Bandwidth Scheduling Algorithm Of Next Generation PON System Supporting Multi-service

With the continuous development of network technology,the next-generation PON system not only needs to provide large bandwidth,but also needs to meet the needs of differentiated business services such as low latency and high reliability.The 5G service carried by the PON system is the main development trend of the PON system.The wide variety of 5G services and the large differences in service requirements pose challenges to the scheduling of multi-service systems based on PON bearers.While the PON system needs to support large capacity and high bandwidth,it also needs to consider its system upgrade cost.Multi-tenant sharing PON network has become a potential solution,and multi-tenant sharing allocation scheduling algorithm has also become one of the research hotspots.To this end,this article focuses on the next-generation PON system bandwidth scheduling algorithm that supports multiple services.The research content of this paper is as follows:(1)Summarizes the next-generation PON system and its bandwidth scheduling algorithm,as well as the research background significance of the multi-service coexistence scenario based on the PON system and the multi-tenant shared PON system;According to the analysis of the current domestic and foreign research status of the bandwidth scheduling algorithm of the next-generation PON system that supports multi-services,the problems to be solved in this paper are obtained.(2)Summarizes the next-generation PON network architecture that supports the coexistence of multiple services;Among them,a detailed description and analysis of the high-level and low-level functional split options for accessing data in Xhaul,a 5G multi-service network architecture based on PON bearers;It also analyzes the 5G frame structure and the structure of the mini-slot introduced to support multiple services,which provides a basis for the research of scheduling algorithms in the Xhaul system that supports multiple services;(3)Regarding the problem of differentiated service requirements caused by the coexistence of multiple services in the PON system,this article establishes a mathematical model based on the PON system carrying EMBB and URLLC services in 5G in the Xhaul network with the goal of maximizing service throughput and the constraint of delay time,And on this basis,a joint optimization algorithm is proposed to simultaneously support 5G fronthaul and midhaul network coexistence,and EMBB and URLLC business coexistence.The specific plan of the algorithm is: reserve bandwidth in each mini-slot for the most delay-sensitive URLLC service to ensure the low latency of the URLLC service;use the dynamic programming method to allocate bandwidth that can meet the delay constraint for the EMBB service in the fronthaul network;on the basis of the above two steps,allocate the remaining bandwidth to the EMBB traffic in the midhaul network in a first-come,first-served manner.At the same time,this paper verifies the feasibility of the algorithm by building a simulation system.The simulation results show that the algorithm designed in this paper can effectively improve the total throughput of service data while meeting the delay requirements of various services.(4)Regarding the resource allocation problem caused by operators' shared PON network,this paper proposes a multi-tenant shared PON system bandwidth scheduling algorithm based on deep reinforcement learning(DRL).This algorithm maps the multi-tenant PON system to the reinforcement learning model.Among them,the environment in DRL is each bandwidth request to be processed and the current remaining bandwidth,and the decision made is the bandwidth request selected for processing at the current moment.The DRL agent interacts with the environment and continuously updates the policy parameters until the model converges,thereby completing the optimization.The feasibility of the algorithm is verified by building a simulation system,and the simulation results show that the proposed algorithm can effectively improve the utilization of bandwidth resources.