Multiplex communications – Diagnostic testing – Determination of communication parameters
Reexamination Certificate
1999-07-30
2003-09-02
Olms, Douglas (Department: 2732)
Multiplex communications
Diagnostic testing
Determination of communication parameters
C370S254000
Reexamination Certificate
active
06614762
ABSTRACT:
TECHNICAL FIELD
The invention concerns the computation of complex node representations and the calculation of a path in a PNNI network.
BACKGROUND OF THE INVENTION
For asynchronous transfer mode (ATM) switches to communicate, a standards-based signalling and routing protocol called Private Network-to-Network Interface (PNNI) is used. PNNI is a comprehensive routing and signalling protocol for use in an ATM network and is a comprehensive signalling standard. Among the major characteristics are signalling for switched virtual circuits (SVCs) and dynamic routing capabilities. It also supports the Quality of Service (QoS) parameters. PNNI was approved by the ATM Forum in 1996 and is found in many ATM systems.
PNNI supports a dynamic information exchange to allow switches to update routing paths and to form alternate rerouting in case of link failure.
In order to support bandwidth request and QoS, a local PNNI switch has to know the network topology. Knowing whether the network can support end-to-end QoS (for example the required bandwidth) and whether the path is available are the only ways the local switch can accept a call without compromising the call integrity. Such information can be established manually when the network is formed. However, having to inform every switch on the network when a new switch is added or when the topology changes is very labor intensive, not to mention the increasing probability for errors. The only effective process is to have the switches exchange information with one another on a regular basis. PNNI requires such an exchange of information as discussed in the next section.
Topology information is exchanged automatically on a regular basis or upon significant changes to ensure that every switch in the network has the most current view. Switches form peer groups under common ATM prefix. A peer-group-leader (PGL) is elected in each peer group to represent the peer group at a higher layer. The PGL does not have to be the connecting node between two peer groups. An efficient procedure governs the frequency and the amount of information being exchanged so that bandwidth is conserved. If update information is received by a switch, it is compared with the existing topology information and changes will automatically be updated. The effect of the information exchange is to increase the ability to reach the destination. By providing alternate rerouting, if a commonly used path fails, an alternate path, if available, will be used to reach the destination. Only by having updated topology information can switches be relied on to make such distributed intelligent decisions.
To reduce the overall complexity, the amount of needed memory, and the path selection complexity in particular, PNNI uses the hierarchical model for topology aggregation, as indicated above. At various levels of this hierarchy, a PNNI peer group is represented one level up by a single node.
PNNI is a hierarchical, link-state routing protocol that organizes switching systems into logical collections called peer groups. Neighboring call establishment in PNNI consists of two operations: the nodes form a peer group by exchanging their peer group identifiers (PGIDs) via Hello packets using a protocol that makes nodes known to each other. If the nodes have the same PGID, they belong to the peer group defined by that particular PGID; if their PGIDs are different, they belong to different peer groups. A border node has at least one link that crosses the peer group boundary. At upper layer (between logical group nodes) hello protocol exchanges occur over logical links called SVCC-based routing control channels (SVCC-RCCs). SVCC stands for switched virtual connection channel. PNNI defines the creation and distribution of a topology database that describes the elements of the routing domain as seen by a node. This topology database provides all the information required to compute a route from the node to any address that is reachable in, or through that routing domain. Nodes exchange database information using PTSEs (PNNI Topology State Elements). PTSEs contain topology characteristics derived from link or node state parameter information. The state parameter information could be either metrics or attributes. PTSEs are grouped to form PTSPs (PNNI Topology State Packets) which are flooded throughout the peer group so that all nodes in one peer group will have an identical topology database. As mentioned already, every peer group has a node called PGL. There is at most one active PGL per peer group. The PGL will represent the current peer group in the parent peer group as a single node called logical group node (LGN). The LGN will also flood the PTSEs in the parent peer group down to the current peer group. Apart from its specific role in aggregation and distribution of information for maintaining the PNNI hierarchy, the PGL does not have any special role in the peer group.
Call establishment in PNNI consists of two operations: the selection of an optimal path and the setup of the connection state at each point along that path. To provide good accuracy in choosing optimal paths in a PNNI network, the PNNI standard provides a way to represent a peer group with a structure which is more sophisticated than the single node. This representation is called ‘complex node representation’. It allows advertisement of the cost of traversing this node and therefore the cost of traversing the whole peer group summarized by the respective complex node representation.
The computation of complex node representations and the aggregation and distribution of information for maintaining identical databases within a peer group and between peer groups is very complex and time consuming in particular when dealing with large networks. In other words, the path calculation becomes slower with increasing size of a network and topology updates use up more and more of the node's and link's capacity.
It is an object of the present invention to provide a fast and reliable method for the computation of complex node representations.
It is an object of the present invention to provide a fast and reliable method for the calculation of a path in a PNNI network.
It is another object of the present invention to provide a method for selection of an optimal path in a PNNI network.
It is a further object to provide improved PNNI nodes and PNNI networks.
SUMMARY OF THE INVENTION
The present invention concerns a scheme for the computation of a restrictive cost between pairs of border nodes of a PNNI peer group which comprises nodes, some of the nodes being border nodes. The peer group further comprises links connecting pairs of nodes. The following steps are carried out to compute the restrictive cost:
a. maintaining a sorted list of said links sorted according to their restrictive cost C,
b. forming logical sets of nodes
by taking one link after the other from said list starting with the cheapest cost C and assigning an identifier to the two nodes connected by the respective link, whereby the nodes of a set of nodes carry a dominant unique identifier if a border node is a member of the respective set, or a unique identifier if there is no border node being a member of the respective set,
uniting two sets, if such a link connects a node of a first set and a node of a second set, by assigning a dominant unique identifier to all nodes of both sets if either nodes of the first or the nodes of the second set carry a dominant unique identifier, or by assigning a unique identifier to all nodes of both sets if neither the nodes of the first nor the nodes of the second set carry a dominant unique identifier,
c. applying a matrix update algorithm if said link connects a node of a first set and a node of a second set that both already carry dominant unique identifiers.
The present scheme can be employed in any kind of network devices, such a routers for example. The scheme can also be used for computation of an optimal path in a PNNI network, or for routing a packet or frame from a source node to a destination node using an optimal path in a PNNI
Illiadis Ilias
Scotton Paolo
Flynn John D.
Grosser George E.
International Business Machines - Corporation
Olms Douglas
Wilson Robert W.
LandOfFree
PNNI topology abstraction does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with PNNI topology abstraction, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and PNNI topology abstraction will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3049084