Multi-protocol over ATM (MPOA) is being considered by the industry as an important short-cut technology that provides an efficient transfer of inter-subnet unicast data in a LANE environment. MPOA was initially considered to carry backbone traffic in the enterprise or campus networks. However, congestion in the public Internet provokes many to consider MPOA as a solution for the carrier or service provider networks as well. In this paper, we investigate the scalability issues of MPOA in the wide area network environment. We use a realistic simulation model driven by real Internet traffic to study crucial metrics such as the SVC setup rate, the number of VCs required, and the percentage of packets switched. We find that MPC ingress cache size provides a three-way trade-off among the percentage of switched packets, the VC usage and the SVC setup rate requirement. We also find that the SVC setup rate is linearly dependent on the packet arrival rate.

NYNEX Science and Technology and Fujitsu Network Switching Systems of North America undertook development of an interface between a NYNEX Network Management System (NetWindow) and Fujitsu's Fetex-150 Broadband Switching System (B-ISDN) for a specific service using a (Bellcore) standard information model based on the OSI Common Management Information Protocol (CMIP). CMIP is a powerful but complex protocol. The modeling flexibility and enhanced vocabulary of the protocol increase the level of effort required to implement both manager and agent code. The areas of CMIP that cause the hardest implementation problems revolve around unbounded recursions in either the model or protocol. Careful selection of limitations on the information model structure, and CMIP features allows sufficient power at controllable implementation expense. By devising and applying a series of simplifying principles during the development of a comprehensive set of implementation agreements, the authors were able to contain the implementation complexity while retaining the power of the protocol. Examples of the simplifying steps are given to illustrate these principles.>

The Integrated Services Digital Network (ISDN) provides basic architecture for existing, as well as future residential plus business communications. ISDN overlayed with CCS#7 of a digital PSTN (Public Switched Telephone Network) can be the ultimate, ubiquitous network for circuit switch (voice, data), packet switch (voice, data), and private line (voice, data) applications. Assuming that the present ISDN has to interwork in the present physically separate overlayed networks (voice and data), significant problems are expected to emerge for designing hardware and linking softwares for handling packet traffic. In this paper, the software-related problems, when ISDN packet distribution nodes have to handle an ISDN interface, will be outlined with an ISDN software protocol solution. An approximation of the delay involved in the telephone switching system which is part of ISDN processing as well as the delay for the interface gateways, the HOST computer nodes, and the LAN and WAN computer nodes will be identified and formulated to reflect the total performance measure defined. Major emphasis is given to flow and congestion control performance measures in the ISDN Gateways, which are analyzed and simulated with the assistance of the basic delay table transfer software model developed for the IMPS and gateways in the ARPANET, MILNET, and MINET. The performance evaluation of this basic ISDN interfacing software, which only involved one ISDN level, i.e., the HOST or gateway and its related subnetworks, is simulated on sections of these networks to illustrate its congestion control effectiveness. There are six mathematical software techniques to account for end-to-end delay, which form the basis for the solution to these ISDN software-hardware problems in the Interface Gateways connecting the electronic switch to the computer network components.

This paper considers the problem of congestion in message-switched, or packet-switched, networks, which employ probabilistic routing schemes. A queueing model for the analysis of a local congestion control scheme is presented. Analytic study of the model shows that local congestion control with random routing is superior to similar control with deterministic routing for a particular example worked out here.

This concise paper extends the results of Pennotti and Schwartz by proposing a queueing model of a communication network with random routing and end-to-end congestion control. Using conventional performance criteria, the performance of this model is compared with that of a previously published model, where fixed routing was employed.