WO1995001029A1 - Flow control method - Google Patents

Flow control method Download PDF

Info

Publication number
WO1995001029A1
WO1995001029A1 PCT/GB1994/001323 GB9401323W WO9501029A1 WO 1995001029 A1 WO1995001029 A1 WO 1995001029A1 GB 9401323 W GB9401323 W GB 9401323W WO 9501029 A1 WO9501029 A1 WO 9501029A1
Authority
WO
WIPO (PCT)
Prior art keywords
network
switch
indication
congested
congestion
Prior art date
Application number
PCT/GB1994/001323
Other languages
French (fr)
Inventor
John Spencer Arnold
Original Assignee
Gpt Limited
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Gpt Limited filed Critical Gpt Limited
Priority to AU69761/94A priority Critical patent/AU6976194A/en
Publication of WO1995001029A1 publication Critical patent/WO1995001029A1/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/54Store-and-forward switching systems 
    • H04L12/56Packet switching systems
    • H04L12/5601Transfer mode dependent, e.g. ATM
    • H04L12/5602Bandwidth control in ATM Networks, e.g. leaky bucket
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/54Store-and-forward switching systems 
    • H04L12/56Packet switching systems
    • H04L12/5601Transfer mode dependent, e.g. ATM
    • H04L2012/5629Admission control
    • H04L2012/563Signalling, e.g. protocols, reference model
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/54Store-and-forward switching systems 
    • H04L12/56Packet switching systems
    • H04L12/5601Transfer mode dependent, e.g. ATM
    • H04L2012/5629Admission control
    • H04L2012/5631Resource management and allocation
    • H04L2012/5632Bandwidth allocation
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/54Store-and-forward switching systems 
    • H04L12/56Packet switching systems
    • H04L12/5601Transfer mode dependent, e.g. ATM
    • H04L2012/5629Admission control
    • H04L2012/5631Resource management and allocation
    • H04L2012/5632Bandwidth allocation
    • H04L2012/5635Backpressure, e.g. for ABR
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/54Store-and-forward switching systems 
    • H04L12/56Packet switching systems
    • H04L12/5601Transfer mode dependent, e.g. ATM
    • H04L2012/5629Admission control
    • H04L2012/5631Resource management and allocation
    • H04L2012/5636Monitoring or policing, e.g. compliance with allocated rate, corrective actions
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/54Store-and-forward switching systems 
    • H04L12/56Packet switching systems
    • H04L12/5601Transfer mode dependent, e.g. ATM
    • H04L2012/5638Services, e.g. multimedia, GOS, QOS
    • H04L2012/5645Connectionless

