A virtual circuit deflection protocol

We propose a communication protocol, called the virtual circuit deflection (VCD) protocol, which combines some of the individual characteristics of virtual circuit switching and deflection routing. An advantage of the VCD protocol over previous (datagram) deflection schemes is that deflections in the former occur on a per session basis (or a per subsession basis, if sessions need to be split to find adequate capacity on the outgoing links), while in the latter, they occur on a per packet basis. This makes packet resequencing at the destination considerably easier to accomplish in the VCD protocol than in datagram deflection schemes. The VCD protocol exploits the storage arising from the high bandwidth-delay product of optical fibers to provide lossless communication with little buffering at the switches and without the need for advance reservations. This makes it particularly suitable for networks that use optical switching, where buffers are expensive to implement with current optical technology. We present a simple implementation of the VCD protocol for such networks, which requires only limited buffering, accomplished through the use of a minimal number of optical delay lines. We also analyze the performance of the protocol for the Manhattan Street network topology by using new analytical models. In particular, we examine the effect of the traffic load and the network size on the throughput and the length of the paths followed by the sessions, and compare the analytical results obtained with corresponding simulation results. The results indicate that the VCD protocol is efficient under both light and heavy traffic conditions, especially when the link capacities are large compared to the basic rate of individual sessions, as is expected to be the case in future multigigabit networks.

[1]  Albert G. Greenberg,et al.  Sharp approximate models of deflection routing in mesh networks , 1993, IEEE Trans. Commun..

[2]  Flaminio Borgonovo,et al.  Deflection Networks: Architectures for Metropolitan and Wide Area Networks , 1992, Comput. Networks ISDN Syst..

[3]  Andrea Fumagalli,et al.  An Optical Switch Architecture for Manhattan Networks , 1993, IEEE J. Sel. Areas Commun..

[4]  Flaminio Borgonovo,et al.  Unslotted deflection routing in all-optical networks , 1993, Proceedings of GLOBECOM '93. IEEE Global Telecommunications Conference.

[5]  Bruce E. Hajek,et al.  Performance of shuffle-like switching networks with deflection , 1990, Proceedings. IEEE INFOCOM '90: Ninth Annual Joint Conference of the IEEE Computer and Communications Societies@m_The Multiple Facets of Integration.

[6]  A. E. Eckberg,et al.  AN APPROACH TO CONTROLLING CONGESTION IN ATM NETWORKS , 1990 .

[7]  Emmanouel Varvarigos,et al.  The "packing" and the "scheduling" packet switch architectures for almost-all optical lossless networks , 1997, GLOBECOM 97. IEEE Global Telecommunications Conference. Conference Record.

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

[9]  Emmanouel A. Varvarigos,et al.  The ready-to-go virtual circuit protocol: a loss-free protocol for multigigabit networks using FIFO buffers , 1997, TNET.

[10]  Arvind Krishna Communication with few buffers: analysis and design , 1991 .

[11]  Flaminio Borgonovo,et al.  On the design of optical deflection-routing networks , 1994, Proceedings of INFOCOM '94 Conference on Computer Communications.

[12]  S. Jamaloddin Golestani,et al.  Congestion-free communication in high-speed packet networks , 1991, IEEE Trans. Commun..

[13]  Flaminio Borgonovo,et al.  A Versatile Model for Predicting the Performance of Deflection-Routing Networks , 1992, Perform. Evaluation.

[14]  Alberto Bononi,et al.  Analysis and comparison of hot-potato and single-buffer deflection routing in very high bit rate optical mesh networks , 1995, IEEE Trans. Commun..

[15]  Zygmunt J. Haas,et al.  The 'staggering switch': an electronically controlled optical packet switch , 1993 .

[16]  Zygmunt J. Haas,et al.  Blazenet: a packet-switched wide-area network with photonic data path , 1990, IEEE Trans. Commun..

[17]  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.

[18]  Emmanouel A. Varvarigos,et al.  Performance of hypercube routing schemes with or without buffering , 1994, TNET.

[19]  Jack Brassil Deflection routing in certain regular networks , 1992 .

[20]  Albert G. Greenberg,et al.  Deflection routing in hypercube networks , 1992, IEEE Trans. Commun..

[21]  Dimitri P. Bertsekas,et al.  Data Networks , 1986 .

[22]  Rene L. Cruz,et al.  COD: alternative architectures for high speed packet switching , 1996, TNET.

[23]  Emmanouel Varvarigos,et al.  An Efficient Reservation Connection Control Protocol for Gigabit Networks , 1998, Comput. Networks.

[24]  Nicholas F. Maxemchuk,et al.  Problems Arising from Deflection Routing: Live-lock, Lockout, Congestion and Message Reassembly , 1991 .

[25]  Nicholas F. Maxemchuk,et al.  Comparison of deflection and store-and-forward techniques in the Manhattan Street and Shuffle-Exchange Networks , 1989, IEEE INFOCOM '89, Proceedings of the Eighth Annual Joint Conference of the IEEE Computer and Communications Societies.

[26]  A G Greenberg,et al.  Sharp approximation models of adaptive routing in mesh networks (preliminary report) , 1986 .

[27]  P. O'Reilly The case for circuit switching in future wide bandwidth networks , 1988, IEEE International Conference on Communications, - Spanning the Universe..

[28]  Baruch Awerbuch,et al.  Distributed control for PARIS , 1990, PODC '90.