主讲人：torsten braun, 瑞士伯尔尼大学计算机学院教授
时间：2018年10月25日（周四）上午 09:00
地点：吉林大学南湖校区第一教学楼304
主讲人简介： torsten braun是瑞士伯尔尼大学计算机学院教授，目前担任通信与分布式系统研究组组长、瑞士switch基金委副主席。1993年毕业于德国卡尔斯鲁厄理工学院并获得博士学位，1994年-1994年在法国inria研究所任研究员，1995年-1997年在德国ibm欧洲网络中心任科学家/高级顾问，自1998年至今担任瑞士伯尔尼大学教授，2007年-2011年期间担任瑞士伯尔尼大学计算机学院院长。torsten braun教授近年来在计算机网络、云计算、移动边缘计算等领域展开研究并取得了丰硕的成果。在ieee journal on selected areas in communications (jsac), ieee transactions on wireless communication, ieee transactions on mobile computing, acm sigcomm, ieee infocom等计算机通信领域国际顶级期刊和会议发表多篇论文。
报告主题：caching and computing in mobile edge networks
移动边缘网络中的缓存与计算
报告摘要: mobile edge computing (mec) provides computing and caching services at the edge of the wireless access network in order to minimize delay and customize services to the characteristics of the wireless access network. we introduced the mobile follow-me cloud concept where we populate caches with the most popular content requested by mobile users. the efficiency of such proactive caching very much depends on the accuracy of mobility prediction. mobility prediction can be improved by using machine learning techniques and combinations of them. however, the mobility prediction accuracy very much depends on the quality of the available mobility data. in cases, where mobility prediction is not accurate enough, we found out that the benefits of caching can be improved by placing content rather in the core network than at the edge. this approach is also beneficial for privacy reasons. moreover, we showed that mec can improve the video quality in case of dynamic adaptive streaming over http in a lte network. we are currently investigating mec to support localization of resource-constrained devices.
