Research On Energy-saving Mechanism Of Cloudlet Enhanced Fiber-Wireless Access Networks With Offload Collaberation

The integrated fiber-wireless access network effectively combines low-cost,highbandwidth and highly-reliable optical networks with highly mobile and highly flexible wireless networks,which becomes one of the important solutions for the network architecture of mobile edge computing systems.The statistical results show that the energy consumption of the access network occupies a large proportion of the energy consumption of the entire communication network,and as the popularity of 5G applications and the number of communication devices increase exponentially,its energy consumption will continue to increase.Therefore,designing an effective offloading energy-saving mechanism is of great significance for reducing the energy consumption of integrated fiber-wireless access network.There is a problem in integrated fiberwireless access network that the utilization of the optical network unit is low,and the controlling overhead is large in the data transmission process.How to reduce the energy consumption of the entire network on the basis of ensuring network performance and offloading efficiency is a major problem in integrated fiber-wireless access network.Firstly,the architecture and key technologies of integrated fiber-wireless access network is introduced,and the dynamic bandwidth allocation mechanism in cloudlet enhanced fiber-wireless access network is analyzed.And then,the existing energy-saving mechanism is introduced,including offloading energy saving mechanism and transmission energy saving mechanism in optical domain and wireless domainSecondly,an energy-saving offloading strategy based on two-layer Stackelberg game is presented.Firstly,based on the characteristics of the traffic carried by the integrated fiber-wireless access network.Considering the quality of service of the offloading traffic,the congestion overhead and the offloading revenue coefficient are quantified.By analyzing the users' offloading request,The utility function of the game participants is given under the offloading collaberation sleep mechanism,which builds the two-layer Stackelberg game model.Then,the inverse inductive method and the twostage distributed algorithm are used to realize the Nash equilibrium of each subgame.The simulation results show that the proposed offloading strategy can effectively reduce the network energy consumption while ensuring the participants' revenue.Thirdly,this thesis proposes an optical network unit offload collaboration sleep mechanism with load transfer.By analyzing the current load of optical network unit,combineing the wireless domain multi-hop transmission delay and the reporting frame transmission time of the target optical network unit to determine the sleeping optical network unit and the destination optical network unit,and then the optical network unit cooperatively considers the arrival time of the receiving data of the edge server and the sending time of the wireless domain's controlling frame.The most suitable sleep duration is selected to reduce the controlling overhead,thereby the energy consumption is reduced when offloading,meanwhile the offloading delay is guaranteed.The simulation results show that the proposed mechanism can guarantee the delay performance of the offloading service while effectively reducing the network energy consumption.Finally,the main work and innovations of this thesis are summarized,and the outlook is presented.