IP/ATM network system adapted for the simultaneous transmission of IP data packets to a plurality of users

• US 6,831,918 B1
• Filed: 09/01/2000
• Issued: 12/14/2004
• Est. Priority Date: 12/01/1997
• Status: Expired due to Fees

• Current Assignee: Teliasonera AB
• Original Assignee: Telia AB (Government of Norway)
• Inventors: Kavak, Nail
• Primary Examiner(s): Nguyen, Chau
• Assistant Examiner(s): Lam, Daniel K.
• Application Number:
  US09/554,613
• Time in Patent Office: 1,565 Days
• Field of Search: 370/395.1, 370/395.51, 370/395-2-3, 370/401
• US Class Current: 370/395.52
• CPC Class Codes: H04L 2012/5621 Virtual private network [VPN]; Private-network - network-interface (P-NNI)
• CPC Class Codes: H04L 2012/5627 Fault tolerance and recovery
• CPC Class Codes: H04L 2012/5642 Multicast/broadcast/point-multipoint, e.g. VOD
• CPC Class Codes: H04L 2012/5667 IP over ATM
• CPC Class Codes: H04Q 11/0478 Provisions for broadband connections

Abstract

An ATM transmission system is adapted for the simultaneous transmission (i.e. multicasting) of IP data packets to a plurality of system users, using ATM as a carrier network, and includes an IP/ATM network including clusters of Logical IP Systems (LISs), a multicast address resolution server (MARS) for each cluster, and a multicast router (MCR) for each cluster. The clusters are interconnected through the MCRs which form an inter-cluster backbone for the IP/ATM network. Each cluster includes an integral number of ATM hosts and LISs, the size of a cluster is dependent on physical and administrative constraints, and each LIS, within a cluster, is adapted to support a number of ATM hosts. Multicasting of IP data packets within a cluster utilizes a MARS protocol, and multicasting between clusters utilizes an inter-cluster multicasting protocol (ILMP). All ATM hosts within a cluster, even if supported by different LISs, are adapted to be configured with an address of the same MARS.

Claims

1. An ATM transmission system, adapted for the simultaneous transmission of IP data packets to a plurality of system users, using ATM as a carrier network, said transmission system includes:

an IP/ATM network including clusters of Logical IP Systems (LISs);

a multicast address resolution server (MARS) for each cluster; and

a multicast router (MCR) for each cluster, in that said clusters are interconnected through said MCRs which form an inter-cluster backbone for said IP/ATM network, in that multicasting of IP data packets within a cluster utilizes a MARS protocol, and in that multicasting between clusters utilizes an inter-cluster multicasting protocol (ILMP),

wherein a MCR is adapted, in advance of multicasting, to select a well known end-node using a selection algorithm that returns a same well known end-node for an inter-cluster multicast group when executed by any of said MCRs and to establish a connection to the well known end-node, said well known end-node being adapted to provide a database for ATM addresses of members of the inter-cluster multicast group, and in that said well known end-node is adapted to establish a point-to-multipoint tree and to distribute multicast traffic sent by a member of said inter-cluster multicast group to other members of said group using said tree.

2. An ATM transmission system, as claimed in claim 1, further comprising an entry point configured to provide access to the Internet.

3. An ATM transmission system, as claimed in claim 2, wherein one of said MCRs is adapted for use as the entry point to the Internet.

4. An ATM transmission system, as claimed in claim 2, wherein said entry point to the Internet includes a non-ATM IP Subnet connected to said inter-cluster backbone.

5. An ATM transmission system, as claimed in claim 1, wherein each cluster consists of an integral number of ATM hosts and LISs, and a size of a cluster being dependent on physical and administrative constraints, and in that each LIS, within a cluster, is adapted to support a number of ATM hosts, and in that all ATM hosts within a cluster, even if supported by different LISs, are adapted to be configured with an address of a same MARS.

6. An ATM transmission system, as claimed in claim 1, wherein, for each cluster, functionality of MARS and MCR is adapted to be co-located.

7. An ATM transmission system, as claimed in claim 1, wherein, in the presence of multiple MCRs belonging to a same cluster, said IP/ATM network is adapted to select a MCR with a highest IP address as a default MCR.

8. An ATM transmission system, as claimed in claim 1, wherein MARS is adapted to operate in server mode.

9. An ATM transmission system, as claimed in claim 1, wherein said MCRs are adapted to promiscuously receive all multicast traffic on a cluster.

10. An ATM transmission system, as claimed in claim 1, wherein said MCRs are adapted to support point-to-multipoint connections provided by UNI 3.0 signalling.

