Throughput Optimization in a Tunable Subcarrier-based Coherent PON

In the Very High Speed Passive Optical Network (VHSP-PON) initiative, ITU-T is currently analyzing which physical-layer transmission solution will be the best for next-generation 200G PON, and one of the possible options under study is digital subcarrier-based coherent PON (DSC-PON). Under this scenario, we present in this paper a study on the heuristic optimization of dynamic time and subcarrier allocation in DSC-PON, under realistic assumptions in terms of limited bandwidth and limited laser tunability at the optical network unit (ONU) side receiver. In this paper, we propose a heuristic Greedy Graph Coloring Algorithm (GGCA) to solve subcarriers (SCs) and timeslot allocation problems that map the traffic generated by the users. Moreover, we also consider tunability of the ONU that switches the subcarrier block in different time-frames. Consequently, our proposed GGCA algorithm follows such tunability constraints of the local oscillator and provides allocation of the SCs to maximize the throughput of DSC-PON. We evaluate the performance of our proposed GGCA algorithm with different tunability times and show that it is improved approximately (35%) of throughput with full tunability compared to slow tunability.