Bandwidth variable transceivers with artificial neural network-aided provisioning and capacity improvement capabilities in meshed optical networks with cascaded ROADM filtering

Xingyu Zhou,Qunbi Zhuge,Meng Qiu,Meng Xiang,Fangyuan Zhang,Baojian Wu,Kun Qiu,David V. Plant

DOI: https://doi.org/10.1016/j.optcom.2017.09.021

IF: 2.4

2018-01-01

Optics Communications

Abstract:We investigate the capacity improvement achieved by bandwidth variable transceivers (BVT) in meshed optical networks with cascaded ROADM filtering at fixed channel spacing, and then propose an artificial neural network (ANN)-aided provisioning scheme to select optimal symbol rate and modulation format for the BVTs in this scenario. Compared with a fixed symbol rate transceiver with standard QAMs, it is shown by both experiments and simulations that BVTs can increase the average capacity by more than 17%. The ANN-aided BVT provisioning method uses parameters monitored from a coherent receiver and then employs a trained ANN to transform these parameters into the desired configuration. It is verified by simulation that the BVT with the proposed provisioning method can approach the upper limit of the system capacity obtained by brute-force search under various degrees of flexibilities.