11. An ATM transmission system, as claimed in claim 1, wherein all ATM hosts and MCRs, associated with a same cluster, are adapted to be configured with an address of a same MARS, in that MARS is adapted to operate in server mode, in that said IP/ATM network is adapted to enable IP/ATM endpoints, within a cluster, to register as a member of an inter-cluster multicast group utilizing MARS procedures, in that said MCRs are adapted to support point-to-multipoint connections provided by UNI 3.0 signalling, and in that a multicast server is provided and adapted to manage said point-to-multipoint connections, and to distribute multicast traffic therefrom to all other leaves of said point-to-multipoint connections.

12. An ATM transmission system, as claimed in claim 1, wherein each group address has a corresponding well known end-node address.

13. An ATM transmission system, as claimed in claim 1, wherein said IP/ATM network is adapted to enable multicast packets to be sent to members of an inter-cluster multicast group, even if a sender is not a member of said group.

14. An ATM transmission system, as claimed in claim 1, wherein said MCRs are adapted to operate in either a server mode, or a Virtual Circuit (VC) mesh mode, dependent on criteria, including a number of inter-cluster multicast groups, a grouping mode, traffic density, charging policies of a network operator, performance, reliability, and system complexity.

15. An ATM transmission system, as claimed in claim 14, wherein said MCRs are adapted to operate in the server mode, wherein said well-known end-node is adapted to provide the database for ATM addresses for members of the inter-cluster multicast group, in that said MCRs are adapted to issue multicast packets to said well known end-node for distribution to members of the inter-cluster multicast group, and in that said well known end-node is adapted to establish the point-to-multipoint tree to said group members and to distributing multicast packets, received from said MCRs, to said group members using said tree.

16. An ATM transmission system, as claimed in claim 14, wherein said MCRs are adapted to operate in the VC mesh mode wherein said well known end-nodes are adapted to keep track of IP multicast addresses, and corresponding ATM members of said MCRs, and said IP/ATM network is adapted to report changes in membership of an MCR to a respective well known end-node, and to establish point-to-multipoint trees between MCRs belonging to a same inter-cluster multicast group, when an association between said IP multicast addresses and said ATM members has been determined.

17. An ATM transmission system, as claimed in claim 16, wherein said IP/ATM network includes a first control VC and a second control VC between the MCR and said well-known end-node, wherein the first control VC is a transient bi-directional point-to-point connection between the MCR and said well-known end-node, and is adapted for use in registering group membership, and wherein the second control VC is a unidirectional point-to-multipoint group-ctrl-VC from said well-known end-node to the MCR, and is adapted to propagate group joining and leaving messages to the MCR.

18. An ATM transmission system, as claimed in claim 12, wherein said IP/ATM network is adapted to enable a multicast packet originating from a non-group member to be distributed to group members using a point-to-multipoint tree of a group member MCR and said well known end-node is adapted to establish a point-to-point connection between said non-group member and said group member MCR to facilitate a transfer of said multicast packet to said group member MCR and thereafter to said group members via said point-to-multipoint tree.

19. An ATM transmission system, as claimed in claim 1, wherein said IP/ATM network is adapted to enable a MCR to join, or leave, an inter-cluster multicast group, at an option of the MCR.

20. In an ATM transmission system including an IP/ATM network, said network including clusters of Logical LP Systems (LISs), a multicast address resolution server (MARS) for each cluster and a multicast router (MCR) for each cluster, a method for a simultaneous transmission of IP data packets to a plurality of system users using ATM as a carrier network, said clusters are interconnected through said MCRs to form an inter-cluster backbone for said IP/ATM network, by using a MARS protocol for multicasting said IP data packets within a cluster, and by using an inter-cluster multicasting protocol (ILMP) for multicasting said IP data packets between clusters, said method comprising:

selecting a well known end-node using a selection algorithm that returns a same well known end-node for an inter-cluster multicast group when executed by any of said MCRs, in advance of multicasting;

establishing a connection to said well known end-node in advance of multicasting, said well known end-node providing a database for ATM addresses of members of an inter-cluster multicast group; and

said well known end-node;

establishing a point-to-multipoint tree to said ATM addresses; and

using said tree to distribute a multicast package to other members of said group on receipt of a multicast packet from a member of said inter-cluster multicast group.

21. The method, as claimed in claim 20, further comprising obtaining access to the Internet for said IP/ATM network.

22. The method, as claimed in claim 20, wherein said IP/ATM network, in a presence of multiple MCRs belonging to a same cluster, selecting a MCR with a highest IP address as a default MCR.

