- VPI power S 8-bit In this case, in the UNI (User-Network Interface), the VPI power S 8-bit, VCI power S16-bit power, and the fields actually used are as follows. May be restricted. In addition, the usage of some values is limited in advance, and one of them has a signaling VCC (Virtual Channel Connection) for setting up a connection. is there.
- This signaling VCC is assigned to VCI 1, 2, or 5 depending on the signaling application, for example. Therefore, when discriminating the type of the input packet, the VCIs 1, 2, and 5 described above are input into the predetermined cells in the packet. If it is used, it can be determined as a bucket for connection settings.
- VCIs in 1, 2, and 5 above indicate the following types.
- V C I 1: Meta signaling
- V C I 2: Broadcast signaling
- V C I 5: Point-to-point signaling filtering table 622 contains such a feature.
- the selection conditions of the buffer queue according to the type of the packet are set in advance and are displayed.
- Knockout 6 2 3 _:! 6 6 2 3 — m is a memory for storing the bucket filtered by the filtering unit 6 21, for example, twenty three -
- the priority is assigned in the order of 1, 6 2 3-2, 6 2 3 — 3....
- the queue management unit 624 controls the output of the buckets stored in the not-acquisition unit 6 2 3 —:! To 6 2 3 — m using a fixed priority method (see FIG. This is done by the robin method (Fig. 49).
- the selector 625 operates under the control of the queue management section 624, and the not-acquisition 623::! The buckets stored in 6 6 23 _ m are selectively output to the ATM switch 6 13.
- each cell input in the form of an ATM cell is No, according to the kit assembly section 62. It is put into the form of a packet and given to the filtering unit 62 1.
- the finalizing unit 621 determines whether the input packet is a normal packet, and, if it is determined that the packet is a normal packet, determines that the packet is a normal packet. Determine what kind of bucket the bucket is. The determination of the bucket type is made by reading the information written in a predetermined cell among the cells constituting the bucket.
- the type of the socket is Trigano.
- the packet includes a packet, a bucket including routing protocol information, control / management information between relay devices such as signaling cells, and a data packet of a user.
- the filtering unit 21 refers to the filtering table 62 to determine which packet of the type is used. ;! 6 6 2 3 — Determines whether to store in m. At this time, simply set the buffer type for each bucket type. ⁇ 6 2 3 — When assigning m, ⁇ ,. It is sufficient to make the bucket type and the Knob Acc. 6 2 3 — :! to 62 3 — m correspond one-to-one, but here the bucket type and other information (for example, The knuckle accumulator 62 3 3 ---! Under the combination of the source of the bucket, the destination of the bucket or the flow to which the bucket is transferred.
- 6 6 2 3 To allow for flexibility, such as allocating m, place a filtering table 62 2 on it. This final lettering table 62 2 is searched to find the relevant information.
- the knob accumulator 6 2 3 — 1 to 6 2 3 -m in which the packet is to be stored is determined, the filtering unit 6 21 determines that the bucket has been determined.
- the buffer queue 6 2 3 —;! To 6 2 3 — m takes the form of a FIFO (first-in-first-out method), which is a method of outputting in order from the first stored bucket.
- each of the queues 6 2 3 — 1 to 6 2 3 — m output to the queue management section 6 2 4
- a signal (key status signal) indicating whether or not one or more buckets to be output are stored in the ATM switch 613 side is input.
- the queue management section 6 24 should output the next ATM switch 6 13 based on the cue status signal.
- Knock-out key that stores the packet 6 2 3 —:! 6 6 23 _ m is determined, and the bucket output to the ATM switch 613 is permitted only for the corresponding off-chip queue 6 2 3 — 1 to 6 23 _ m Signal to be activated (output enable signal).
- Roh Tsu off A Kiyu 6 2 3 queue management unit 6 2 4 is applicable -:! ⁇ 6 2 3 - Nono of m force et al. 0 Ke Tsu door mosquito Ni you'll be output to the ATM sweep rate Tutsi 6 1 3 Switch the selector 6 2 5.
- How to perform priority control depends on the algorithm implemented in the queue management unit 624. In the present embodiment, two methods, a fixed priority method (FIG. 48) and a round robin method (FIG. 49), are shown. Here, the cue status (#X) signal is output to the #N. Indicates whether or not more than one packet has been stored. When "H" level, 1 no.
- the output enable (#X) signal indicates whether or not packet output is allowed to the buffer queue (#X). When the signal is at "H" level, the output enable (#X) signal indicates that output is enabled. There is. The highest priority is assigned to the buffer queue (# 1), and the lowest priority is assigned to the buffer queue (#m).
- the packet when determining the next packet to be output to the ATM switch 613, the packet is stored in one or more buckets and the packet is stored.
- the highest priority is the Nokia Aki Power. This is a method for outputting a packet.
- the output enable (# 2) which allows the bucket to be output to the buffer pool (# 2), has a higher priority than the buffer pool (# 2). Only when no packet is stored in (# 1) (queue status (# 1) two "L” level), the level becomes "H,”.
- the decoder 631 decodes the output enable signal (# 1) to the output enable signal (#m), and outputs the data from the buffer buffer for which the output is permitted.
- the packet is output to ATM switch 613.
- a signal (selector control signal) for controlling the selector 625 is generated.
- a packet is always output from the buffer access power having the highest priority.
- the buffer queue that previously output the packet to the ATM switch 613 is used. This method gives the right to output a packet to the buffer with the next highest priority. However, since only the right to output is given, if the packet to be output is not stored in the entitled buffer queue, the packet with the next highest priority is output. Grant rights to the buffer queue.
- the queue management unit 62 4 has a configuration as shown in FIG. 49, and outputs, for example, the last buffer cue (# 1) force, the last packet, and the ° packet.
- the last buffer cue (# 1) force for example, the last buffer cue (# 1) force
- the last packet for example, the last packet
- the ° packet In this case, first, it is determined whether or not the bucket to be output is stored in the Knock Queue (# 2). If it is stored, the round robin scheduler 64 1 gives output permission to the buffer pool (# 2), and the output enable generator 64 2 outputs it. Only enable (# 2) is set to "H" level and output.
- the decoder 643 is output from the knockout accumulator (# 2). No, A signal for generating a signal (selector control signal) for controlling the selector 625 so as to be output to the ATM switch 613. Generate a signal that controls the knockout accumulator (# 2).
- the round-robin scheduler 641 places the packets in the buffer queue (# 3). Give the right to output, and determine the power at which the packet to be output is stored in the Knockout Queue (# 3). The same operation is repeated until a Knock-Away storing the bucket to be output is found.
- the round-robin scheduler 64 1 When a round-robin scheduler 641 finds a buffer that stores a packet to be output, the round-robin scheduler 64 1 has a buffer having the next highest priority after the buffer queue. When the packet is stopped at the queue and the packet to be output next time is determined, the right to output from the buffer queue is given, and the packet to be output to the buffer queue is given. It is determined whether the data is stored.
- a bucket is output to the ATM switch 613 preferentially from a specific buffer, like the fixed priority method described above. Instead, this method targets all buffer queues and outputs buckets in order of priority from each buffer queue.
- the priority control section 6 1 2 — :! is provided before the ATM switch 6 13.
- 6 6 1 2 — Provided in (n-1) input data is handled in units of buckets before switching and routing processing, and priority is given according to the packet type.
