Distributed partial information management (DPIM) schemes for survivable networks .1

This paper describes a novel framework, called distributed partial information management (or DPIM). It addresses several major challenges in achieving efficient shared path protection under distributed control with only partial information, including (1) how much partial information about existing active and backup paths (or APs and BPs respectively) is maintained and exchanged; (2) how to obtain a good estimate of the bandwidth needed by a candidate BP, called BBW, and subsequently select a pair of AP and BP for a connection establishment request so as to minimize total bandwidth consumption and/or maximize revenues; (3) how to distributively allocate minimal BBW (and deallocate maximal BBW) via distributed signaling; and (4) how to update and subsequently exchange the partial information. A DPIM-based scheme using integer linear programming is described to illustrate our approach. In addition, an ultrafast and efficient heuristic scheme is described. With about the same amount of partial information, such a heuristic-based DPIM scheme can achieve almost as a good performance as the ILP-based DPIM scheme, and a much better performance than another ILP-based scheme described by Kodialam and Lakshman (see INFOCOM'00, p.902-911, 2000). The paper also presents an elegant method to support dynamic requests for protected, unprotected, and pre-emptable connections in the unified DPIM framework.

[1]  Chunming Qiao,et al.  Distributed Partial Information Management (DPIM) schemes for survivable networks - Part I , 2002 .

[2]  Yufei Wang,et al.  Optical network design and restoration , 1999, Bell Labs Technical Journal.

[3]  W. Gander,et al.  Adaptive Quadrature—Revisited , 2000 .

[4]  Parameswaran Ramanathan,et al.  Resource aggregation for fault tolerance in integrated services networks , 1998, CCRV.

[5]  Xun Su,et al.  An online distributed protection algorithm in WDM networks , 2001, ICC 2001. IEEE International Conference on Communications. Conference Record (Cat. No.01CH37240).

[6]  Surajit Chaudhuri,et al.  Capacity performance of dynamic provisioning in optical networks , 2001 .

[7]  S. Chaudhuri,et al.  Comparison of centralized and distributed provisioning of lightpaths in optical networks , 2001, OFC 2001. Optical Fiber Communication Conference and Exhibit. Technical Digest Postconference Edition (IEEE Cat. 01CH37171).

[8]  Lorne Mason,et al.  Restoration strategies and spare capacity requirements in self-healing ATM networks , 1997, Proceedings of INFOCOM '97.

[9]  Robert E. Tarjan,et al.  A quick method for finding shortest pairs of disjoint paths , 1984, Networks.

[10]  Murali S. Kodialam,et al.  Dynamic routing of bandwidth guaranteed tunnels with restoration , 2000, Proceedings IEEE INFOCOM 2000. Conference on Computer Communications. Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies (Cat. No.00CH37064).

[11]  Lorne Mason,et al.  Restoration strategies and spare capacity requirements in self-healing ATM networks , 1999, TNET.