23. The method, as claimed in claim 20, further comprising MARS operating in server mode.

24. The method, as claimed in claim 20, further comprising said MCRs promiscuously receiving all multicast traffic on a cluster.

25. The method, as claimed in claim 20, further comprising said MCRs supporting point-to-multipoint connections provided by UNI 3.0 signalling.

26. The method, as claimed in claim 20, further comprising:

configuring all ATM hosts and MCRs, associated with a same cluster, with an address of a same MARS;

operating the same MARS in server mode;

IP/ATM endpoints, within a cluster, utilizing MARS procedures to register as a member of the inter-cluster multicast group;

said MCRs supporting point-to-multipoint connections provided by UNI 3.0 signalling; and

managing said point-to-multipoint connections and distributing multicast traffic to all other leaves of said point to-multipoint connections using a multicast server.

27. The method, as claimed in claim 20, further comprising each group address having a corresponding well known end-node address.

28. The method, as claimed in claim 20, further comprising:

said well known end-node forming a root of a point-to-multipoint tree when said MCRs are operating in a server mode; and

an MCR, wishing to distribute a multicast packet to members of an existing group of which it is not a member;

establishing a point-to-point connection to said well-known end-node; and

sending said multicast packet to said well-known end-node using the point-to-point connection; and

on receipt of said multicast packet, said well known end-node distributing said multicast packet, via an existing point-to-multipoint tree to said group members.

29. The method, as claimed in claim 20, further comprising said MCRs operating in either a server mode, or a Virtual Circuit (VC) mesh mode, dependent on criteria, including a number of inter-cluster multicast groups, a grouping mode, traffic density, charging policies of a network operator, performance, reliability and system complexity.

30. The method, as claimed in claim 29, further comprising:

said MCRs operating in a server mode;

said well-known end-node providing the database for ATM addresses for members of the inter-cluster multicast group;

said MCRs issuing multicast packets to said well known end-node for distribution to members of the inter-cluster multicast group; and

said well known end-node;

establishing a point-to-multipoint tree to said group members; and

using said tree to distribute multicast packets, received from said MCRs, to said group members.

31. The method, as claimed in claim 29, further comprising:

said MCRs operating in the VC mesh mode;

said well known end-nodes keeping track of IP multicast addresses, and corresponding ATM members of said MCRs; and

said IP/ATM network;

reporting changes in membership of an MCR to a respective well known end-node; and

on determining an association between said IP multicast addresses and said ATM members, establishing point-to-multipoint trees between MCRs belonging to a same inter-cluster multicast group.

32. The method, as claimed in claim 31, said IP/ATM network performing further steps comprising:

establishing a first control VC and a second control VC between the MCR and said well-known end-node, the first control VC being a transient bi-directional point-to-point connection between the MCR and said well-known end-node and the second control VC being a unidirectional point-to-multipoint group-ctrl-VC from said well-known end-node to the MCR;

using said first control VC to register membership of the inter-cluster multicast group; and

using said second control VC to propagate group joining and leaving messages to the MCR.

33. The method, as claimed in claim 30, wherein said MCRs, at start-up performing steps of:

establishing a point-to-point connection to said well-known end-node; and

informing said well-known end-node of their ATM addresses and of an inter-cluster multicast group from which they want to receive multicast traffic;

on receipt of information concerning the multicast group from which an the MCR wants to receive multicast traffic, said well known end-node returning to said MCR, the ATM addresses which support said group; and

said MCR, on receipt of said ATM addresses, establishing point-to-multipoint connections with each of the multicast group members.

34. The method, as claimed in claim 33, further comprising a MCR, on receipt of a multicast packet from its cluster, using said point-to-multipoint tree to forward the multicast packet to other clusters.

35. The method, as claimed in claim 33, further comprising:

said IP/ATM network using a point-to-multipoint tree of a group member MCR to distribute a multicast packet originating from a non-group member to group members; and

said well known end-node establishing a point-to-point connection between said non-group member and said group member MCR to facilitate transfer of said multicast packet to said group member MCR and thereafter to said group members via said point-to-multipoint tree.

36. The method, as claimed in claim 33, further comprising:

a leaving MCR, wishing to leave an inter-cluster multicast group, contacting said well-known end-node;

said well known end-node, upon being contacted by said leaving MCR, propagating a leave message, in respect of said leaving MCR, to the other MCRs on its group membership list; and

each of said other MCRs, on receipt of said leave message, dropping a leaf for said leaving MCR from their respective point-to-multipoint trees.

