多主多从Stackelberg博弈下的边缘缓存资源分配算法  

Edge Cache Resource Allocation Algorithm Based on Multi-Leaders and Multi-Followers Stackelberg Game

作　　者：王磊[1] 李金城 康彬 崔景伍[1] 郑宝玉[1] Wang Lei;Li Jincheng;Kang Bin;Cui Jingwu;Zheng Baoyu(College of Telecommunication & Information Engineering,Nanjing University of Posts and Telecommunications,Nanjing,Jiangsu 210003,China;College of Internet of Things,Nanjing University of Posts andTelecommunications,Nanjing,Jiangsu 210003,China)

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003 [2]南京邮电大学物联网学院,江苏南京210003

出　　处：《信号处理》2019年第4期574-581,共8页Journal of Signal Processing

基　　金：国家自然科学基金(61571240,61671253,61801242);江苏省高校自然科学研究重大项目(16KJA510004);江苏省自然科学基金青年基金(BK20170915);南京邮电大学科研基金(NY218012)

摘　　要：为鼓励视频服务提供商参与到缓存过程中,本文提出一种基于Stackelberg博弈的激励缓存资源分配算法。与传统激励缓存资源分配方案不同,本文考虑同时存在多个网络运营商和多个视频服务提供商,视频服务提供商从网络运营商处购买存储空间以缓存热门视频。针对该场景,本文将该激励缓存模型建模为多主多从Stackelberg博弈问题,分别构建主方和从方的效用函数,证明了在网络运营商价格确定的情况下,视频服务提供商之间的非合作博弈存在纳什均衡。文章利用分布式迭代算法对该博弈模型进行求解,获得了视频服务提供商的最优缓存策略和网络运营商的最优价格策略。仿真结果表明,本文提出的激励缓存机制可使视频服务提供商获得比其他缓存分配算法更高的单位成本收益。In order to encourage video service providers(VSPs)to participate in the caching process,an incentive cache resource allocation algorithm based on Stackelberg game is proposed in this paper.Different from traditional incentive cache resource allocation scheme,this paper considers the circumstance that there are multiple network operators(MNOs)and multiple VSPs at the same time.The VSPs purchase storage space from the MNOs to cache popular videos.We formulate the incentive cache model as a multi-leaders and multi-followers Stackelberg game problem,and construct the utility functions of the leaders and the followers respectively.We prove that in the case of given MNOs’price,there is a Nash equilibrium in VSPs’non-cooperative game.We use the distributed iterative algorithm to solve the game model,and obtain the optimal cache strategy of the VSPs and the optimal price strategy of the MNOs.The simulation results show that the incentive caching mechanism proposed in this paper can make the VSPs obtain higher unit cost benefit than other cache allocation algorithms.

关 键 词：内容缓存 资源分配 STACKELBERG博弈 激励机制 纳什均衡 

分 类 号：TN92[电子电信—通信与信息系统]