A 6G-Driven Multiclass Power-Efficient Dynamic Bandwidth Allocation (MPE-DBA) Scheme for Passive Optical Network (PON)

The latest ITU IMT-2030 recommendations for sixth generation (6G) networks have imposed strict specifications on communication systems. Some of these requirements include increased throughput, ultra-low delay and jitter, differentiated services, and energy efficiency. Current dynamic bandwidth allocation (DBA) schemes for passive optical networks (PON) may meet some of these requirements, yet they fail to fulfill other recommendations, especially energy efficiency. Therefore, we propose a new PON architecture, whose objective is to offer flexibility in meeting the IMT-2030 recommendations. It uses a multiclass power efficient dynamic bandwidth allocation (MPE-DBA) scheme that helps achieving both differentiated services and sustainability in terms of energy and cost. For computational efficiency, we propose a two stages Mamdani Fuzzy Inference System (FIS). The inputs of the first FIS are the latency and cost of the 6G traffic, whereas the latency and cost of the non-6G (N6G) traffic are the inputs of the second FIS stage. Both FISs use the variation in the number of channels as output. The proposed algorithm achieves less than 100μ s delay, less than 10μ s jitter and high aggregate throughput for the 6G packets. In addition, it reduces the power consumption by three times and the cost of traffic transmission by four times as compared to the state-of-the art solution.