Dynamic routing of anycast and unicast traffic in elastic optical networks with various modulation formats — Trade-off between blocking probability and network cost

The continuous use of Internet services and applications triggers a sustainable growth of data transfer, therefore generating a necessity to develop an efficient and scalable transport platform. A novel-networking concept - Elastic Optical Networks (EONs) - can support bandwidth-demanding traffic from an advent of new services including cloud computing and content delivery networks. In this paper, we focus on dynamic routing of anycast and unicast traffic; anycast refers to cloud computing or content traffic served by data centers. We assume that EONs support various distance-adaptive modulation formats, namely BPSK, QPSK, and m-QAM, where m belongs to {8, 16, 32 and 64}. A higher modulation format consumes less spectrum, which can lead to a smaller blocking probability (BP) in dynamic routing. However, at the same time, a higher modulation format provides a smaller transmission distance, which triggers the need to use regenerators and increases CAPEX/OPEX network costs as well as power consumption. Therefore, the main goal of our research is to find a trade-off between the BP and using extra regenerators in the context of dynamic routing of anycast and unicast traffic in EONs.