Performance Evaluation of WSS-based All-Optical Spine-Leaf Data Center Network

In this paper, we consider service provisioning in a wavelength selective switch (WSS) based all-optical data center network with spine-leaf topology. Its related routing, wavelength, time-slot assignment (RWTA) problem is modelled and solved in the context of two WSS reconfiguration scenarios. Specifically, for the all-stop (AS) WSS reconfiguration scenario, an integer linear programming (ILP) model is formulated and an efficient heuristic algorithm is developed. For the not-all-stop (NAS) WSS reconfiguration scenario, an efficient heuristic algorithm is developed. Simulation results show that the proposed schemes are efficient to significantly reduce the overall task completion time (TCT) compared with a benchmark scheme. Also, the NAS scenario can significantly reduce the overall TCT compared with the AS scenario since the former can maximally eliminate the WSS reconfiguration time.