37. The method, as claimed in claim 33, further comprising:

a new MCR, wishing to join an inter-cluster multicast group, contacting said well-known end-node;

said well known end-node, request upon being contacted by said new MCR;

returning ATM addresses of existing group members to said new MCR; and

propagating an ATM address of said new MCR to existing group members, through a point-to-multipoint tree;

said new MCR, on receipt of the ATM addresses, establishing a point-to-multipoint connection to each of said existing group members; and

each of said existing group members, on receipt of the ATM address of the new MCR, updating their respective point-to-multipoint trees to include a leaf for said new MCR.

38. The method, as claimed in claim 29, further comprising:

said IP/ATM network operating in server mode ILMP, in conjunction with MARS;

MCRs, on obtaining information concerning existing groups, propagating said information to a corresponding well known end-node residing within an inter-cluster backbone, any one of said MCRs acting as said well known end-node for a multicast group, each group having only one corresponding well known end-node;

a MCR sending group membership reports to the corresponding well-known end-node via a transient control channel established between said MCR and the corresponding well known end-node, said control channel being released after a predetermined period of time;

the MCR performing steps of;

establishing point-to-point data channels for tunneling multicast packets in unicast fashion; and

establishing a point-to-multipoint tree for each multicast group using itself as the root of the tree;

each existing group member reporting their own ATM address to said corresponding well-known end-node when the MCR propagates information concerning an existing group on their cluster side; and

using group member IDs to eliminate multicast packet reflections, said IDs being assigned by said corresponding well known end-node.

39. The method, as claimed in claim 38, further comprising MCRs selecting the well known end-node for a particular inter-cluster multicast group by invoking algorithmic mapping of multicast addresses and corresponding well known end-nodes.

40. The method, as claimed in claim 38, further comprising:

an ATM host, wishing to transmit a multicast packet to members of a multicast group, sending said packet to a Multicast Server (MCS) entity;

on receipt of said multicast packet, said MCS;

copying the multicast packet to all its interfaces where there are group members within the cluster, including the MCR (since MCRs behave promiscuously); and

unicasting said multicast packet to the corresponding well known end-node via a preestablished data channel; and

if said multicast group is reported to said corresponding well-known end-node, said corresponding well known end-node;

extracting the multicast packet from the unicast packet; and

sending the extracted multicast packet to group members using its point-multipoint tree to said group members.

41. The method, as claimed in claim 40, further comprising a MCS wishing to participate in the multicast group, configuring an ATM address of a primary well known end-node and an ATM address of a secondary well known end-node.

42. The method, as claimed in claim 41, further comprising a MCR maintaining:

a single bi-directional-control connection to the well-known end-node for registration/de-registration purposes;

a single data unidirectional data channel for tunneling multicast packets; and

a number (N) of incoming multicast VCs, where N is a total number of multicast groups.

43. The method, as claimed in claim 40, further comprising:

in the event of failure of the primary well known end-node, dismantling all point-to-point connections between the MCS and said primary well-known end-node;

each MCR, on detecting dismantlement of said connections, opening a new circuit to the secondary, well known end-node;

each MCR registering ATM addresses of group members of which it is aware to the secondary well known end-node;

said secondary well known end-node, on receipt of the ATM addresses, establishing point-multipoint trees for each group;

said MCRs continuously trying to reestablish a connection with said primary well known end-node; and

on reestablishment of said connection, said MCRs dismantling their connection to said secondary well known end-node and re-establishing their connection to said primary well known end-node.

44. A telecommunication system, including an ATM transmission system, adapted for the simultaneous transmission of IP data packets to a plurality of system users, using ATM as a carrier network, said transmission system includes:

an IP/ATM network including clusters of Logical IP Systems (LISs);

a multicast address resolution server (MARS) for each cluster; and

a multicast router (MCR) for each cluster, in that said clusters are interconnected through said MCRs which form an inter-cluster backbone for said IP/ATM network, in that multicasting of IP data packets within a cluster utilizes a MARS protocol, and in that multicasting between clusters utilizes an inter-cluster multicasting protocol (ILMP),

wherein a MCR is adapted, in advance of multicasting, to select a well known end-node using a selection algorithm that returns a same well known end-node for an inter-cluster multicast group when executed by any of said MCRs and to establish a connection to the well known end-node, said well known end-node being adapted to provide a database for ATM addresses of members of the inter-cluster multicast group, and in that said well known end-node is adapted to establish a point-to-multipoint tree and to distributing multicast traffic sent by a member of said inter-cluster multicast group to other members of said group using said tree.