在高頻寬的波長分波多工網路上，網路元件的錯誤會造成網路服務很大的損失。因此，對於網路元件錯誤的快速保護和回復便成為未來網路技術發展上相當重要的議題。長久以來，網路回復機制只能在環狀網路和網狀網路之間做選擇，直到有人提出事先配置保護環(以下簡稱保護環)的概念，打破了這樣的窘境。保護環的優點在於其不但具備環狀網路的回復速度，同時也可以和網狀網路一樣只需要很少的備用頻寬來提供保護，因此逐漸引起大家的注意。
目前以保護環為基礎的網路容錯設計已經有相當的發展，之前的研究者在這方面提出了許多的方法。這些方法主要可以分成兩種配置保護環的方式，一種是採用靜態的保護環，另一種是採用可重配置的保護環。採用靜態保護環的網路在設計上的主要目標是達到100%的單一錯誤回復率，這也是所有保護環網路設計上的起始點。而採用可重配置保護環的網路在設計上的主要目標是達到100%的雙重錯誤回復率，依改變保護環配置的方式可分成以下三種方法，完全性的保護環重配置、增加性的保護環重配置和動態修復保護環。動態修復保護環的優點是備用頻寬需求量和平均重配置成本皆介於完全性保護環重配置的方法和增加性保護環重配置的方法之間，缺點是使用了兩種保護環在其網路設計中，包括一般保護鏈結的保護環和保護路徑片段的保護環，這將使得網路系統在錯誤處理上更加複雜，增加了系統的負擔。
在本篇論文中，我們提出了一個以保護環為基礎的新保護策略。我們將新保護策略的形成分成兩個解法，第一個解法改進了動態修復保護環的方法中使用兩種保護環的方式，只使用一般保護鏈結的保護環來提供保護。只使用一種保護環來提供保護，雖然減少了系統在錯誤處理上的負擔，但是在備用頻寬需求量方面卻有所增加。因此，第二個解法將第一個解法結合強迫者概念中強迫者填滿的應用，讓取代保護環回復錯誤鏈結上工作頻寬的工作繞路能盡量通過非強迫者鏈結，如此便能利用原本已配置的保護環來保護工作繞路，減少需要額外配置的保護環，進而達到減少備用頻寬需求量的目的。在效能模擬與分析中，我們針對提出的新保護策略和其他可重配置保護環網路設計的方法，在四個不同的網路拓樸下，做備用頻寬需求量和平均重配置成本效能的比較。模擬結果顯示，新保護策略的第二個解法在備用頻寬需求量和平均重配置成本的效能表現上和動態修復保護環的方法一樣介於完全性保護環重配置的方法和增加性保護環重配置的方法之間，並且在備用頻寬需求量方面更優於動態修復保護環的方法。

In high-capacity Wavelength-Division-Multiplexing (WDM) networks, a failure of a network component can lead to a severe disruption of network services. Hence, fast protection and restoration against network component failures is critically important in future WDM networks. Ring-based and mesh protection which recover network traffic from link failure are the most familiar fault-tolerant mechanisms in WDM networks. Beside, a fascinating protection mechanism called pre-configured protection cycle (p-cycle) becomes a focused research issue in recent years.
There are many methods which have been introduced in p-cycle network design. These methods can be divided into two categories, i.e., static p-cycle and reconfigurable p-cycle. In static p-cycle network design, 100% single-failure restorability must be achieved. And in reconfigurable p-cycle network design, the objective is to reach 100% dual-failure restorability. There are three principal methods in reconfigurable network design, i.e., complete reconfiguration, incremental reconfiguration, and dynamic repair of p-cycles. The performance of dynamic repair of p-cycles (DRP) in terms of spare capacity requirement and average reconfiguration overhead is between that of complete reconfiguration and incremental reconfiguration methods. But DRP uses two kinds of p-cycles including span-protecting p-cycle and path-segment-protecting p-cycle to protect whole network. This will increase the burden of network system on failure processing. Therefore, our objective is to improve dynamic repair of p-cycles by using only span-protecting p-cycle instead of two kinds of p-cycle.
