Impairment-aware manycast routing, modulation level, and spectrum assignment in elastic optical networks

In this paper, we study the impairment-aware manycast routing, modulation level, and spectrum assignment problem in elastic optical networks. We formulate a mixed-integer linear program (MILP) that serves all the given manycast requests at once, referred to as a joint impairment-aware MILP. The formulated MILP can be used to jointly serve diverse traffic types, i.e., manycast, multicast, anycast, and unicast requests. In this formulation, nonlinear interference noise generated in fibers, the noise of optical amplifiers, the limitation of the maximum splitting degree of multicast-capable nodes (MCNs), and the power penalty of splitting the data flow into multiple branches at MCNs are considered. Furthermore, by modifying the joint MILP, two decomposed MILPs and corresponding heuristic algorithms are proposed to find a light-tree and assign modulation level and spectrum to the given requests, sequentially. We simulated the proposed heuristic algorithms and joint MILP (as a benchmark) and compared them with their distance-adaptive (DA) alternatives by considering a small-scale network. Our results reveal that our heuristic algorithms perform close to the joint MILP and have lower spectrum consumption than the DA alternatives. Furthermore, the performance of the heuristic algorithms is evaluated by considering a large-scale network. We observe that the number of occupied frequency slots and the average SNR of the network depend on the transmission power, and there is an optimum value for transmission power that maximizes the spectral efficiency. Moreover, the results show that the impairment-aware schemes outperform their DA alternatives in terms of spectrum consumption.