面向资源使用的TTL缓存服务计费机制与存储策略
|
摘要
为刻画缓存服务提供商和内容提供商在网络缓存服务中的经济学行为,研究了网络边缘TTL缓存的服务计费机制与存储策略,提出了基于内容文件缓存命中速率以及内容文件缓存逗留时间的计费模型。采用Stackelberg博弈模型对缓存服务定价与存储策略问题进行建模,并求解了在均一定价和差异定价策略下的最优缓存价格和内容文件缓存时间。仿真实验对比了在不同计费模型和定价策略下缓存服务提供商获得的投资回报以及内容提供商产生的收益,结果表明:差异定价策略性能要优于均一定价策略;在差异定价策略以及基于内容文件缓存逗留时间的均一定价策略下,内容提供商获得的收益显著大于缓存服务提供商得到的回报,而在基于内容文件缓存命中速率的均一定价策略下情况正好相反;在均一定价策略下,基于内容文件缓存命中速率的计费模型比基于内容文件缓存逗留时间的模型带给缓存服务提供商更多回报。
Pricing schemes and caching policy for TTL(time-to-live)caches located at the network edges are studied to characterize the economic behaviors of content providers and caching service providers in web cache services.Both pricing models based on cache hit rate and cache sojourn time of content files are proposed.The problem of setting appropriate prices and TTLs of files for maximum revenue is modeled as a Stackelberg game,and the solutions are obtained under both uniform and non-uniform pricing schemes.The revenue of caching service provider and the gain of content provider are compared through simulation,and it is found that non-uniform pricing schemes outperform uniform pricing ones.Furthermore,the gain of content provider is much larger than the revenue of caching service provider under non-uniform pricing schemes and the sojourn time based uniform pricing schemes,whereas the situation is the opposite under the hit rate based uniform pricing schemes.Finally,the hit rate based uniform pricing schemes bring more gains for content providers than the sojourn time based uniform pricing schemes do.Simulation results indicate that the proposed pricing schemes and caching policies can be applied in real network environments.
Pricing schemes and caching policy for TTL(time-to-live)caches located at the network edges are studied to characterize the economic behaviors of content providers and caching service providers in web cache services.Both pricing models based on cache hit rate and cache sojourn time of content files are proposed.The problem of setting appropriate prices and TTLs of files for maximum revenue is modeled as a Stackelberg game,and the solutions are obtained under both uniform and non-uniform pricing schemes.The revenue of caching service provider and the gain of content provider are compared through simulation,and it is found that non-uniform pricing schemes outperform uniform pricing ones.Furthermore,the gain of content provider is much larger than the revenue of caching service provider under non-uniform pricing schemes and the sojourn time based uniform pricing schemes,whereas the situation is the opposite under the hit rate based uniform pricing schemes.Finally,the hit rate based uniform pricing schemes bring more gains for content providers than the sojourn time based uniform pricing schemes do.Simulation results indicate that the proposed pricing schemes and caching policies can be applied in real network environments.
引文
[1]Sandvine.Global internet phenomena report 2H 2012[EB/OL].(2012-11-22)[2017-09-26].https:∥www.sandvine.com/hubfs/downloads/archive/2012-2h-global-internet-phenomena-report.pdf.
[2]SHAMMA D A,FRIEDLAND G,ELIZALDE B,et al.YFCC100M:the new data in multimedia research[J].Communications of the ACM,2016,59(2):64-73.
[3]Cisco.Cisco visual networking index:forecast and methodology,2014-2019[EB/OL].(2015-05-27)[2017-09-26].http:∥s2.q4cdn.com/230918913/files/doc_downloads/report_2014/white_paper_c11-481360.pdf.
[4]Wikipedia.Content delivery network[EB/OL].(2017-09-01)[2017-09-26].https:∥en.wikipedia.org/wiki/Content_delivery_network.
[5]PALLIS G,VAKALI A.Insight and perspectives for content delivery networks[J].Communications of the ACM,2006,49(1):101-106.
[6]JACOBSON V,SMETTERS D K,THORNTON J D,et al.Networking named content[C]∥Proceedings of the International Conference on Emerging Networking Experiments and Technologies.New York,USA:ACM,2009:1-12.
[7]CHOI J,HAN J,CHO E,et al.A survey on contentoriented networking for efficient content delivery[J].Communications Magazine IEEE,2011,49(3):121-127.
[8]吴超,张尧学,周悦芝,等.信息中心网络发展研究综述[J].计算机学报,2015,38(3):455-471.WU Chao,ZHANG Yaoxue,ZHOU Yuezhi,et al.A survey for the development of information-centric networking[J].Chinese Journal of Computers,2015,38(3):455-471.
[9]Wikipedia.Stackelberg competition.[EB/OL].(2017-01-23)[2017-09-26].https:∥en.wikipedia.org/wiki/Stackelberg_competition.
[10]BERGER D S,GLAND P,SINGLA S,et al.Exact analysis of TTL cache networks[J].Performance Evaluation,2014,79:2-23.
