Research On Energy Saving Mechanism In Hybrid Optical Fiber Wireless Access Network

With the rapid development of communication technology,people's life style has changed greatly.Especially with the emergence of industry 4.0,more and more intelligent machines are connected to the network,followed by the exponential growth of sensing and monitoring data,which leads to high energy consumption caused by high load.As one of the goals of green communication network,energy saving has been an important research topic of fiber wireless hybrid access network(FiWi).In order to meet the higher service quality requirements of the Internet of Things users with low energy consumption,this paper integrates the metro and access network based on the traditional FiWi network architecture,introduces edge computing,and proposes Edge Cloud Enhanced Metro-FiWi.With a more energy-saving architecture,Edge Cloud Enhanced Metro-FiWi can address the greater challenges of a large number of IoT applications and devices in terms of connectivity,reliability,bandwidth,delay and cost.However,the extended access scope of the architecture inevitably brings the dense deployment of the equipment,and thus brings about a large energy consumption.Therefore,in order to achieve the green communication goal,an effective Edge Cloud Enhanced Metro-FiWi energy saving strategy is very important to reduce the energy consumption of the entire communication network.Aiming at the problem of energy wastage of ONU in Edge Cloud Enhanced Metro-FiWi,this paper studies the energy consumption minimization from the aspects of capacity resource allocation and routing algorithm,and proposes a sleep scheme design.The main contents and innovations of this paper are as follows:1.The network architecture design of Edge Cloud Enhanced Metro-FiWiThe traditional FiWi is composed of tree like passive optical network(PON)and wireless network.It combines high capacity,reliability of the optical back end and low cost,ubiquitous connectivity of the wireless fronthaul.In order to further improve the bandwidth capacity and expand the coverage of the network with low energy consumption,this paper combines the metro network with the access network to form the metro fiber wireless hybrid access network.Meanwhile,edge computing,as a promising computing paradigm,can meet the strict service quality requirements of more and more delay sensitive applications.It transfers traditional cloud resources to edge servers,which is conducive to the energy saving of terminal devices.In this paper,mobile edge computing server is deployed at the ONU,which greatly reduces latency.Therefore,this paper proposes the network architecture of Edge Cloud Enhanced Metro-FiWi.2.Design of load adaptive and location aware sleep strategyThe sleep strategies that previous research commonly used are cyclic sleep and threshold-based sleep.The cyclic sleep is to enter sleep periodically without considering the load in the network.The threshold-based sleep is to enter sleep when there is no traffic load(i.e.threshold is 0)or the traffic load is lower than the predefined threshold.However,these ONU independent energy saving methods do not consider two important issues:the first is that the number of sleep ONUs can't fully adapt to the changes of traffic load or flow distribution in the network,which results in low ONU utilization rate and limits the energy saving potential of ONU under medium traffic conditions;the second problem is that due to the lack of location awareness,the active ONU is not well distributed in geographical location,and it is possible that all active ONUs are concentrated in a heavy load area,which leads to the delay of the data flow of the user equipment in the light load area.Therefore,for the first problem,this paper proposes a load adaptive ONU sleep scheme.From the perspective of the load state of the whole network,greedy algorithm is applied to determine the optimal number of active ONUs that can carry the current network load through iteration under different load conditions,so that the number of sleep ONUs can be fully adapted to the load,thus further improving the utilization and energy efficiency of ONU.In order to optimize the configuration of ONU resources,the optimal active ONU location deployment strategy is designed based on K-means clustering and matching game,so as to realize the optimal configuration of ONU resources,while ensuring the quality of service and minimizing the number of active ONUs.3.Research on capacity delay aware energy efficient routing algorithmTraditional rerouting algorithms include shortest path algorithm,residual capacity algorithm,delay sensing algorithm(DARA),capacity delay aware algorithm(CaDAR),etc.These algorithms determine the selection of destination ONU and link with the weight of path length and residual capacity of link nodes from the view of MAP.Traditional rerouting algorithm finds the most favorable path for MAP from the view of MAP.Unlike traditional rerouting algorithm,this paper considers different optimization objectives of ONU energy saving and MAP delay reduction,and aggregates traffic to the destination nodes as few as possible to maximize the utilization of capacity resources and increase the number of sleep devices.Based on Floyd algorithm and K-shortest Path algorithm,this paper provides a self-organizing resource allocation solution by using matching game theory.It breaks through the traditional first come first service routing strategy,and uses delay acceptance algorithm to realize the optimal allocation of resources,so as to realize the multi-objective optimization of energy saving and service quality assurance of the whole network.