Model Features: PNNI

This section provides a list of the main features available in the PNNI model.

The PNNI model suite captures the following protocol behavior.

Table 177: PNNI Model Features

Feature Description
Flexibility The model has many configurable parameters. The PNNI model allows you to configure the following:
• Up to 104 hierarchical levels
• External static routes to external PNNI domains
• Routing metric — metrics can be administrative weights, cell transfer delay, or cell delay variation
• Complex node parameters to govern the costs of traversal of logical nodes
Report generation The model provides several ways to easily display a network’s configuration information. The PNNI model allows you to print the following:
• Node information, such as group IDs, peer group leader status, and node ID
• Route information for the PNNI domain
• Hierarchical addressing information
• Topology Database on a node at various times during the simulation
NNI Signaling NNI (Network Node Interface) signaling is used during call setup and call tear-down procedures.
Topology Discovery in Fast-Mode PNNI Topology information is instantaneously propagated to all nodes in the network. No latency is involved in updating changes in the network.
Topology databases are maintained on a per-group basis instead of a per-node basis.
Topology database is not subject to aging.
Load Balancing (Flow Analysis only) Based on algorithms implemented on Nortel devices
• Widest Path
• Random
• Stochastic—Based on available cell rate, routing metric, or both
Topology Discovery in Explicit PNNI Topology discovery is done in the explicit mode, where the Hello protocol first discovers the neighbor links and nodes. Then, database synchronization and flooding is done to propagate the topology information in the network.
All changes in the network take time to propagate to other nodes in the network.
Topology databases are maintained on a per-node basis.
Topology database is subject to aging.
Peer Group Leaders Peer group leaders are selected based on:
1) the peer group leader election algorithm
2) the peer group leader priority configured on each node in a peer group.
  • The PNNI model follows the ATM forum’s PNNI 1.0 specification.
  • Djkstra’s algorithm is used to compute shortest path routes through the network.
  • PNNI models are implemented based on information available from the following sources:

Table 178: Reference Documents

P-NNI V1.0 PNNI 1.0 ATM Forums: Private Network-Network Interface Specification Version 1.0 (PNNI1.0)