Energy-Efficient Sleep-Aware Slicing-Based Scheduler (SA-SBS) for Multi-Operators Virtualized Passive Optical Networks

Abstract

Due to the long reach of the next-generation passive optical network (PON), the idea of sharing infrastructure between multiple PON operators has emerged to reduce capital and operational expenditures. However, the standard transmission convergence layer requires an independent virtual bandwidth management process for each operator. Each operator uses a virtual dynamic bandwidth assignment (vDBA) for upstream bandwidth control. An additional bandwidth scheduler is required to create a single physical bandwidth map (phyBW-map) with non-overlapping time slots for each optical network terminal (ONT) or optical network unit (ONU) in the upstream frame. However, the existing slicing-based scheduler (SBS) using dynamic bandwidth assignment schemes does not perform well with the cyclic sleep process (CSP) for energy conservation in PON due to inefficient utilization of the residual bandwidth of the asleep ONUs of an operator. This work introduces an energy-efficient sleep-aware slicing-based scheduler (SA-SBS) for the multi-operator scenario of ten gigabit-capable symmetric PONs (XGS-PONs). Our approach (SA-SBS) achieves 7-10% better energy savings than the existing slicing-based scheduler (SBS) without compromising upstream delays and delay variance. Moreover, we demonstrate that our scheme outperforms the original SBS, resulting in an average delay(s) reduction of up to 28.5% for operators 1, 2, 3, and 4, respectively. Our findings can significantly improve the energy efficiency and latency performance of virtual PONs in multi-operator scenarios, paving the way for a sustainable and reliable communication infrastructure.

Keywords

• Bandwidth assignment
• bandwidth efficiency
• energy efficiency
• MT-vPON
• vPON