Efficient Switch Architectures for Pre-configured Backup Protection with Sharing in Elastic Optical Networks (EON)

TL;DR

This paper proposes new switch architectures that enable efficient, shared backup protection in elastic optical networks, addressing limitations of existing designs and improving resource utilization and survivability.

Contribution

The paper introduces novel switch architectures supporting PBPS in EONs, overcoming issues of resource inefficiency and sharing limitations present in existing designs.

Findings

Proposed switch architectures support PBPS in EONs effectively.

New architectures reduce resource wastage and blocking.

Support for shared backup paths enhances survivability.

Abstract

In this paper, we address the problem of providing survivability in elastic optical networks (EONs). EONs use fine granular frequency slots or flexible grids, when compared to the conventional fixed grid networks and therefore utilize the frequency spectrum efficiently. For providing survivability in EONs, we consider a recently proposed survivability method for conventional fixed grid networks, known as pre-configured backup protection with sharing (PBPS), because of its benefits over the traditional survivability approaches such as dedicated and shared protection. In PBPS, backup paths can be pre-configured and at the same time they can share resources. Therefore, both short recovery time and efficient resource usage can be achieved. We find that the existing switch architectures do not support both PBPS and EONs. Specifically, we identify and illustrate that, if a switch architecture is not carefully designed, several key problems/issues might arise in certain scenarios. Such problems include unnecessary resource consumption, inability of using existing free resources, and incapability of sharing backup paths. These problems appear when PBPS is adopted in EONs and they do not arise in fixed grid networks. In this paper, we propose new switch architectures which support both PBPS and EONs. Particularly, we illustrate that, our switch architectures avoid the specific problems/issues mentioned above. Therefore, our switch architectures support using resources more efficiently and reducing blocking of requests.