Definitions

  • the second patent application describes a signalling system which utilises the same routing mechanism on the destination address but where the source address is replaced by one of a small number, for example three, reserved values which are used to identify one of several types of equipment unit which may be connected to a peripheral port (user port) of the network; including at least an Exchange Terminator (ET) adjacent to the switch, a Network Terminator (NT) at the user end of the connection link, and a Multiplexer between ET and NT: the source address is displaced to the cell payload.
  • ET Exchange Terminator
  • NT Network Terminator
  • a network of the kind described requires regulation of the traffic flows carried on shared bearers within the core of the network in order to ensure an equitable sharing of the available resource and a satisfactory quality of service.
  • One of the important advantages of this type of network is that there is no centralised control; the network is fully distributed and this principle needs also to be applied to traffic management; the principal mechanism used is therefore self-regulation.
  • the CCITT standard cell header includes a 3-bit "Payload Type” field which carries a "congested” indication: all switches are required to set this condition in cells when conditions within the switch, for example excessively large queues, suggests the approach of congestion on relevant bearers. The way in which this indication is to be used in not specified.
  • An AMiS network being a "connectionless” network does not carry channels recognisable to the core network and is thus totally dependent upon the operation of the self-regulating mechanism, thus a formal mechanism is necessary.
  • the 16 Virtual Chanels identified by the 4-bit reserved field are monitored in the upstream direction in the ET and "throttled" to a default value which is a proportion of the bandwidth of core network bearers P n , a typical value of which is 1/16.
  • the ET will examine the Payload Type field for the "congested" indication. Persistence of this indication in a significant proportion of cells received on a Virtual Channel will result in a signalling message being returned to the peer ET using the last received source address in a cell carrying the "congested" indication.
  • receipt of this message results in the value of P n being decremented by a value P ⁇ , typically 1/64. If no congestion messages are received in an interval T d the value of P n is incremented by a value P.,-, typically 1/256.
  • the decrement and increment algorithms have limits set for P n of typically 1/64 and 1/8 respectively, however these limits may be modified by instructions from the NT; for example if the user port is used for a Constant Bit Rate (CBR) service then P n may be frozen at some intermediate value.
  • CBR Constant Bit Rate
  • a method of self-regulating traffic to avoid congestion in an ATM network where the "congested" indication provided for in the Payload Type field of the standard ce 7 l format is set by a switch when the size of queues provided in the switch to statistically multiplexed switched inputs onto output ports suggests the approach of congestion and where the said "congested" indication is subject to a check at the receiving port of the network and a message is sent to the source port if the proportion of cells carrying the "congested" indication on a particular channel is relatively high and where, on receipt of such a message, the permitted bandwidth of the channel is reduced and, if no congestion messages are received over a specified period the permitted bandwidth is increased, said reductions and increases of bandwidth being in small increments between absolute limits.
  • a "disable self-regulation" message may be sent from NT to ET on the same port; in this case, "throttling" is held at the default value P ⁇ , typically 1/16 of the core bearer rate, and regulation is taken over by the NT.
  • the NT is, in fact, a sub-layer of an adaption unit and collaborates closely with the Segmentation And Reassembly (SAR) sub-layer.
  • the SAR sub-layer is required to recognise "Virtual Calls"; a Virtual Call is set-up on receipt of a cell with an unrecognised Source + VCI/MUX address and is terminated by receipt of a cell carrying the "end-of-message" indication in the Payload Type field.
  • the NT/SAR sub-layers are therefore able to provide self-regulation of Virtual Calls but at the price of increased complexity and hence cost; thus regulation at the ET is preferred for the majority of cases.
  • Each individual switch in such a network will interpret the Destination Address to a switch output port, thus successive cells between the same source and destination will follow the same path; this ensures that sequence integrity is retained.
  • the interpreter table at each switch input port (or common to all the inputs on the one switch) is set up and maintained by a Resource Manager which is an application run on a PC on any network port. In order to determine optimum routings the Resource Manager must have access to information on the traffic carried by each core network bearer.
  • the queue size indication used to set the "congestion" indication into the cell Payload Type field is not a good indication for this purpose since the relationship between traffic and queue size is highly non-linear and provides little useful guidance under normal operating conditions with relatively light loads. Instead, the running average occupancy of each switch output bearer is monitored and held in the switch control memory where it is available for polling by the Resource Manager; for this purpose a signalling address is allocated to each switch.
  • bandwidth is an expensive commodity which must be conserved; a broadband system will, however, use optical fibre interconnections at least in the core network and for fibre technology bandwidth is a relatively cheap commodity (and the cost is reducing as the technology matures).
  • the target load under normal operating conditions should be in the region of 40-60%.
  • the Resource Manager software provides a set of tools which will enable high occupancy bearers to be easily identified and paths to be re-routed; the software suite will also provide for storing approved changes which will be implemented during low traffic periods, say at night, since path changes can result in temporary loss of cells and loss of sequence integrity.
  • the self-regulating mechanism previously described can only operate effectively in a well-balanced network thus it is important that a periodic check is made by the Resource Manager; the self-regulating mechanism will however tend to distort the results. Because of the non-linear relationship between queue size and bearer occupancy, this will only be a problem if the network becomes heavily loaded or seriously unbalanced and for this case a broadcast message is sent to all ETs to instruct them to turn off the self-regulator and set the bandwidth to the default value P d .

Abstract

In an ATM network, traffic can be self-regulated to avoid congestion by a method where the 'congested' indication provided for in the Payload Type field of the standard cell format is set by a switch when the size of queues provided in the switch to statistically multiplexed switched inputs onto output ports suggests the approach of congestion and where the said 'congested' indication is subject to a check at the receiving port of the network and a message is sent to the source port if the proportion of cells carrying the 'congested' indication on a particular channel is relatively high and where, on receipt of such a message, the permitted bandwidth of the channel is reduced and, if no congestion messages are received over a specified period the permitted bandwidth is increased, said reduction and increase of bandwidth being in small increments between absolute limits.

