Energy Efficiency Management Mechanism Of Access Points In Software Defined Wireless Networks

Software defined wireless network(SDWN)is referred as one of prospective key techniques for 5G,which has been increasing the network's manageability,controllability,scalability,optimizing the scheduling of network resources,and can provide users with reliable high-bandwidth.A large number of wireless access points(AP)are deployed densely in SDWN to meet users' increasing demands of network performance.Concerning users' behaviors present tidal characteristics and lots of idle APs running with full power when the system maintaining a relatively low load,these issues bring up a serious energy problem.Aiming at above sloving these problems,this thesis proposes an AP energy efficiency management mechanism based on reverse auctions using SDWN's network programmable features and centralized control characteristics.This mechanism reduces the network redundancy energy consumption by sleeping idle and low-load AP,chossing reasonable APs for offloading users to reduce energy consumption furtherly while protecting the users' QoE.The factors affecting AP energy consumption are measured and analyzed,and the relationship between load and energy consumption is quantified by using regression analysis.Based on the analysis results,a reasonable AP sleep wake-up threshold is obtained,and the sleep wake-up matrix is determined according to the association between users and APs.This thesis introduces the Vickrey-Clarke-Groves(VCG)reverse auction mechanism to optimize AP selection of sleeping-APs' offloading users,constructs the Mixed Integer Linear Programming(MILP)to determine the resource allocation,and uses the VCG rules to set the payment method,then designs the AP energy efficiency management mechanism based on the VCG reverse auction and the optimal algorithm.However,when handling with enoumous network nodes,the optimal algorithm cannot meet the real-time demands of the network due to the high computational complexity.Therefore,this thesis puts forward the AP energy efficiency management mechanism based on greedy reverse auction,uses the idea of greedy algorithm to solve the resource allocation problem,improves the second price mechanism to determine the payment method,and designs the fast AP management mechanism and the fast implementation algorithm.The experimental results show that the AP energy efficiency management mechanism can reduce 50%~75% energy consumptions while guaranteeing the effective throughput and coverage of the system.The energy-saving effect has been improved significantly compared with existing designs.This mechanism can reduce the network energy consumption effectively,thus saving the network costs which is essential for the construction of the green network.