- the line connecting ATM switch 613 and SAR 616 is Even when traffic exceeding the bandwidth is input, the most important buckets are preferentially transferred to the SAR 616 and processed, so they are discarded in the ATM switch 613. Is less likely to be done.
- cutting trigano By giving high priority to control and management information between relay devices, such as packets, control cells, and routing information, it is possible to quickly share control and management information between relay devices. Become.
- the relay device can preferentially process the user data.
- the priority setting is based on the scenario of the user of the relay device (high priority given to user data transferred by hop-by-hop, high priority given to control information, high priority given to management information) Etc.) can be performed flexibly.
- FIG. 50 is a block diagram showing a schematic configuration of a relay device in a case where an ATM switch is used as a switch and a hop-by-hop bucket is subjected to routing processing by hardware.
- 6 5 1 —:! ⁇ 6 5 1 — n is the input line
- 6 5 2 —:! -65 2-(n-1) is the priority control section
- 65 3 is the ATM switch
- 65 4 is the switching table
- 65 5-:! ⁇ 65 5 — n is the output line
- Reference numeral 6556 denotes 31; 6557 denotes a processor; 6558 denotes a frame memory; and 6559 denotes a routing processing unit.
- the difference from Fig. 46 is that the hop-by-hop ⁇ is transferred from the output line 655_n to the SAR 656 and assembled on the frame memory 658.
- the notting processing of the packet is performed by the notching processing unit 6559.
- the routing processing unit 659 When the SAR 656 informs the frame memory 658 that the assembly of the packet has been completed, the routing processing unit 659 outputs the packet header. Routing processing is performed based on the sender information (sender address, destination address, etc.). After the end of the routing process, the SAR 656 is notified that the bucket to be transmitted exists in the frame memory 658. The SAR 656 divides the bucket into ATM cells and inputs the packet to the ATM switch 653 through the line 651-n.
- the packet includes only a packet containing the trigger packet and routing protocol information, and a packet destined for the relevant relay device, such as a signaling cell used for establishing an ATM connection. Become.
- FIG 51 shows the schematic configuration of the routing processing unit 659.
- the routing processing section 6559 is composed of a frame memory management section 661, a header analysis section 662, a table I / F 666, and a foraging table 664. Become. Inside the routing processing unit 659, the SAR 656 is connected. In addition to receiving notification of the completion of the bucket assembly, receiving the address of the frame memory 658 in which the bucket is stored, etc., and managing the frame memory It is managed in part 6 6 1.
- the frame memory management unit 661 generates the header information of the bucket based on the address of the frame memory 658 in which the notified bucket is stored. Is extracted and transferred to the header analysis unit 662.
- the header analysis section 662 analyzes the extracted header information and searches a forwarding table 664 that stores the routing information of the packet. And notify table IZF 663.
- the table IF 663 uses the search key of the notified forwarding table 664 to output the routing information, which is the output route information of the bucket, to the forwarding table 664. Search and extract from. The extracted routing information is transferred to the frame memory management unit 661 together with information for identifying which bucket the routing information is related to.
- the frame memory management unit 661 writes the extracted routing information to the frame memory address where the bucket is stored, and the SAR 656 completes the routing process. Notify this.
- the priority control section 652--! Provided before the ATM switch 653 is provided.
- ⁇ 652-(n-1) handles input data in packets and performs priority control according to the packet type.
- Priority control section 6 5 2 —:! 6 652 — (n—) is the same as that in Fig. 47, and the configuration of the key management unit 624 provided therein (fixed priority method / round robin method). Expression) is also the same as in FIGS. 48 and 49, and the description thereof is omitted here.
- the priority control section 65 2 _ 1 to 65 2 — (n-1) controls the bandwidth of the line connecting the ATM switch 653 and the SAR 656. Even if more traffic is input, the more important buckets are preferentially transferred to the SAR 656 and processed, so they are discarded in the ATM switch 653. Is less likely to occur.
- the relay device can preferentially process user data.
- the priority settings are based on the scenario of the user of the relay device (high priority given to user data transferred by hop-by-hop, high priority given to control information, high priority given to management information) , Etc.).
- input data is input in buckets, and the priority is determined according to the bucket type.
- high-priority packets can be transferred and processed preferentially, and efficient data can be transmitted. Communication can be realized.
- FIG. 53 is a diagram showing the configuration of the seventh embodiment of the present invention.
- CSR consists of an ATM switch 710 and an IP processing unit 711 connected to the ATM switch 710 via an ATM—IF 712. Consists of forces.
- the IP processing section 71 1 is the same as the CPU 71 1a and the AAL 5 section 71 1b, and the CSR protocol processing between adjacent nodes is not processed or cut through. IP no. This is the part where the Ket, that is, the hop-by-hop processing is processed by software (CPU).
- This A plurality (n) of the IP processing units 711 may be connected to improve the hop-by-hop processing capability.
- n l,
- the ATM switch 710 has a large number of ATMI / Fs 712 for connecting a large number of devices (routers, terminals, ATM switches, etc.) that can be connected via the ATMI / F 712. Te, ru.
- the ATM switch 710 of the present embodiment performs queuing of VCs for each PDU by the ATM switch, that is, performs cell sorting by a function based on perV C queuing.
- the ATM switch is equipped with an ATM switch core.
- the ATM switch core 720 has an input processing unit 721 for processing input cells and an address conversion table for performing address conversion for processing inside the ATM switch (this is not essential. However, queuing is performed for each VC, that is, for each per PDU, and in consideration of QOS support, VC or PDU units, and each output is conscious of priority.
- An ATM switching function 723 that provides an ATM switching function including a per VC (PDU) scheduler section 724 that performs scheduling to a port, and a cell buffer that stores received cells. 7 2 5 and the cell A control section for storing or reading data, that is, a cell buffer address, attributes (QOS, priority, etc.), pointer control, etc.
- a response table 7 26 and an address conversion table 7 for VPI / VCI address exchange for outputting cells arranged and output in units of VCs or PDUs to the ATM I / F. 2 2 'and a table that allows you to refer to the information on whether or not to perform VC merging.
- Selected VP 1 / VCI—> Merged VC section 7 27 You. This Selected VPI / VCI—> Merged VC section 727 may share the function with the address conversion table 722 ′.
- the cells output-processed by the output processing unit 728 are sent to the ATMIZF 714.
- the AAL 5 unit 713 is connected to the ATM switch core 720, but basically the CPU 715 CPU of the ATM switch 710 IP processing unit 7 This is for connecting both CPUs via the ATM switch core 720 and the ATM I / F 714 when communicating with the 11 CPU 71 1a.
- the CPU 715 of the ATM switch 710 is a device management CPU, but receives an instruction from the IP processing unit 711 in accordance with the CSR protocol, and receives the instruction from the ATM switch 711. Controls 0. In other words, it controls the ATM switch core 720 such as setting and releasing the VPI / VCI and the various types of tables described above.
- FIG. 54 is a view for explaining the operation of the seventh embodiment of the present invention.
- the VCs are queued for each PDU for each port, that is, the cells are sorted by the shaving function using perV C queuing.
- FIG. 55 shows another example of the seventh embodiment, in which per VC queuing, that is, per PDU queuing is performed intensively at one location of the ATM switching unit.