Description

FLOW CONTROL METHOD
A concept for an intrinsically connectionless ATM network is described in Patent Application No. GB 2,255,257A "ATM Mini-Switch for Business Systems": addressing in this solution closely follows International Telegraph and Telephone Consultative Committee (CCITT) standards with the Destination Address (which defines the path through the network) carried in the 12-bit VPI field of the NNI format and the Source Address (which defines the "channel" at the destination) carried in the most-significant 12 bits of the VCI field, the least significant 4 bits being reserved to provide up to 16 Virtual Channels between source and destination or to provide source and/or destination addresses for 4-way multiplexers.
The above solution has the disadvantage that the number of addresses available is limited to 4,096 and a second solution is described in Patent Application No. GB 9300279.8, published as GB2274227A, which uses the available address space in the standard cell header format more efficiently to provide for 32,678 possible addresses (still retaining the reserved 4 bits) at the cost of ignoring the standard division between VPI and VCI fields.
The second patent application describes a signalling system which utilises the same routing mechanism on the destination address but where the source address is replaced by one of a small number, for example three, reserved values which are used to identify one of several types of equipment unit which may be connected to a peripheral port (user port) of the network; including at least an Exchange Terminator (ET) adjacent to the switch, a Network Terminator (NT) at the user end of the connection link, and a Multiplexer between ET and NT: the source address is displaced to the cell payload. By this means signalling messages may be transported between a network manager and peripheral units, between peripheral units on the one network port, and between peer units on different network ports.
A network of the kind described requires regulation of the traffic flows carried on shared bearers within the core of the network in order to ensure an equitable sharing of the available resource and a satisfactory quality of service. One of the important advantages of this type of network is that there is no centralised control; the network is fully distributed and this principle needs also to be applied to traffic management; the principal mechanism used is therefore self-regulation.
The CCITT standard cell header includes a 3-bit "Payload Type" field which carries a "congested" indication: all switches are required to set this condition in cells when conditions within the switch, for example excessively large queues, suggests the approach of congestion on relevant bearers. The way in which this indication is to be used in not specified.
An AMiS network, being a "connectionless" network does not carry channels recognisable to the core network and is thus totally dependent upon the operation of the self-regulating mechanism, thus a formal mechanism is necessary.
In order to ensure good statistical smoothing within the network, the 16 Virtual Chanels identified by the 4-bit reserved field (multiplexer ports are also treated as Virtual Channels) are monitored in the upstream direction in the ET and "throttled" to a default value which is a proportion of the bandwidth of core network bearers Pn, a typical value of which is 1/16.
At the destination port the ET will examine the Payload Type field for the "congested" indication. Persistence of this indication in a significant proportion of cells received on a Virtual Channel will result in a signalling message being returned to the peer ET using the last received source address in a cell carrying the "congested" indication. At the source ET receipt of this message results in the value of Pn being decremented by a value P^, typically 1/64. If no congestion messages are received in an interval Td the value of Pn is incremented by a value P.,-, typically 1/256. The decrement and increment algorithms have limits set for Pn of typically 1/64 and 1/8 respectively, however these limits may be modified by instructions from the NT; for example if the user port is used for a Constant Bit Rate (CBR) service then Pn may be frozen at some intermediate value.
The mechanism described is illustrated in the accompanying single figure.
According to the present invention there is provided a method of self-regulating traffic to avoid congestion in an ATM network where the "congested" indication provided for in the Payload Type field of the standard ce7l format is set by a switch when the size of queues provided in the switch to statistically multiplexed switched inputs onto output ports suggests the approach of congestion and where the said "congested" indication is subject to a check at the receiving port of the network and a message is sent to the source port if the proportion of cells carrying the "congested" indication on a particular channel is relatively high and where, on receipt of such a message, the permitted bandwidth of the channel is reduced and, if no congestion messages are received over a specified period the permitted bandwidth is increased, said reductions and increases of bandwidth being in small increments between absolute limits.
Note that a "throttle" is provided at the source end for each possible value of the 4-bit VCI/MUX field; similarly, persistence of the "congested" indication is separately checked for each value of the VCI/MUX -^ld at the destination end. This solution is not perfect since the source, the same VCI/MUX value may be used to several destina s not all of which may be congested and, at the destination, the sa... i/CI/MUX field value may be received from several sources. Nevertheless this mechanism is considered to be adequate for the majority of uses of network ports.
In order to cater for network ports where tighter discrimination is considered to be necessary a "disable self-regulation" message may be sent from NT to ET on the same port; in this case, "throttling" is held at the default value P^, typically 1/16 of the core bearer rate, and regulation is taken over by the NT. The NT is, in fact, a sub-layer of an adaption unit and collaborates closely with the Segmentation And Reassembly (SAR) sub-layer. The SAR sub-layer is required to recognise "Virtual Calls"; a Virtual Call is set-up on receipt of a cell with an unrecognised Source + VCI/MUX address and is terminated by receipt of a cell carrying the "end-of-message" indication in the Payload Type field. The NT/SAR sub-layers are therefore able to provide self-regulation of Virtual Calls but at the price of increased complexity and hence cost; thus regulation at the ET is preferred for the majority of cases.
Each individual switch in such a network will interpret the Destination Address to a switch output port, thus successive cells between the same source and destination will follow the same path; this ensures that sequence integrity is retained. The interpreter table at each switch input port (or common to all the inputs on the one switch) is set up and maintained by a Resource Manager which is an application run on a PC on any network port. In order to determine optimum routings the Resource Manager must have access to information on the traffic carried by each core network bearer. The queue size indication used to set the "congestion" indication into the cell Payload Type field is not a good indication for this purpose since the relationship between traffic and queue size is highly non-linear and provides little useful guidance under normal operating conditions with relatively light loads. Instead, the running average occupancy of each switch output bearer is monitored and held in the switch control memory where it is available for polling by the Resource Manager; for this purpose a signalling address is allocated to each switch.
A conventional wisdom in traditional telecommunications systems is that bandwidth is an expensive commodity which must be conserved; a broadband system will, however, use optical fibre interconnections at least in the core network and for fibre technology bandwidth is a relatively cheap commodity (and the cost is reducing as the technology matures). Thus, in order to provide the network with resilience to short-term traffic peaks, the target load under normal operating conditions should be in the region of 40-60%. The Resource Manager software provides a set of tools which will enable high occupancy bearers to be easily identified and paths to be re-routed; the software suite will also provide for storing approved changes which will be implemented during low traffic periods, say at night, since path changes can result in temporary loss of cells and loss of sequence integrity.
The self-regulating mechanism previously described can only operate effectively in a well-balanced network thus it is important that a periodic check is made by the Resource Manager; the self-regulating mechanism will however tend to distort the results. Because of the non-linear relationship between queue size and bearer occupancy, this will only be a problem if the network becomes heavily loaded or seriously unbalanced and for this case a broadcast message is sent to all ETs to instruct them to turn off the self-regulator and set the bandwidth to the default value Pd.

