Dynamic Path Routing with Maximum Allowable p-Cycle Length and with m-Cycle in Optical Mesh Networks

Abstract

During modern age, fast communication is an essential requirement of everyone and communication between large distances is mainly provided by optical mesh networks. Network failures even for a small interval lead to large disruptions in providing service of transferring of data. Therefore, for reliable network services, network survivability plays a vital role. Preconfigured protection cycle (p-cycle) combines the benefits of ring as well as mesh-based shared protection schemes which are two fundamental techniques of network survivability. Hamiltonian p-cycles (crosses every node of the network at one time) are mostly selected to achieve the level of optimal redundancy due to which the total length of the protection path may become longer than the length of the primary path, so the signal quality may fall below the desired threshold. In order to maintain required signal quality, the maximum possible length of protection path through p-cycle is to be limited. Shortest path routing (SPR) with restricted p-cycle length is mostly employed in current existing strategies of the optical networks. The monitoring cycles (m-cycles) are used for detection of fault and after that the affected service is protected which was interrupted due to span failure. This paper considers dynamic path routing (DPR) using the restricted p-cycle length with m-cycle against single-link failure. p-Cycles combined with m-cycles are used for identification of the location of the faults and its protection. An integer linear program (ILP) is expressed in optimizing the needed protection capacity with length restriction of p-cycles using DPR, and at the same time m-cycle properties are combined. Numerical results show that the total spare capacity with DPR under the maximum allowed p-cycle length combining m-cycles considering fault localization and protection is lesser than SPR.