- the function of the ATM switch core 740 will be described. The explanation starts from the ATMI / F input side. The function may be simpler than the ATM switch core 720 described in FIG.
- the ATM switch core 740 includes an input processing unit 714 for processing input cells and an address conversion table for performing address conversion for processing inside the ATM switch. Not required) 7 4 2
- An ATM switching and scheduling unit 744 that provides an ATM switching and scheduling function; a cell buffer unit 745 that stores received cells; Control section for storing or reading, that is, the address of the cell buffer, the attribute (QOS, priority, etc.), and the knob address for performing pointer control, etc.
- the cell data is output-processed by the output processing unit 748, and then input to the ATM IZF750.
- the output port of the ATM switch core 740 is connected to the ATM I / F 750, but this ATMI / F 750 supports the topology drain.
- address management of the AAL5 section 753 that provides the SAR (Segmentation and Reassembly) function, the buffer section 751, and the buffer section 751 is performed.
- No-fade-response table 752, address conversion table 749 for performing address conversion, and table for referencing information about VC-merged and non-recorded layers Selected The VPI / VC I-> Merged VC block consists of 747 and power. This Selected VPI / VCI_> Merged VC section 747 may have the same function as the address conversion table 749.
- AAL5 section 753 of this embodiment is an ATM switch.
- AAL5 unit 72 3 is connected to the ATM switch core 74 0, basically, the CPU 72 5 of the ATM switch 72 0 SIP processing unit 7 This is for connecting both CPUs via the ATM switch core 740 and the ATM I / F 724 when communicating with the CPU 21 21a of the 21st.
- the CPU 725 of the ATM switch 720 is a device management CPU, but receives an instruction from the IP processing unit 721 in accordance with the CSR protocol, and receives the instruction from the ATM switch 722. Controls 0. That is, it controls the ATM switch core 740 such as VPI / VCI setting and release, and various tables described above.
- FIG. 57 is a view for explaining the operation of the modification of the seventh embodiment of the present invention.
- CS-PDU is taken out by connecting AAL5 to the output port of the ATM switch, and then disassembled and assembled again. This is equivalent to scheduling in PDU units, and the cells can be sorted by such a method.
- FIG. 57 when input to the AYM switching + scheduling unit 744, as shown in FIG. 57A, cells having different QOS (priority 1 and priority 2) are obtained. PDUs cannot be assembled correctly. Then, the signal is input to the AAL5 unit 753 and decomposed for each cell in the AAL5 decomposing unit 753a under the control of the CPU 754. Next, the cells disassembled in the AAL5 assembling unit 753b are assembled into packets in units of PDUs according to the QOS (priority 1, priority 2). As a result, as shown in Figure 57B You. It is sorted and output in PDU units in descending order of priority.
- FIG. 58 is a diagram showing another example of the modification of the above-described seventh embodiment, in which an AAL5 section 753 is provided at a specific port of the ATM switch.
- the present embodiment supports topology driven as a function of CSR. Then, the data is transferred to a subnet having the same address.
- the output ports of each ATM switch are concentrated in the manner described above, and the VC is queued for each PDU and shaving is performed. Cells are sorted by function.
- Another method is to connect an AAL5 to the output port or a specific port of the ATM switch, take out the CS-PDU, disassemble it, and reassemble it. Scheduling is being performed. Therefore, it is possible to provide an ATM relay device having a dual mode in which a cut-through path can be used for each class of communication quality.
- cells are sorted and merged and transferred in units of PDUs according to the communication quality. Therefore, a power-through path is selectively used for each class of communication quality. It is possible to provide an ATM repeater having a dual mode capable of performing this.
- IP In terms of the packet, the IP address to transfer. It monitors the protocol information and port number of the packet, and monitors the source IP address and destination IP address of the IP bucket to be transferred.
- IP mode This mode shifts to the cut-through from the source IP address and destination IP address of the header of the KET.
- FIGS. 59A and 59B are diagrams for explaining the outline of the packet transfer state monitoring device of the present embodiment.
- the monitoring / analyzing unit 80 shown in FIG. 59A monitors and analyzes the IP packet transfer status over a medium and long term.
- the trend predicting unit 81 predicts the medium- and long-term IP packet transfer trends.
- a flow in which the cut-through mode is considered to function effectively from the predicted result Determine the topology.
- the cancel unit 803 automatically sets the cut-through / trigger information in accordance with the change over time, and also performs this setting. If it is found that is invalid, the setting is canceled.
- the following processing is performed even if the cut-through trigger information has already been set.
- the cut-through state is already set by the use state monitoring unit 810, and the use state is monitored for the trigger information that has been set, and the validity state is checked.
- the judgment section 8 2 1 judges whether the setting is functioning effectively. For those that are determined not to be working effectively, the release 'reset section 8 22 2 automatically releases the cut-through' trigger information and If more effective conditions are found, reset them.
- FIG. 60 is an example of a network incorporating a CSR.
- Host A (820)-Host B (822) Host B (822)-Between Host C (822)
- Host A (820)- Consider the operation on CSR822 when files are transferred between hosts C (822), respectively.
- the IP knock that has entered CSR 822 is indicated by the arrow in Figure 61.
- the CSR 82 3 IP processing unit (IP controller) 1311 is sent and processed at the network layer level (IP processing) 1fi is performed and the next Transferred to the hop (CSR 825).
- IP processing IP processing
- the IP node At the time of the IP processing at this time, the IP node.
- the protocol information, port number, source IP address, and destination IP address are read from the header of the kit, and the cut-through trigger information and The comparison will determine the transition to power-through mode.
- communication is performed with the adjacent CSR using a protocol called FANP to negotiate the cut-through path.
- FIG. 62 shows the movement of the CSR when the cut snoring trigger force S is applied. Do not switch to cut-through mode
- the number of transferred IP buckets is counted every three hours.
- the part indicated by _ _ / x means a specific month and day, and ⁇ and ⁇ mean that it changed on the next day. If this was collected for one week, it is assumed that the following trends were predicted from the aggregated data.
- the number of transfer buckets per unit time during which the cut-through mode is valid must be used as a reference. No.
- the reference number of transfer buckets per unit time was calculated in consideration of the performance of the process of shifting the CSR to the cut-through mode.
- the recommended cut-through trigger information determined in this way is automatically set according to its temporal change.
- the usage status of the network changes daily, and it is possible that the recommended cut-through trigger information once determined becomes invalid. Therefore, by monitoring whether the operating cut-through mode is working effectively, it can be disabled. Make sure to release the trigger information.
- the reference is the number of cells per unit time. If the number of cells per unit time in the monitoring result does not exceed the reference number, it is determined that cut-through is not working effectively. Automatically cancels the invalid cut-through trigger information that has been set. In addition, if the cut-through transfer has already been performed at the time of this release processing, the invalid cut-through trigger information is released and the cut-through trigger as shown in Fig. 64 is released. Forced switch from through-thru transfer to hop-by-hop transfer.
- the cut-through trigger information is set by monitoring the packet transfer status over time, but the present invention is not limited to this.
- the cut-through trigger information may be set according to the geographical condition where the router is located or the priority of the bucket to be transferred.
- the router using the label switching technology itself determines conditions for shifting to efficient cut-through, automatically sets the conditions, and performs subsequent cutting. The effectiveness of the transfer is also monitored so that efficient network The operation of the network can be realized, and the load on the network administrator can be reduced.