[11]FOFACK N C,DEHGHAN M,TOWSLEY D,et al.On the performance of general cache networks[C]∥Proceedings of the 8th International Conference on Performance Evaluation Methodologies and Tools.New York,USA:ACM,2014:106-113.
[12]DEHGHAN M,MASSOULIE L,TOWSLEY D,et al.A utility optimization approach to network cache design[C]∥Proceedings of the 2016IEEE INFOCOM.Piscataway,NJ,USA:IEEE,2016:1-9.
[13]HONIG M L,STEIGLITZ K.Usage-based pricing of packet data generated by a heterogeneous user population[C]∥Proceedings of the 14th Joint Conference of the IEEE Computer and Communication Societies.Piscataway,NJ,USA:IEEE,1995:867-874.
[14]NEVO A,TURNER J L,WILLIAMS J W,et al.Usage-based pricing and demand for residential broadband[J].Econometrica,2016,84(2):411-443.
[15]MA R T.Usage-based pricing and competition in congestible network service markets[J].IEEE ACM Transactions on Networking,2016,24(5):3084-3097.
[16]SHEN H,BASAR T.Optimal nonlinear pricing for a monopolistic network service provider with complete and incomplete information[J].IEEE Journal on Selected Areas in Communications,2007,25(6):1216-1223.
[17]ZEKRI M,HADJI M,JOUABER B,et al.A Nash Stackelberg approach for network pricing,revenue maximization and vertical handover decision making[C]∥Proceedings of the 36th IEEE Conference on Local Computer Networks.Piscataway,NJ,USA:IEEE,2011:622-629.
[2]SHAMMA D A,FRIEDLAND G,ELIZALDE B,et al.YFCC100M:the new data in multimedia research[J].Communications of the ACM,2016,59(2):64-73.
[3]Cisco.Cisco visual networking index:forecast and methodology,2014-2019[EB/OL].(2015-05-27)[2017-09-26].http:∥s2.q4cdn.com/230918913/files/doc_downloads/report_2014/white_paper_c11-481360.pdf.
[4]Wikipedia.Content delivery network[EB/OL].(2017-09-01)[2017-09-26].https:∥en.wikipedia.org/wiki/Content_delivery_network.
[5]PALLIS G,VAKALI A.Insight and perspectives for content delivery networks[J].Communications of the ACM,2006,49(1):101-106.
[6]JACOBSON V,SMETTERS D K,THORNTON J D,et al.Networking named content[C]∥Proceedings of the International Conference on Emerging Networking Experiments and Technologies.New York,USA:ACM,2009:1-12.
[7]CHOI J,HAN J,CHO E,et al.A survey on contentoriented networking for efficient content delivery[J].Communications Magazine IEEE,2011,49(3):121-127.
[8]吴超,张尧学,周悦芝,等.信息中心网络发展研究综述[J].计算机学报,2015,38(3):455-471.WU Chao,ZHANG Yaoxue,ZHOU Yuezhi,et al.A survey for the development of information-centric networking[J].Chinese Journal of Computers,2015,38(3):455-471.
[9]Wikipedia.Stackelberg competition.[EB/OL].(2017-01-23)[2017-09-26].https:∥en.wikipedia.org/wiki/Stackelberg_competition.
[10]BERGER D S,GLAND P,SINGLA S,et al.Exact analysis of TTL cache networks[J].Performance Evaluation,2014,79:2-23.
[11]FOFACK N C,DEHGHAN M,TOWSLEY D,et al.On the performance of general cache networks[C]∥Proceedings of the 8th International Conference on Performance Evaluation Methodologies and Tools.New York,USA:ACM,2014:106-113.
[12]DEHGHAN M,MASSOULIE L,TOWSLEY D,et al.A utility optimization approach to network cache design[C]∥Proceedings of the 2016IEEE INFOCOM.Piscataway,NJ,USA:IEEE,2016:1-9.
[13]HONIG M L,STEIGLITZ K.Usage-based pricing of packet data generated by a heterogeneous user population[C]∥Proceedings of the 14th Joint Conference of the IEEE Computer and Communication Societies.Piscataway,NJ,USA:IEEE,1995:867-874.
[14]NEVO A,TURNER J L,WILLIAMS J W,et al.Usage-based pricing and demand for residential broadband[J].Econometrica,2016,84(2):411-443.
[15]MA R T.Usage-based pricing and competition in congestible network service markets[J].IEEE ACM Transactions on Networking,2016,24(5):3084-3097.
[16]SHEN H,BASAR T.Optimal nonlinear pricing for a monopolistic network service provider with complete and incomplete information[J].IEEE Journal on Selected Areas in Communications,2007,25(6):1216-1223.
[17]ZEKRI M,HADJI M,JOUABER B,et al.A Nash Stackelberg approach for network pricing,revenue maximization and vertical handover decision making[C]∥Proceedings of the 36th IEEE Conference on Local Computer Networks.Piscataway,NJ,USA:IEEE,2011:622-629.