Sun and Oracle Community Voices How to Buy Log In United States [Change] English

Navbar Here$navbar =
$url_id =
$ROOT_ID = 1
$url = /techrep/1995/annualreport95/hispeed.html

High-Speed Networks

Fiscal 1995 Project Portfolio Report

Israel Cidon
israel.cidon@eng.sun.com

Overall Objective

To explore new opportunities and become a knowledge seed in the emerging technologies of high-speed networking, especially asynchronous transfer mode (ATM), and to investigate the value-added synergy between computing and communication platforms.

Objective for FY95

To define high-speed networking projects which combine leading edge research activities and tangible business opportunities for Sun by:

  • Recruiting a small group of networking experts;

  • Developing relationships with our internal Sun customers in SMCC, SunSoft and STB;

  • Initiating external collaboration with leading ATM vendors and researchers.

Description

The OPENET project seeks to develop a novel and efficient architecture for a high-performance vendor-independent ATM network control platform. It addresses one of the factors hampering widespread deployment of ATM equipment for computer networks, namely the lack of interoperability among switches from several different vendors. Although standards for network control are being developed, they are not being rapidly embraced and implemented by equipment manufacturers. The OPENET platform will manage ATM network connections through a heterogeneous network, choosing optimal routes subject to capacity and quality-of-service (QOS) requirements, and configuring the switches to route traffic accordingly. The existence of such a platform is crucial for the success of the ATM market, and will enable the development of many universal network service applications.

The high-speed networking project is also exploring new ways to provision networks of workstations with suitably scalable ATM switches. This effort explores the technologies required to include network switching elements (Ethernet and ATM) as add-on cards to Sun systems. We have examined a broad range of possible solutions and have identified a novel and promising technology (an active bus or point-to-point ring) to be superior to the other alternatives in terms of simplicity, cost, integration, system scalability, future growth, and functionality. The recent new mandatory traffic management requirements set by the ATM Forum are especially easy to accommodate in the new architecture.

Accomplishments

The basic OPENET architecture design has been completed, ranging from the conceptual stage to a detailed design of message formats and finite state machines. The basic architecture provides only an essential set of processes, including topology and utilization database maintenance and update (the first follows the ATM Forum PNNI structure); leader election and setup of a distribution tree as required by the utilization multicast; routing procedures for unicast-cast and multicast, including the handling of route caching and QOS constraints; and call setup, modifications, maintenance, and takedown procedures for both unicast, multicast, and third party (add, drop) cases. These call control procedures include the ability to negotiate the amount of bandwidth to be reserved, to deliver information about congestion in cases of failed attempts, and to handle link and switch outage. In particular, they solve the problem of reservation garbage collection.

Currently, two leading ATM vendors (GDC and Newbridge) and a leading network research group (Jon Turner, Washington University, Saint Louis) have agreed to collaborate with us.

References

Publications

Cidon, I., L. Georgiadis, R. Guerin, and A. Khamisy. "Optimal Buffer Sharing." IEEE Journal on Selected Areas in Communications 13, no. 7., Also Technion EE PUB No. 932 (August 1994). Also INFOCOM'95 (April 1995).

Cidon, I., O. Gerstel, and S. Zaks. "A Scalable Approach to Routing in ATM Networks." 8th International Workshop on Distributed Algorithms (September 1994).

Cidon, I., I. S. Gopal, and S. Kutten. "New Models and Algorithms for Future Networks." IEEE Transactions on Information Theory 41, no. 3 (May 1995).

Cidon, I., R. Guerin, I. Kessler, and A. Khamisy. "Analysis of a Statistical Multiplexer with Generalized Periodic Sources." Queueing Systems: Theory and Applications (Forthcoming).

Cidon, I., R. Guerin, and A. Khamisy. "An Investigation of Application Level Performance in ATM Networks." INFOCOM '95 (April 1995).

Cidon, I., R. Guerin, A. Khamisy, and M. Sidi. "On Queues with Inter-Arrival Times Proportional to Service Times." Queuing Systems (Forthcming).

Cidon, I. and Y. Shavitt. "Message Terminating Algorithms for Anonymous Rings of Unknown Size." Information Processing Letters, no. 54 (1995): 111-119.

Georgiadis, L., R. Guerin, and A. Parekh. "Optimal Multiplexing on a Single Link: Delay and Buffer Requirements." IEEE Transactions on Information Theory.

Jiang, S. and M. T. Hsiao. "Performance Evaluation of a Receiver-Based Handshake Protocol for CDMA Networks." IEEE Transactions on Communications 43, no. 5 (May 1995).

Parekh, A. "Selecting Routers in Ad-hoc Wireless Networks." SBT/IEEE International Telecommunications Symposium (August 1994). Also appeared as an invited paper in a special issue of the Brazilian Journal of Computer-Science.

Tseng, J. K. H. and M. T. Hsiao. "Optimal Control of Arrivals to Token Ring Networks with Exhaustive Service Discipline." Operations Research 43, no. 1 (January 1995): 89-101.

View previous page

Return to the Table of Contents

View next page

webmaster@sunlabs.eng.sun.com