- Fig. 66 shows a system configuration to which the method for setting a designated path according to the present embodiment is applied.
- the network of a label switch and a network composed of other devices or other devices are used.
- Routers 903 and 905 located at the boundary with the network constituted by the level switches of the above, and router 904 located within the label switch network, and other devices.
- the routers 902 and 906 of a network composed of other or composed of other label switches are arbitrarily connected by a physical line 901. It is a network.
- reference numeral 907 denotes a path for passing all data except when a logical line designation path is set.
- 908 is a designated path set when a strict QOS request is made
- 909 is a designated path set when a strict QOS request is made.
- the QOS information is, as described later,. It shall be inserted into the socket header. In this designated path, it is possible to reserve the bandwidth in advance.
- FIG. 66 shows only one-way data flow, data can be transmitted in both directions.
- QOS information will be described in two cases: severe and non-strict. S requests can be divided into as many levels as the designated path can be set. In addition, there are various QOS requirements, and 1 to 4 in Fig. 67 show examples. Here, if disposal is permitted, it is possible to set multiple stages. It is also possible that the bandwidth used has not been determined.
- FIGS. 68 to 74 are diagrams showing a state where a designated path is set based on the QoS information used in the present embodiment.
- Fig. 68 shows a state in which no data is flowing, and all data is transferred to the label switch network by a hop-by-hop method. Only path 907 is set. This path 907 may be set immediately after the data starts flowing. When its QOS Nono 0 of router 9 0 2 Power et al router 9 0 6 bound that with a not rather stringent information requirements Ke Tsu door mosquito S norator 9 0 3 to reach rather than to the child, the router 9 0 3 will be described later to set.
- the specified path setting information for each QOS information set by the method determine whether to set the specified path. If it is determined that the designated path is to be set, the designated path 908 with less severe QOS request is set and the state changes as shown in FIG. In this case, if it is determined not to set the designated path, the state shown in FIG. 68 is maintained.
- the packet is transferred via path 907 in a hop-by-hop manner.
- the bucket of the same destination, the same QOS, and the same destination port information is transferred to the designated path 908. Also, in this state, the same Q The same designated path is used when a packet with a different destination port is received in the os information.
- the specified path is set by referring to the specified path setting information.
- the designated path 909 is set, and the state changes as shown in FIG. Thereafter, packets of the same destination and the same destination port with the same QoS information are transferred to the designated path 909.
- the QOS request is severe (it does not matter if the same QOS information is the same as the previous QOS information), and a bucket with the same destination and a different destination port is received.
- the specified path is newly set by referring to the QOS information.
- the designated path 909 for which the QoS requirement is strict is set, and the state changes as shown in FIG. Thereafter, the same destination, the same destination port, and the same QoS are used.
- the packet is transferred to the designated path 909.
- the designated path of the corresponding information is set by referring to the designated path setting information.
- the designated path 909 for which the QOS requirement is strict is set, and the state changes as shown in FIG. Thereafter, packets with the same destination and the same destination port and the same QoS information are transferred to the designated path 909.
- the same destination and the same Q ⁇ S are not set.
- the packets are transferred in a hop-by-hop manner.
- a bucket with the same QOS request (it may be the same or not the same as the previous QOS information)
- the same destination and a different destination port is received.
- the designated path is newly set by referring to the Q ⁇ S information.
- a new designated path 909 is set and the state changes as shown in FIG.
- buckets with the same destination, same destination port, and same QoS are transferred to the designated path.
- Figure 75 is a diagram in which a designated path is set for each piece of QoS information at startup. In this way, at the time of start-up, it is specified for all QoS types. There is also a method in which a packet 907, 908, 909 is set, and when a certain packet is received, it is transferred to the designated path of the corresponding QoS.
- FIG. 76 and FIG. 77 are diagrams for explaining another setting method of the designated path. This section shows how to determine which path to set for each QOS when setting a specified path for each piece of QOS information.
- the information such as the bandwidth, delay, and discarding should be transmitted from the user terminal 910. I do.
- reference numeral 911 denotes a message from the user terminal 910 relating to designated path setting information for each piece of QOS information
- reference numeral 912 denotes a request notification message from the user terminal 910.
- FIG. 77 there is also a method in which a maintenance person inputs a message regarding designated path setting information for each piece of QoS information from the maintenance terminal 914.
- reference numeral 913 denotes a physical line for connecting a maintenance terminal.
- default information may be provided so that operation is possible even when no setting is made from anywhere.
- Fig. 78 is a diagram showing a state in which a designated path for the required conditions considered necessary at the time of startup is set, and a new designated path is generated when a data transmission request occurs.
- 915 is a designated path set for a specific request at the time of startup.
- Reference numeral 916 denotes a path not applicable to the specified path set at the time of startup, and is a specified path created when a data transmission request occurs.
- This method has an excellent advantage if the bandwidth can be secured from the beginning when starting up, especially from the viewpoint of securing the bandwidth.
- many types of requirements are not used, but in practice, a specified path can be set for each requirement.
- Figure 79 shows that in the header configuration of IPV4, the type of service (Type of Service) is used as the field into which the QOS information is inserted.
- Figure 80 shows the IP FIG. 10 is a diagram showing that a flow label (Flow Label) is used as a field into which QOS information is inserted in the header configuration of v6.
- Flow Label a flow label
- the QoS information of the bucket it is possible to guarantee the QoS information of the bucket and realize the transfer of the packet in consideration of the difference in service quality. For example, in the case of a bucket having a strict QoS requirement for which the time delay is unacceptable, setting a path for preferentially passing this bucket can be used. It can prevent the transfer of the packet from being delayed. Similarly, if the bucket cannot be discarded, a path for preferentially passing the bucket is set, thereby discarding the bucket. Can be prevented.
- the present invention can be applied to similar information, and a designated path can be set using such information. For example, a packet could be sent only to a specific destination with a specific quality.
- Routing protocols use data link On the Internet, IP, IPX, Apple Talk, and other network-based protocols are used. Routing protocols running on these protocols include RIP (Routing Information Protocol), OSPF (Open Shortest Path Fast), and BGP (Border Gateway Protocol).
- RIP Raster Information Protocol
- OSPF Open Shortest Path Fast
- BGP Border Gateway Protocol
- IP IP over ATM power S.
- a connection is set based on a routing table generated by a routing protocol as described later, the topology of the network is configured by the connection. It is called a topology driven method because it triggers a trigger for setting a session.
- a connection for routing information is set by the above-described IP over ATM in order to perform routing by a routing protocol used in the IP protocol.
- the upper layer has an ATM layer control function to set up an ATM connection by SVC (Switched Virtual Connection) and a function that the signaling processing part of SVC setting refers to the routing table entry. To be added to the ATM switch This is realized by and.
- the routing protocol used in the IP protocol is used as it is on the ATM switch, the SVC is used for setting the ATM connection, and the signaling processing unit performs the above-mentioned routing processing.
- the routing table generated by the routing protocol the connection setting in the end window can be performed at high speed and the number of connections can be reduced. And are possible.
- FIG. 81 is a diagram showing a configuration of an ATM network to which the tenth embodiment of the present invention is applied.
- the network shown in FIG. 81 is composed of ATM switches 1001 to: 1005 and ATM terminals 101:! To 107, and each of the switches is configured as follows. Switches and switches are connected by ATM lines.
