Trunk reservation for elastic optical networks

Abstract

From the logical performance point of view elastic optical networks are multi-service circuit switched networks where bandwidth allocation has to be managed to guarantee fair performance to channels of different capacity. In this manuscript are discussed and investigated the implementation options for a trunk reservation policy aiming at solving unfairness in the call blocking probability and network resources utilization.

Fairness enhanced dynamic routing and spectrum allocation in elastic optical networks

“…The fragmentation problem is an important issue in EONs. Fragmentation will arise as a consequence of services ending and leaving ‘voids’ of unused capacity, which will lead to unfairness [14]. To avoid fragments, we should minimise the ‘voids’.…”

“…As the multi‐granularity services are assigned with spectrum areas of the same size ignoring their differences in capacity sizes, it may increase unfairness. Fairness is addressed in [12–14] by introducing trunk reservation (TR) policy. TR is a common threshold of acceptance for all services to equalise performance [12].…”

Hybrid Optimized Approaches for Routing and Resource Reservation Protocols in Optical Networks

“…For WA/RA sub-problem, channels are indexed and different protocols are used in the literature including first-fit, most-used, random-fit, last-fit, least-used for dynamic traffic. Different algorithms have been used in the literature for wavelength reservation in optical networks including a trunk reservation policy which has been investigated in [23] for fair channel utilization. A resource reservation algorithm using tabu search has been proposed in [24] for scheduled traffic.…”