In this paper, we first propose a new strategy based on using only span-protecting p-cycle to protect alternative working route in dynamic repair of p-cycles. This change will decrease the burden of network system since only one kind of p-cycle is used in the network, but increase the requirement of spare capacity. Consequently, we incorporate the forcer concept into the first solution in order to reduce the requirement of spare capacity. To evaluate the performance in spare capacity requirement and average reconfiguration overhead, four network topologies are used as the test networks in our simulation. And the simulation results show that our new strategy has better performance than dynamic repair of p-cycles in terms of the requirement of spare capacity. So we not only improve dynamic repair of p-cycles by only using one kind of p-cycle to provide protection, but also reduce the spare capacity requirement for dual-failure restoration.

[1] W. D Grover, D. Stamatelakis “Cycle-oriented distributed preconfiguration: Ring-like speed with mesh-like capacity for self-planning network restoration,” Proc. IEEE International Conference on Communications (ICC) 1998, Atlanta, Georgia, USA, pp. 537-543, 7-11 Jun. 1998.
[2] W. D. Grover, D. Stamatelakis, “Bridging the ring-mesh dichotomy with p-cycles,” Proc. Second International Workshop on the Design of Reliable Communication Networks (DRCN 2000), Munich, Germany, pp. 92-104, 9-12 Apr. 2000.
[3] D. A. Schupke, C. Gruber, and A. Autenrieth, “Optimal configuration of p-cycles in WDM networks,” in Proc. of ICC 2002, pp. 2761-2765, 2002.
[4] T. Flanagan, “Fiber network survivability,” IEEE Communications Magazine, Vol. 28, No. 6, pp. 46-53, June 1990.
[5] R. Iraschko, M. MacGregor, W. D. Grover, “Optimal capacity placement for path restoration in STM or ATM mesh survivable networks,” IEEE/ACM Trans. on Networking, pp. 325-336, 3 June, 1998.
[6] M. Clouqueur, W.D. Grover, “Availability analysis of span-restorable mesh networks,” IEEE JSAC Special Issue on Recent Advances in Fundamentals of Network Management, vol.20, no. 4, pp. 810-821, May 2002.
[7] D. A. Schupke, A. Autenrieth, and T. Fischer, “Survivability of multiple fiber duct failures,” in Proc. of DRCN 2001, pp. 213-219, 2001.
[8] D. A. Schupke, “The tradeoff between the number of deployed p-cycles and the survivability to dual fiber duct failures,” in Proc. of ICC 2003, pp. 1428-1432, 2003.
[9] D. A. Schupke, “Multiple failure survivability in WDM networks with p-cycles,” in Proc. of ISCAS 2003, vol. 3, pp. III/866-869, 2003.
[10] D. A. Schupke, W. D. Grover, M. Clouqueur, “Strategies for enhanced dual failure restorability with static or reconfigurable p-cycle networks,” Proc. IEEE International Conference on Communications (ICC) 2004, pp. 1628-1633, 20-24 Jun. 2004.
[11] D. S. Mukherjee, C. Assi, A. Agarwal, “Alternate Strategies for Dual Failure Restoration Using p-Cycles,” Proc. IEEE International Conference on Communications (ICC) 2006, vol. 6, pp. 2477-2482, June 2006.
[12] G. Shen, W. D. Grover, “Extending the p-Cycle Concept to Path Segment Protection for Span and Node Failure Recovery,” IEEE JSAC Optical Communications and Networking Series, vol. 21, no.8, pp. 1306-1319, Oct. 2003.
[13] H. Hwang, S. Ahn, Y. Yoo, and C. S. Kim, “Multiple shared backup cycles for survivable optical mesh networks,” in Proc. of ICCCN 2001, pp. 284-289, 2001.
[14] W. D. Grover, D. Y. Li, “The Forcer Concept and Express Route Planning in Mesh-survivable Networks,” Journal of Network and Systems Management, vol. 7, no. 2, pp. 199-223, Jun. 1999.
[15] G. Shen, W. D. Grover, “Exploiting Forcer Structure to Serve Uncertain Demands and Minimize Redundancy of p-Cycle Networks,” OptiComm. 2003, pp. 59-70, October 2003.
[16] http://www.ece.ualberta.ca/~grover/book/protected-material/
[17] http://www.ampl.com