- ATM switch 1001 will be described as an example.
- one IP over ATM connection is set for routing information between adjacent nodes and terminals (see Fig. 81).
- This connection can be set by PVC (Permanent Virtual Connection) or by ARP server or SVC.
- ATM switch 1001 sets up this connection, and then switches over IP routing over existing routing protocols such as RIP, OSPF, and BGP. This is done using IP at ATM.
- the routing information is exchanged between adjacent nodes and terminals via the management layer 1 1 1 4 (1 1 1 4 ', 1 1 1 4-1).
- 1 1 1 0 is generated.
- FIG. 82 shows only the exchange of routing information between the ATM switches 1001 and 1002 and between the switches 1001 and 1005. Also, it is assumed that the routing is performed in the adjacent ATM switch and ATM terminal in the same manner.
- 1 1 16 (1 1 1 6 ', 1 1 6-1) is a signaling processing section.
- ATM switch 1001 As a notifying table entry of ATM switch 1001 (destination: ATM terminal 106, relay node: ATM switch 1005) and relay ⁇ If the entry is completed, the ATM switch 1001 is sent to the ATM terminal 1005 by the connection setting protocol over IP. Request connection setting by SVC up to 6.
- connection setting request is based on the IP protocol (connection setting daemon) 111 over IP over ATM, and the ATM layer 11 1
- the ATM layer 1 1 1 3 sends a request to the signaling processing section 1 1 1 6 without sending it to the physical layer 1 1 1 4 .
- the ATM switches 1005 and 1004 of the relay nodes receive the same connection setting request by the normal SVC, and the signaling processing unit 111 ′ , 1116 "by directly referring to the routing table entries 1110 'and 1110 generated by the routing protocol over IP in IP over ATM.
- Next step to set up the connection Code shall be sought.
- the ATM switch 1001 is connected to an ATM terminal other than the directly connected ATM terminal (11014, 101, 115, 106, 1). 0 17 7) Automatically set connection (A, B, C, D) by SVC at the optimum route in the topology-driven manner.
- the ATM terminal 101:! ⁇ 1103 When there is a connection request by the SVC to the ATM terminal 106, the already set ATM terminal 1 Set the switch so that the data from each host will flow for the connection to 016. However, this requires that the ATM switch 1001 has the function of VC merging (matching multiple connections to one path), as shown in Figure 85. .
- the IP routing protocol can be used in the ATM switch, the end-to-end connection in the ATM network is possible. High-speed session setting and reduction in the number of connections can be realized at the same time. Industrial applicability
- a plurality of forging means each having a unique fading processing function, are provided for each output port of the ATM switch core. Then, for each of the sockets processed by each of the plurality of forwarding means and having the same address, Different destination information is assigned to each ATM packet that constitutes each bucket, and at the same time, the same ATM packet is assigned to the ATM cells.
- the predetermined amount of ATM cells is stored and output in a form in which the predetermined amount of ATM cells is stored.
- the first switch connected to the first port of the ATM switch.
- An ATM cell indicating that the destination information of the ATM cell is for forwarding is received via the ATM switch to perform IP packet forwarding processing, and the ATM switch is connected to the ATM cell.
- ATM indicating that the destination information of the ATM cell is for signaling processing via a second path which is physically different from the first path connected to the second port of the switch.
- the cell receives the cell, performs signaling processing of the IP packet, and is physically different from the first and second paths connected to the third port of the ATM switch.
- An ATM cell indicating that the destination information of the ATM cell is for management and control is received and managed and controlled via a different third path.
- the present invention it is possible to reduce the software processing of each function of the IP forwarding processing, the signaling processing, the management and control functions, and to reduce the IP packet processing.
- ⁇ Dating processing, signaling processing, and management.
- the efficiency of control processing can be improved, thereby increasing operating speed and processing speed (CPU performance, formal performance).
- high-speed, large-capacity data can be relayed.
- a signaling bucket is received from the ATM switch via the first link of the ATM switch, and signaling processing is performed by signaling processing means.
- the signaling processing means provided separately from the signaling processing means causes a second link physically different from the first link to pass through the second link. Then, the IP packet is received, and the IP packet transfer process is performed by hardware using predetermined memory.
- the net data by the CSR system In a packet relay device, especially when supporting SVC, signaling performance can be obtained, and the packet transfer capability at the time of hop-by-hop transfer is greatly improved. Can be done.
- a routing table is searched based on the header information of the cell input from the incoming line, and the header information of the input cell is rewritten to the information of the outgoing line to output the predetermined output line.
- the ATM switch to be exchanged a part of the header information of the input cell is masked, and the routing table is searched based on the masked header information. Then, based on the search result at this time, the header information of the input cell is rewritten to the outgoing side information specified in the routing table. And then
- the present invention it is possible to realize a network that unifies a plurality of VCCs into one by sharing outgoing exchange information of each cell input through each incoming line. And can be. This reduces the amount of VPI / VCI resources, especially in mission-critical networks, and enables a VCC merging method that does not hinder packet transfer. And can be.
- the ATM switch periodically detects the time when the cell is not conducting for each connection, and stores the non-conducting time of the detected cell.
- the stored non-conducting time of the cell is polled at a predetermined polling period, and the connection is performed based on the non-conducting time of the polled cell. Is being released.
- connection resources can be used effectively.
- the above-described routing processing and switching processing are performed in a relay apparatus that performs routing processing in a network layer and switching processing in a data link layer using an ATM switch.
- the input data is handled in packet units, the packet type is determined, and each packet is transmitted to the above-mentioned ATM switch according to the priority according to the packet type. ing.
- the routing process and the switch Prior to performing the tuning process the input data is sent in packets and sent to the ATM switch according to the priority according to the bucket type.
- High-priority buckets can be preferentially transferred and processed, and efficient data communication can be realized.
- the present invention also provides a high-speed bucket transfer mode using hardware using an ATM switch, and the same path is used for buckets having the same transfer destination address.
- an ATM relay device that transfers data in a shared manner, packets having the same transfer destination address are stored in different queues in PDU units according to the communication quality of each bucket. Then, buckets in PDU units stored in different queues are sorted in order of communication quality level, and are transferred using a single path.
- cells are sorted and merged and transferred in PDU units according to the communication quality, so that a power-through path is used for each communication quality class. It is possible to provide an ATM relay device having a dual mode capable of performing this.
- the packet transfer status monitoring device monitors the transfer status of the packet transferred to the destination by performing routing processing in the network layer. In order to predict the transfer tendency of the bucket based on the monitoring result at this time, and to shift to the hardware processing below the data link layer based on the predicted result at this time. We try to determine the conditions.
- the label switching technology is used.
- the router itself, which determines the conditions for transitioning to efficient cut-through, and automatically sets it, and also monitors the effectiveness of subsequent cut-through forwarding. Therefore, efficient operation of the network can be realized, and the load on the network administrator can be reduced.
- a router or a label switch host located at a boundary between a network constituted by a label switch and a network constituted by other devices is provided.
- the packet data to be transferred in the method of setting the designated path in the router located at the boundary between the same label switch network and another label switch network A step of referring to the QOS information or similar information in the header of the header, and setting of a designated path for the bucket data and unsetting are determined based on the referenced QOS information or similar information.
- the ATM switch has a SIP over ATM function at each node of the ATM network, and the connection is set by such an ATM switch.