Claims

1. A method of self-regulating traffic to avoid congestion in an Asynchronous Transfer Mode (ATM) network where the "congested" indication provided for in the Payload Type field of the standard cell format is set by a switch when the size of queues provided in the switch to statistically multiplexed switched inputs onto output ports suggests the approach of congestion and where the said "congested" indication is subject to a check at the receiving port of the network and a message is sent to the source port if the proportion of cells carrying the "congested" indication on a particular channel is relatively high and where, on receipt of such a message, the permitted bandwidth of the channel is reduced and, if no congestion messages are received over a specified period the permitted bandwidth is increased, said reduction and increases of bandwidth being in small increments between absolute limits.
2. An ATM network using the self-regulating mechanism of Claim 1 of the "direct-connectionless" type where the absolute source and destination address is carried in the header address field of each cell and where the destination address is interpreted in each switch in the path to a switch output port address so that successive cells between the same source and destination will follow the same path so retaining sequence integrity and where a Resource Manager is provided which has signalling access to all switches in the network, said switches providing means to determine the running average occupancy of each switch port and to hold this information for polling by the Resource Manager and where the said Resource Manager tests the network periodically for excessive bearer loading, makes recommends on the re-routing of paths through the network and stores approved recommendations for implementation at a time when the traffic in the network is low.
3. An ATM network using the self-regulating mechanism of Claim 1 of the "connection-mode" type where the header address field carries a "Virtual Path" identifier and a "Virtual Channel" identifier, said identifiers being labels which identify respectively a path or a channel on a single bearer and where the identifier is translated to a new unique value on the outgoing bearer of a switch using values held in a table in each switch, said table values being set up according to information supplied by signalling messages at the commencement of a call and where the values are determined by a routing algorithm utilising information supplied by the user on the characteristics of the service required when the call demand is originated and the average occupancy of relevant network bearers, said bearer occupancies being maintained in an up-to-date state by polling, from time-to-time, all switches in the network for the running average occupancy on ports of said switches.
PCT/GB1994/001323 1993-06-21 1994-06-20 Flow control method WO1995001029A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU69761/94A AU6976194A (en) 1993-06-21 1994-06-20 Flow control method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
GB9312741.3 1993-06-21
GB939312741A GB9312741D0 (en) 1993-06-21 1993-06-21 Nitwork management

