A loss-free connection control protocol for the Thunder and Lightning network

The ready-to-go virtual circuit protocol (or RGVC) is a novel connection control protocol for gigabit networks that is designed to ensure lossless transmission for delay-sensitive traffic and for traffic whose rate changes with time. The RGVC protocol is one of the two connection control protocols that will be used in the 40 Gbit/s fiber-optic ATM-based Thunder and Lightning network, currently being developed at UCSB. We introduce the RGVC protocol, discuss its main features, and indicate its lossless character. The RGVC protocol can be viewed as a reservation protocol where the reservation and the data transmission phases overlap. The source need not wait for an end-to-end round-trip delay for reservations to be made before transmitting the data. Instead the data packets follow the setup packet after a short offset-interval, which is much smaller than the round-trip delay. As a result, the protocol does considerably better than wait-for reservation protocols in terms of minimizing pre-transmission delay, and is useful for connection establishment for traffic with strict delay requirements. If the setup packet is unsuccessful in reserving the required capacity, or if the rate of the session changes without there being sufficient capacity to accommodate the change, the packets are buffered at intermediate nodes and back-pressure is exercised to the upstream nodes to control the source transmission rate. The back-pressure mechanism uses the concept of freezing of capacity to ensure that the protocol is lossless.

[1]  Ian M. Leslie,et al.  A testbed for wide area ATM research , 1989, SIGCOMM 1989.

[2]  Steven E. Butner,et al.  Architecture and design of a 40 gigabit per second ATM switch , 1995, Proceedings of ICCD '95 International Conference on Computer Design. VLSI in Computers and Processors.

[3]  Craig Partridge.  Protocols for High-Speed Networks: Some Questions and a Few Answers , 1993, Comput. Networks ISDN Syst..

[4]  Israel Cidon,et al.  Paris: An approach to integrated high‐speed private networks , 1988 .

[5]  Ian M. Leslie,et al.  A Testbed for Wide Area ATM Research , 1989, SIGCOMM.

[6]  W. David Sincoskie,et al.  The AURORA Gigabit Testbed , 1993, Comput. Networks ISDN Syst..

[7]  Pierre E. Boyer,et al.  A Reservation Principle with Applications to the ATM Traffic Control , 1992, Comput. Networks ISDN Syst..

[8]  Emmanouel Varvarigos,et al.  An efficient reservation connection control protocol for gigabit networks , 1995, Proceedings of 1995 IEEE International Symposium on Information Theory.

[9]  Björn Pehrson.  MultiG-Distributed Multimedia Applications on Gigabit Networks with Wireless Extensions , 1994 .

[10]  Adrian Segall,et al.  Route Setup with Local Indentifiers , 1986, IEEE Trans. Commun..

[11]  D. P. Tranchier,et al.  Fast bandwidth allocation in ATM networks , 1992 .

[12]  Joseph Y. Hui.  Resource allocation for broadband networks , 1988, IEEE J. Sel. Areas Commun..

[13]  I. Cidon,et al.  Bandwidth management and congestion control in plaNET , 1991, IEEE Communications Magazine.

[14]  Roch Guérin,et al.  The plaNET/ORBIT High Speed Network , 1993, J. High Speed Networks.

[15]  Roch Guérin,et al.  Bandwidth Management and Congestion Control Framework of the Broadband Betwork Architecture , 1993, Comput. Networks ISDN Syst..

[16]  Michel Servel,et al.  The 'Prelude' ATD experiment: assessments and future prospects , 1988, IEEE J. Sel. Areas Commun..

[17]  Andy Hopper,et al.  The Cambridge Backbone Ring , 1990, Proceedings. IEEE INFOCOM '90: Ninth Annual Joint Conference of the IEEE Computer and Communications Societies@m_The Multiple Facets of Integration.

[18]  Israel Cidon,et al.  Connection establishment in high-speed networks , 1993, TNET.