- a connection for routing information between adjacent ATM switches and ATM terminals is set by IP over ATM, and after this connection is set, IP ov- er Performs routing according to a predetermined routing protocol over IP in ATM. Let's do it.
- the present invention it is possible to use the IP routing protocol in the ATM switch, so that the end-to-end in the ATM network can be used. High-speed connection setting and reduction in the number of connections can be realized at the same time.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Description
Claims
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CA002282441A CA2282441C (en) | 1997-12-25 | 1998-12-25 | Atm relay device and network including same |
US09/367,956 US6781994B1 (en) | 1997-12-25 | 1998-12-25 | Distributing ATM cells to output ports based upon destination information using ATM switch core and IP forwarding |
EP98961608A EP0967756A4 (en) | 1997-12-25 | 1998-12-25 | ATM INTERMEDIATE REGENERATOR AND NETWORK WITH THIS |
Applications Claiming Priority (24)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP35843497A JP3805880B2 (ja) | 1997-12-25 | 1997-12-25 | Atm中継装置 |
JP9/358448 | 1997-12-25 | ||
JP9/358447 | 1997-12-25 | ||
JP35844797A JP3805881B2 (ja) | 1997-12-25 | 1997-12-25 | Atmスイッチ |
JP9/358439 | 1997-12-25 | ||
JP9/358434 | 1997-12-25 | ||
JP35843997A JP3842416B2 (ja) | 1997-12-25 | 1997-12-25 | 指定パスの設定方法 |
JP9/358449 | 1997-12-25 | ||
JP9/358437 | 1997-12-25 | ||
JP9358448A JPH11191778A (ja) | 1997-12-25 | 1997-12-25 | 中継装置 |
JP35843797A JP3810914B2 (ja) | 1997-12-25 | 1997-12-25 | Atm中継装置 |
JP35844997A JP3842417B2 (ja) | 1997-12-25 | 1997-12-25 | Atmスイッチ |
JP9/366779 | 1997-12-26 | ||
JP36678397A JPH11196101A (ja) | 1997-12-26 | 1997-12-26 | パケット転送状態監視装置 |
JP36677997A JPH11196100A (ja) | 1997-12-26 | 1997-12-26 | Atmネットワークにおけるコネクション設定方法 |
JP9/366783 | 1997-12-26 | ||
JP36679897 | 1997-12-26 | ||
JP9/366784 | 1997-12-26 | ||
JP36678497A JP3805884B2 (ja) | 1997-12-26 | 1997-12-26 | Atm中継装置 |
JP9/366798 | 1997-12-26 | ||
JP20961498 | 1998-07-24 | ||
JP10/209614 | 1998-07-24 | ||
JP10/311369 | 1998-10-30 | ||
JP31136998A JP4044226B2 (ja) | 1997-12-26 | 1998-10-30 | Atm中継装置及びルータ装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO1999034558A1 true WO1999034558A1 (fr) | 1999-07-08 |
Family
ID=27583502
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP1998/005964 WO1999034558A1 (fr) | 1997-12-25 | 1998-12-25 | Repeteur mta et reseau le comprenant |
Country Status (4)
Country | Link |
---|---|
US (1) | US6781994B1 (ja) |
EP (1) | EP0967756A4 (ja) |
CA (1) | CA2282441C (ja) |
WO (1) | WO1999034558A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2001015386A2 (en) * | 1999-08-24 | 2001-03-01 | Telefonaktiebolaget Lm Ericsson (Publ) | Differentiated services provisioning for legacy systems |
Families Citing this family (48)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7664115B1 (en) * | 1999-04-30 | 2010-02-16 | Alcatel-Lucent Canada, Inc. | Method and apparatus for merging virtual connections |
WO2001028166A1 (fr) * | 1999-10-14 | 2001-04-19 | Fujitsu Limited | Procede de gestion de reseau |
ATE268097T1 (de) * | 1999-12-23 | 2004-06-15 | Nokia Corp | Verbindungsverwaltung in atm netzwerk und atm netzwerkeinheiten |
AU2001232784A1 (en) * | 2000-01-30 | 2001-08-07 | Celox Networks, Inc. | Device and method for packet inspection |
US6810031B1 (en) | 2000-02-29 | 2004-10-26 | Celox Networks, Inc. | Method and device for distributing bandwidth |
US6184409B1 (en) * | 2000-03-01 | 2001-02-06 | General Electric Company | Method and catalyst system for producing aromatic carbonates |
US7116680B1 (en) * | 2000-03-02 | 2006-10-03 | Agere Systems Inc. | Processor architecture and a method of processing |
US7215637B1 (en) * | 2000-04-17 | 2007-05-08 | Juniper Networks, Inc. | Systems and methods for processing packets |
US7075926B2 (en) * | 2000-05-24 | 2006-07-11 | Alcatel Internetworking, Inc. (Pe) | Programmable packet processor with flow resolution logic |
US7184440B1 (en) * | 2000-07-26 | 2007-02-27 | Alcatel Canada Inc. | Multi-protocol switch and method therefore |
US8213460B1 (en) * | 2000-09-07 | 2012-07-03 | Cisco Technology, Inc. | Method and system for processing traffic in an access network |
DE10045205A1 (de) | 2000-09-13 | 2002-03-28 | Siemens Ag | Verfahren zum Aufbau von Verbindungen mit vorgegebener Dienstgüte für ein paketorientiertes Kommunikationsnetz mit einem Resourcenmanager |
US7120159B2 (en) * | 2000-10-30 | 2006-10-10 | Matsushita Electric Industrial Co., Ltd. | Apparatus and method for packet transmission |
JP2002261807A (ja) * | 2001-02-28 | 2002-09-13 | Matsushita Electric Ind Co Ltd | Atmセルヘッダ変換回路及び方法 |
US6944159B1 (en) * | 2001-04-12 | 2005-09-13 | Force10 Networks, Inc. | Method and apparatus for providing virtual point to point connections in a network |
US7203170B2 (en) * | 2001-05-01 | 2007-04-10 | Integrated Device Technology, Inc. | Network switch port with weighted random early discard |
JP3808736B2 (ja) * | 2001-08-28 | 2006-08-16 | 株式会社エヌ・ティ・ティ・ドコモ | 多重伝送装置及び多重伝送方法 |
US7221678B1 (en) * | 2001-10-01 | 2007-05-22 | Advanced Micro Devices, Inc. | Method and apparatus for routing packets |
US7295563B2 (en) * | 2001-10-01 | 2007-11-13 | Advanced Micro Devices, Inc. | Method and apparatus for routing packets that have ordering requirements |
US7274692B1 (en) * | 2001-10-01 | 2007-09-25 | Advanced Micro Devices, Inc. | Method and apparatus for routing packets that have multiple destinations |
CA2410137C (en) * | 2001-11-02 | 2008-04-15 | Nippon Telegraph And Telephone Corporation | Optical dynamic burst switch |