Publications (1)

Publication Number Publication Date
WO1995001029A1 true WO1995001029A1 (en) 1995-01-05

Family

ID=10737497

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/GB1994/001323 WO1995001029A1 (en) 1993-06-21 1994-06-20 Flow control method

Country Status (4)

Country Link
AU (1) AU6976194A (en)
GB (2) GB9312741D0 (en)
WO (1) WO1995001029A1 (en)
ZA (1) ZA944435B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0767595A1 (en) 1995-10-06 1997-04-09 ALCATEL BELL Naamloze Vennootschap Method and apparatus for shaping and multiplexing bursty traffic
EP0909106A2 (en) * 1997-10-10 1999-04-14 Mitsubishi Denki Kabushiki Kaisha Implementation schemes for a queue length based ABR flow control system
US5987007A (en) * 1996-02-14 1999-11-16 Nokia Telecommunications Oy Manipulation of header field in ATM cell
US6026075A (en) * 1997-02-25 2000-02-15 International Business Machines Corporation Flow control mechanism
CN1086883C (en) * 1996-09-02 2002-06-26 三星电子株式会社 Apparatus for noticing and comtrolling crowd condition in asynchronous transmission type network and method thereof

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2337895B (en) * 1995-06-05 2000-01-19 Fujitsu Ltd Detecting congestion in ATM networks
GB9511319D0 (en) * 1995-06-05 1995-08-02 Gen Datacomm Adv Res Controlling the flow of ATM cells in an ATM network
GB2306849B (en) * 1995-10-30 1999-10-20 Gen Datacomm Adv Res Atm network switch
JPH10190733A (en) 1996-12-25 1998-07-21 Hitachi Ltd Ip switch, interface circuit and atm switch used for the ip switch and ip switch network system
US6667985B1 (en) 1998-10-28 2003-12-23 3Com Technologies Communication switch including input bandwidth throttling to reduce output congestion
GB9823489D0 (en) * 1998-10-28 1998-12-23 3Com Technologies Ltd Communication switch including input bandwidth throttling to reduce output congestion
GB0321570D0 (en) 2003-09-15 2003-10-15 British Telecomm Inter-domain congestion charging
US7907519B2 (en) 2005-01-28 2011-03-15 British Telecommunications Plc Packet forwarding

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2753294B2 (en) * 1988-12-23 1998-05-18 株式会社日立製作所 Packet congestion control method and packet switching device
GB8914983D0 (en) * 1989-06-29 1989-08-23 Digital Equipment Int Congestion control in computer networks
US5313454A (en) * 1992-04-01 1994-05-17 Stratacom, Inc. Congestion control for cell networks

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
D.E. COMER: "A rate-based congestion avoidance and control scheme for packet switched netwoks", 10TH INTERNATIONAL CONFERENCE ON DISTRIBUTED COMPUTER SYSTEMS, June 1990 (1990-06-01), USA, pages 390 - 397 *
D.M. CHIU ET AL.: "Analysis of the increase and decrease algorithm for congestion avoidance in computer networks", COMPUTER NETWORKS AND ISDN SYSTEMS, vol. 17, no. 1, June 1989 (1989-06-01), AMSTERDAM NL, pages 1 - 14 *
D.W. GLAZER ET AL.: "A new metric for dynamic routing algorithms", IEEE TRANSACTIONS ON COMMUNICATIONS, vol. 38, no. 3, March 1990 (1990-03-01), NEW YORK, US, pages 360 - 367 *
H. SAITO ET AL.: "Traffic control technologies in ATM networks", TRANSACTIONS OF THE INSTITUTE OF ELECTRONICS AND COMMUNICATION ENGINEERS OF JAPAN, SECTION E, vol. E74, no. 4, April 1991 (1991-04-01), TOKYO JP, pages 761 - 771 *
L. MONGIOVI ET AL.: "A proposal for inteconnecting FDDI networks through B_ISDN", IEEE INFOCOM '91, vol. 3, April 1991 (1991-04-01), NEW YORK, USA, pages 1160 - 1167 *

