Abstract: Network virtualization is a key enabling technology for “Infrastructure as a Service” provisioning, increasing the flexibility and cost savings offered to customers. By extending the concept of server virtualization to the network infrastructure, the allocation of different, independent virtual networks over a single physical network is carried out on demand. A fundamental challenge in network virtualization systems is to choose which physical nodes and links to use for hosting virtual networks in the physical infrastructure, known as the “virtual network allocation” problem. All virtual network allocation proposals on elastic optical networks assume a centralized operation, deploying a single node with access to the network state global information and assigning resources accordingly. However, such configuration might exhibit the inherent problems of centralized systems: survivability and scalability. In this paper, we present a distributed protocol for resource discovery, mapping, and allocation of network virtualization systems. The distributed protocol is generic enough as to be used with different substrate networks. However, in this paper, it has been adapted to work over an elastic optical network infrastructure, where further considerations regarding the spectrum continuity and contiguity constraints must also be taken into account. The distributed protocol is based on the concept of alliances: upon the arrival of a virtual network request, agents located in the physical network nodes compete to form the first alliance able to host the virtual network. Because the first alliances to be formed are also the ones composed by nearby nodes, a good network resource usage is achieved. The feasibility of the distributed protocol was studied by evaluating its ability to successfully establish virtual networks within acceptable time and with low bandwidth consumption from the coordination messages.

Abstract: Multi-band elastic optical networks are a promising alternative to meet the bandwidth demand of the ever-growing Internet traffic. In this letter, we propose a family of band allocation algorithms for multi-band elastic optical networks. Employing simulation, we evaluate the blocking performance of 3 algorithms of such a family and compare their performance with the only heuristic proposed to date. Results show that the three new algorithms outperform the previous proposal, with up to one order of magnitude improvement. We expect these results to help advance the area of dynamic resource allocation in multi-band elastic optical networks.

Abstract: This paper studies the physical layer's impact on the blocking probability and energy consumption of wide-area dynamic elastic optical networks (EONs). For this purpose, we consider five network configurations, each named with a network configuration identifier (NCI) from 1 to 5, for which the Routing, Modulation Level, and Spectrum Assignment (RMLSA) problem is solved. NCI 1-4 are transparent configurations based on all-EDFA, hybrid Raman/EDFA amplifiers (with different Raman gain ratio Gamma(R)), all-DFRA, and alternating span configuration (EDFA and DFRA). NCI 5 is a translucent configuration based on all-EDFA and 3R regenerators. We model the physical layer for every network configuration to determine the maximum achievable reach of optical signals. Employing simulation, we calculate the blocking probability and the energy consumption of the different network configurations. In terms of blocking, our results show that NCI 2 and 3 offer the lowest blocking probability, with at least 1 and 3 orders of magnitude of difference with respect to NCI 1 and 5 at high and low traffic loads, respectively. In terms of energy consumption, the best performing alternatives are the ones with the worst blocking (NCI 1), while NCI 3 exhibits the highest energy consumption with NCI Gamma(R) = 0.75 following closely. This situation highlights a clear trade-off between blocking performance and energy cost that must be considered when designing a dynamic EON. Thus, we identify NCI 2 using Gamma(R) = 0.25 as a promising alternative to reduce the blocking probability significantly in wide-area dynamic EONs without a prohibitive increase in energy consumption.

Abstract: A deep reinforcement learning (DRL) approach is applied, for the first time, to solve the routing, modulation, spectrum, and core allocation (RMSCA) problem in dynamic multicore fiber elastic optical networks (MCF-EONs). To do so, a new environment was designed and implemented to emulate the operation of MCF-EONs – taking into account the modulation format-dependent reach and intercore crosstalk (XT) – and four DRL agents were trained to solve the RMSCA problem. The blocking performance of the trained agents was compared through simulation to 3 baselines RMSCA heuristics. Results obtained for the NSFNet and COST239 network topologies under different traffic loads show that the best-performing agent achieves, on average, up to a four-times decrease in blocking probability with respect to the best-performing baseline heuristic method.