US6856991B1 (en) * | 2002-03-19 | 2005-02-15 | Cisco Technology, Inc. | Method and apparatus for routing data to a load balanced server using MPLS packet labels |
JP3896879B2 (ja) * | 2002-03-22 | 2007-03-22 | 日本電気株式会社 | トラヒック監視システム |
KR100896616B1 (ko) * | 2002-06-27 | 2009-05-08 | 주식회사 케이티 | 큐를 이용한 트래픽 혼잡 제어 장치 및 방법 |
TW589832B (en) * | 2002-11-12 | 2004-06-01 | Inst Information Industry | QoS router for effectively processing fragmented packets and method thereof |
US7725886B2 (en) * | 2003-04-01 | 2010-05-25 | Intel Corporation | Merger of tables storing protocol data unit related data |
US7366184B2 (en) * | 2003-04-17 | 2008-04-29 | Alcatel | SVC/SPVC with L3 IP forwarding |
WO2004112318A1 (ja) * | 2003-06-11 | 2004-12-23 | Ntt Docomo, Inc. | パケット通信方法、制御装置及び移動局 |
US7621162B2 (en) * | 2003-12-30 | 2009-11-24 | Alcatel Lucent | Hierarchical flow-characterizing multiplexor |
US20050213595A1 (en) * | 2004-03-23 | 2005-09-29 | Takeshi Shimizu | Limited cyclical redundancy checksum (CRC) modification to support cut-through routing |
US20060114907A1 (en) * | 2004-11-30 | 2006-06-01 | Broadcom Corporation | Cut-through switching in a network device |
US7633972B1 (en) * | 2005-03-01 | 2009-12-15 | Nortel Networks Limited | Dynamic packet concatenation |
US7644147B1 (en) * | 2005-03-25 | 2010-01-05 | Marvell International Ltd. | Remote network device management |
US7839875B1 (en) * | 2005-07-12 | 2010-11-23 | Oracle America Inc. | Method and system for an efficient transport loopback mechanism for TCP/IP sockets |
US8842687B1 (en) * | 2006-06-30 | 2014-09-23 | Blue Coat Systems, Inc. | By-pass port facilitating network device failure detection in wide area network topologies |
CN101340364A (zh) * | 2007-07-04 | 2009-01-07 | 鸿富锦精密工业(深圳)有限公司 | 网络设备及其发送封包的方法 |
US8265056B2 (en) * | 2008-02-19 | 2012-09-11 | Qualcomm Incorporated | Packet decoding for H-ARQ transmission |
US8159944B2 (en) * | 2008-12-24 | 2012-04-17 | At&T Intellectual Property I, L.P. | Time based queuing |
US8493981B2 (en) * | 2010-11-03 | 2013-07-23 | Broadcom Corporation | Switch module |
US8787388B1 (en) * | 2011-08-29 | 2014-07-22 | Big Switch Networks, Inc. | System and methods for forwarding packets through a network |
JP6079348B2 (ja) * | 2013-03-22 | 2017-02-15 | 富士通株式会社 | スイッチ装置、スイッチ装置の制御方法、及びネットワークシステム |
CN104811379B (zh) * | 2015-04-29 | 2017-12-08 | 北京北信源软件股份有限公司 | 用于即时通信服务器互联互通的路由器寻址方法 |
CN104836734B (zh) * | 2015-05-11 | 2018-04-03 | 北京北信源软件股份有限公司 | 一种全新的即时通信路由方法和路由器 |
RU2651186C1 (ru) * | 2017-08-16 | 2018-04-18 | Общество с ограниченной ответственностью "Научно-Технический Центр ПРОТЕЙ" | Способ обмена данными и устройство управляющего узла сети |
US11275632B2 (en) | 2018-09-14 | 2022-03-15 | Advanced Micro Devices, Inc. | Broadcast command and response |
CN113010173A (zh) | 2019-12-19 | 2021-06-22 | 超威半导体(上海)有限公司 | 并行处理中矩阵数据广播的方法 |
CN113094099A (zh) | 2019-12-23 | 2021-07-09 | 超威半导体(上海)有限公司 | 矩阵数据广播架构 |
US11403221B2 (en) | 2020-09-24 | 2022-08-02 | Advanced Micro Devices, Inc. | Memory access response merging in a memory hierarchy |
Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05219098A (ja) * | 1991-09-13 | 1993-08-27 | Internatl Business Mach Corp <Ibm> | フレーム変換方法及び装置 |
JPH06141038A (ja) * | 1992-10-26 | 1994-05-20 | Hitachi Ltd | 分散交換機及び交換機構成方式 |
JPH0730554A (ja) * | 1993-07-13 | 1995-01-31 | Nec Corp | アドレス変換方式 |
JPH0766811A (ja) * | 1993-08-25 | 1995-03-10 | Toshiba Corp | Atmセル化装置およびatmデセル化装置 |
JPH07221762A (ja) * | 1994-01-27 | 1995-08-18 | Hitachi Ltd | パケット処理方法および通信インターフェイス装置 |
JPH0837530A (ja) * | 1994-07-22 | 1996-02-06 | Oki Electric Ind Co Ltd | タイムアウト監視装置 |
JPH08223181A (ja) * | 1995-02-17 | 1996-08-30 | Hitachi Ltd | Atm交換装置およびネットワーク間接続装置 |
JPH08288952A (ja) * | 1995-04-17 | 1996-11-01 | Nippon Telegr & Teleph Corp <Ntt> | パケット中継方法及び装置 |
JPH09149050A (ja) * | 1995-08-11 | 1997-06-06 | Fujitsu Ltd | データ受信装置 |
JPH09172457A (ja) * | 1995-12-20 | 1997-06-30 | Toshiba Corp | パケット送信ノード装置、パケット受信ノード装置及びコネクション設定方法 |
WO1997050276A1 (en) * | 1996-06-21 | 1997-12-31 | British Telecommunications Public Limited Company | Atm partial cut-through |
JPH1051465A (ja) * | 1996-08-06 | 1998-02-20 | Fujitsu Ltd | Atm網におけるサーバー予測型呼管理方式 |
JPH1065711A (ja) * | 1996-05-31 | 1998-03-06 | Mmc Networks Inc | Atmネットワーク内のatmセルをルートする方法及びatmセルをルートする装置 |
JPH10200533A (ja) * | 1997-01-08 | 1998-07-31 | Nec Corp | Atm vlanクライアント−サーバシステム |
JPH10313316A (ja) * | 1997-05-12 | 1998-11-24 | Nec Corp | Atmコネクションレス通信網におけるセッション管理方式及びコネクション管理方式 |
JPH1188427A (ja) * | 1997-09-05 | 1999-03-30 | Nippon Telegr & Teleph Corp <Ntt> | ルーティングテーブル検索装置 |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE59209115D1 (de) * | 1992-08-28 | 1998-02-12 | Siemens Ag | Verfahren und Schaltungsanordnung zum Übertragen von Nachrichtenzellen innerhalb eines ATM-Netzes |
JPH06132974A (ja) * | 1992-10-20 | 1994-05-13 | Toshiba Corp | パケット・ディスアセンブル用バッファ |
US5418781A (en) * | 1993-07-23 | 1995-05-23 | Digital Equipment Corporation | Architecture for maintaining the sequence of packet cells transmitted over a multicast, cell-switched network |
-
1998
- 1998-12-25 CA CA002282441A patent/CA2282441C/en not_active Expired - Fee Related
- 1998-12-25 WO PCT/JP1998/005964 patent/WO1999034558A1/ja active Application Filing
- 1998-12-25 US US09/367,956 patent/US6781994B1/en not_active Expired - Fee Related
- 1998-12-25 EP EP98961608A patent/EP0967756A4/en not_active Withdrawn
Patent Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05219098A (ja) * | 1991-09-13 | 1993-08-27 | Internatl Business Mach Corp <Ibm> | フレーム変換方法及び装置 |
