Enhancing consecutiveness in elastic optical networks

Abstract

The Elastic Optical Network (EON) has brought several studies to improve optical spectrum efficiency. In the EONs, the bandwidth is divided into frequency slot units (FSUs). When a connection is requested for a service type, the network management system determines how many FSUs are required. As the number of service requests increases, the optical network goes into optical spectrum fragmentation, leaving non-contiguous FSUs over time in the optical spectrum, thus hampering new assignments and generating blocking probability (BP), which is the most significant performance metric for optical networks considering the network layer level, resulting from the unavailability of resources at the time of the request. In this article, the concept of the consecutiveness factor is presented and used to propose an improved heuristic technique to address the problem. The proposed algorithm uses a candidate spectral block with the same amount of FSUs as the requested service demand. This way, the BP is decreased, the resources’ utilization of the network is improved, the number of hops for the connections in the EON is reduced, and the consecutiveness of the services assigned in the network is increased. Some simulation scenarios on the two well-known benchmarks are provided to show the practical efficiency of our proposed technique compared to other available algorithms in the literature.