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0767595A1 (en) 1995-10-06 1997-04-09 ALCATEL BELL Naamloze Vennootschap Method and apparatus for shaping and multiplexing bursty traffic
US5898670A (en) * 1995-10-06 1999-04-27 Alcatel N.V. Bursty traffic multiplexing arrangement and method for shaping and multiplexing bursty input flows
AU725085B2 (en) * 1995-10-06 2000-10-05 Alcatel A bursty traffic multiplexing arrangement
EP0767595B1 (en) * 1995-10-06 2001-04-04 Alcatel Method and apparatus for shaping and multiplexing bursty traffic
US5987007A (en) * 1996-02-14 1999-11-16 Nokia Telecommunications Oy Manipulation of header field in ATM cell
CN1086883C (en) * 1996-09-02 2002-06-26 三星电子株式会社 Apparatus for noticing and comtrolling crowd condition in asynchronous transmission type network and method thereof
US6026075A (en) * 1997-02-25 2000-02-15 International Business Machines Corporation Flow control mechanism
US7002911B1 (en) 1997-02-25 2006-02-21 International Business Machines Corporation Flow control mechanism
EP0909106A2 (en) * 1997-10-10 1999-04-14 Mitsubishi Denki Kabushiki Kaisha Implementation schemes for a queue length based ABR flow control system
EP0909106A3 (en) * 1997-10-10 2000-04-12 Mitsubishi Denki Kabushiki Kaisha Implementation schemes for a queue length based ABR flow control system

Also Published As

Publication number Publication date
GB9412331D0 (en) 1994-08-10
GB2281005A (en) 1995-02-15
ZA944435B (en) 1995-02-15
GB9312741D0 (en) 1993-08-04
AU6976194A (en) 1995-01-17

Similar Documents

Publication Publication Date Title
US5949757A (en) Packet flow monitor and control system
US5864539A (en) Method and apparatus for a rate-based congestion control in a shared memory switch
US5629928A (en) Dynamic fair queuing to support best effort traffic in an ATM network
US6549938B1 (en) System and method for prioritizing multicast packets in a network service class utilizing a priority-based quality of service
US6411627B1 (en) Switching protocol providing controlled access to available asynchronous network service
US7920471B2 (en) Call admission control method and system
EP0781068A1 (en) Method and system for adaptive bandwidth allocation in a high speed data network
EP0512495A2 (en) Switching node in a network with label multiplexed information
US5914937A (en) B-ISDN access
WO1995001029A1 (en) Flow control method
US6088359A (en) ABR server
EP0838970B1 (en) Method for shared memory management in network nodes
US20040076118A1 (en) Explicit rate flow control for multicast connections
US8369320B2 (en) Multiplexing apparatus and cell discard method
US6510165B1 (en) Band controlling apparatus
EP1145588B1 (en) Mechanism and method for dynamically allocating atm connections between exchanges
GB2255257A (en) Telecommunications switching
CA2296688C (en) Method for optimizing the load level on connecting sections for vbr traffic
JP3001052B2 (en) ATM communication device
JPH11136252A (en) Atm switch and shaping method
JP2002237842A (en) Communication network system and line concentrating node device
JPH09261245A (en) Atm switch
JP2001119411A (en) 0/1 system selection circuit
Cummings et al. An ATM cell format and media access protocol for a network of switch-interconnected shared-medium ATM rings
KR980013086A (en) How to distinguish the signal channel of an ATM switch

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AM AT AU BB BG BR BY CA CH CN CZ DE DK ES FI GB GE HU JP KE KG KP KR KZ LK LU LV MD MG MN MW NL NO NZ PL PT RO RU SD SE SI SK TJ TT UA US UZ VN

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

122 Ep: pct application non-entry in european phase
NENP Non-entry into the national phase

Ref country code: CA