JPH06141038A (ja) * | 1992-10-26 | 1994-05-20 | Hitachi Ltd | 分散交換機及び交換機構成方式 |
JPH0730554A (ja) * | 1993-07-13 | 1995-01-31 | Nec Corp | アドレス変換方式 |
JPH0766811A (ja) * | 1993-08-25 | 1995-03-10 | Toshiba Corp | Atmセル化装置およびatmデセル化装置 |
JPH07221762A (ja) * | 1994-01-27 | 1995-08-18 | Hitachi Ltd | パケット処理方法および通信インターフェイス装置 |
JPH0837530A (ja) * | 1994-07-22 | 1996-02-06 | Oki Electric Ind Co Ltd | タイムアウト監視装置 |
JPH08223181A (ja) * | 1995-02-17 | 1996-08-30 | Hitachi Ltd | Atm交換装置およびネットワーク間接続装置 |
JPH08288952A (ja) * | 1995-04-17 | 1996-11-01 | Nippon Telegr & Teleph Corp <Ntt> | パケット中継方法及び装置 |
JPH09149050A (ja) * | 1995-08-11 | 1997-06-06 | Fujitsu Ltd | データ受信装置 |
JPH09172457A (ja) * | 1995-12-20 | 1997-06-30 | Toshiba Corp | パケット送信ノード装置、パケット受信ノード装置及びコネクション設定方法 |
JPH1065711A (ja) * | 1996-05-31 | 1998-03-06 | Mmc Networks Inc | Atmネットワーク内のatmセルをルートする方法及びatmセルをルートする装置 |
WO1997050276A1 (en) * | 1996-06-21 | 1997-12-31 | British Telecommunications Public Limited Company | Atm partial cut-through |
JPH1051465A (ja) * | 1996-08-06 | 1998-02-20 | Fujitsu Ltd | Atm網におけるサーバー予測型呼管理方式 |
JPH10200533A (ja) * | 1997-01-08 | 1998-07-31 | Nec Corp | Atm vlanクライアント−サーバシステム |
JPH10313316A (ja) * | 1997-05-12 | 1998-11-24 | Nec Corp | Atmコネクションレス通信網におけるセッション管理方式及びコネクション管理方式 |
JPH1188427A (ja) * | 1997-09-05 | 1999-03-30 | Nippon Telegr & Teleph Corp <Ntt> | ルーティングテーブル検索装置 |
Non-Patent Citations (4)
Title |
---|
ESAKI H.: "HIGH SPEED IP PACKET FORWARDING OVER INTERNET USING ATM TECHNOLOGY.", OPTOMECHATRONIC MICRO/NANO DEVICES AND COMPONENTS III : 8 - 10 OCTOBER 2007, LAUSANNE, SWITZERLAND, SPIE, BELLINGHAM, WASH., vol. 2608., 1 January 1995 (1995-01-01), Bellingham, Wash., pages 44 - 51., XP002920000, ISBN: 978-1-62841-730-2, DOI: 10.1117/12.224189 * |
HORIKAWA K., IWATA A., SUZUKI H.: "NHRP ARCHITECTURE FOR LARGE SCALE ATM INTERNET.", IEICE TECHNICAL REPORT, DENSHI JOUHOU TSUUSHIN GAKKAI, JP, vol. 96., no. 215., 1 August 1996 (1996-08-01), JP, pages 49 - 54., XP002919997, ISSN: 0913-5685 * |
KATSUBE Y., NAGAMI K.-I., MATSUZAWA S.: "CELL SWITCH ROUTER - BASIS CONCEPT AND MIGRATION SCENARIO.", IEICE TECHNICAL REPORT, DENSHI JOUHOU TSUUSHIN GAKKAI, JP, 1 January 1996 (1996-01-01), JP, pages 173 - 178., XP002919999, ISSN: 0913-5685 * |
TAKASHI IKEDA, ET AL.: "ATM KOUKAN SYSTEM AX-4100/AX-1500", TOSHIBA REVIEW., TOSHIBA CORPORATION. TOKYO., JP, vol. 51, no. 09, 1 January 1996 (1996-01-01), JP, pages 39 - 42, XP002919998 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2001015386A2 (en) * | 1999-08-24 | 2001-03-01 | Telefonaktiebolaget Lm Ericsson (Publ) | Differentiated services provisioning for legacy systems |
WO2001015386A3 (en) * | 1999-08-24 | 2001-06-07 | Ericsson Telefon Ab L M | Differentiated services provisioning for legacy systems |
Also Published As
Publication number | Publication date |
---|---|
EP0967756A1 (en) | 1999-12-29 |
EP0967756A4 (en) | 2005-11-30 |
US6781994B1 (en) | 2004-08-24 |
CA2282441A1 (en) | 1999-07-08 |
CA2282441C (en) | 2004-07-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO1999034558A1 (fr) | Repeteur mta et reseau le comprenant | |
US6693909B1 (en) | Method and system for transporting traffic in a packet-switched network | |
US6553030B2 (en) | Technique for forwarding multi-cast data packets | |
JP4439122B2 (ja) | インターネット・プロトコル・コネクション指向サービスの管理方法 | |
US6147989A (en) | ATM communication system with high speed connection-less service function | |
US6859842B1 (en) | Method and apparatus for selection of paths on a communication network | |
US6587469B1 (en) | Telecommunications system | |
US7123620B1 (en) | Apparatus and method for scalable and dynamic traffic engineering in a data communication network | |
US6385204B1 (en) | Network architecture and call processing system | |
US20020085565A1 (en) | Technique for time division multiplex forwarding of data streams | |
US20020085567A1 (en) | Metro switch and method for transporting data configured according to multiple different formats | |
US20010053149A1 (en) | Method and system for quality of service (QoS) support in a packet-switched network | |
US20020085548A1 (en) | Quality of service technique for a data communication network | |
Aweya | On the design of IP routers Part 1: Router architectures | |
JPH11103298A (ja) | パケット伝送制御方法および装置 | |
US20020085507A1 (en) | Address learning technique in a data communication network | |
US20020085545A1 (en) | Non-blocking virtual switch architecture | |
WO2004112342A1 (en) | Emulated multi-qos links | |
US7061919B1 (en) | System and method for providing multiple classes of service in a packet switched network | |
JPH11103297A (ja) | パケット伝送制御方法および装置 | |
US7206310B1 (en) | Method and apparatus for replicating packet data with a network element | |
Cisco | Configuring Tag Switching and MPLS | |
JP3543953B2 (ja) | ラベル化パケット転送処理方法およびその通信装置 | |
JP3052924B2 (ja) | ショートカットルート確立システムおよびデータフロー検出装置 | |
WO2005050933A1 (en) | Point-to-point route monitoring in a packet-based core network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): CA US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): DE FR GB SE |
|
ENP | Entry into the national phase |
Ref document number: 2282441 Country of ref document: CA Ref country code: CA Ref document number: 2282441 Kind code of ref document: A Format of ref document f/p: F |
|
WWE | Wipo information: entry into national phase |
Ref document number: 09367956 Country of ref document: US |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 1998961608 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 1998961608 Country of ref document: EP |