Abstract: A method of performing flow control in a PNNI network wherein a node that becomes congested advertises this fact to the network thus eliminating additional call requests and other messages that it cannot handle due to the current load. The method utilizes the Maximum Cell Rate (MCR) and Available Cell Rate (ACR) parameters provided for in the PNNI standard. When the congestion level on a local node exceeds a predetermined threshold, a PTSE with MCR set to zero and ACR set to zero is advertised to the network. The PTSE causes nodes (originating nodes in particular) to not consider that node in route calculations. When the congestion level recedes on the local node, another PTSE is generated with the original MCR and ACR that existed before the congestion occurred causing the resumption of call request messages to be forwarded to the network.

Abstract: An apparatus for and method of enabling the debugging and testing of complex multilevel PNNI based ATM networks. The invention has applications in networks wherein one or more nodes implement only the Minimum Function PNNI implementation and wherein these modes must operate correctly in a PNNI hierarchy environment. A plurality of PTSEs representing simulated virtual portions of an ATM network are injected into a node under test. The PTSEs represent hierarchical portions of ATM networks that are difficult or impossible to implement. The virtual portions of the networks may or may not have been able to be created using real physical network elements. The method includes first generating the injection file containing all the PTSEs to be simulated and then injecting this file into the node under test.

Abstract: The plurality of nodes in a computer network is divided into a plurality of peer groups. The plurality of peer groups can further be divided into a plurality of upper level peer groups. Routing information for each peer group is summarized according to an algorithm. This summary routing information is then stored in nodes which are not a member of that peer group. The algorithm for summarizing the routing information, represents each peer group as a summary node where nodes of the peer group which connect to other peer groups, are considered ports of the summary node. Metric values for the characteristics of paths connecting these ports are determined. A metric average is computed for all of the port-to-port paths and this average is then halved to give a peer default metric value for a routing path entering into, or exiting from that particular peer group.