Title:Counter-Propagating Core Assignment in Multi-Core Fiber Optical Networks to Reduce Inter-Core Crosstalk and Capacity Wastage

Abstract:Inter-core crosstalk is one of the most serious impairments for signal transmission in a multi-core fiber (MCF) optical network. On the other hand, because of wide deployment of data centers (DCs), we are seeing an increasing bidirectional traffic demand asymmetry, which leads to significant capacity wastage in designing and operating an optical transport network. To alleviate these effects, for an MCF optical network, we propose to assign fiber cores in an MCF in an asymmetric and counter-propagating manner. This can not only significantly reduce inter-core crosstalk between counter-propagating fiber cores but also flexibly assign different numbers of fiber cores in the opposite directions of a fiber link, thereby overcoming network capacity wastage due to the bidirectional traffic demand asymmetry. To evaluate the benefits of the proposed strategy, we consider the routing, spectrum, and core assignment (RSCA) problem for the MCF optical network. An integer linear programming (ILP) model and an auxiliary graph (AG) based heuristic algorithm are developed to optimize network spectrum resource utilization. Simulation studies show the effectiveness of the proposed core counter-propagation strategy, which can significantly outperform its counterpart, i.e., the co-propagation scheme, in terms of the total number of MCFs used and average inter-core crosstalk. In addition, the proposed RSCA heuristic algorithm is efficient to perform close to the ILP model, which can minimize the number of MCFs used and crosstalk between neighboring cores.