IEEE国际通信会议 (IEEE International Conference on Communications）汇报报告
会议名称：IEEE ICC 2018
全称：IEEE国际通信会议(IEEE International Conference on Communications）
报告题目：Ultra-low Latency Service Provision in Edge Computing
报告摘要：Edge Computing is emerging as a promising solution to meet the ultra-low latency requirement of data processing at the edge of the Internet, and it is close to end users and the smart devices of Internet of Things. We propose an Edge Computing task scheduling model which utilizes the existing resources to achieve low latency by cooperative computing through multiple edge servers and close-range communication at the edge of the Internet. We treat the latency minimization design as an optimization problem. We formulate the latency minimization as an integer programming problem and solve it efficiently via dynamic programming, then we propose an optimal scheduling algorithm based on dynamic programming (OSA-DP). Considering the limited heterogeneous resources shared among tasks, we further propose a cooperative taskserver matchmaking scheduling heuristic (CTMS), which jointly optimizes the computation and communication cost. Extensive simulations demonstrate that ultra-low latency service provision can be achieved by the cooperative design of computation and communication in Edge Computing.
全称：IEEE国际通信会议(IEEE International Conference on Communications)
报告题目：Mobile Resource Aware Scheduling for MobileEdge Environment
报告摘要：In stream processing applications, a data stream is acontinuous stream of data items that are generated from multiplesources distributed at various geographic locations. A commonmethod of streaming processing is to transfer raw data streamsto a data center for unified processing. However, the methoddoes not scale well when a huge amount of data for streamprocessing is generated at the edge of the Internet, with thedevelopment of smartphones, Internet of things, 5G and othertechnologies in recent years. For stream processing applications,processing data at the edge can significantly reduce the responselatency of the applications. However, the mobility of edge nodesin a mobile edge environment poses a significant challenge toscheduling stream processing tasks efficiently to achieve highsystem throughputs. In this paper, we introduce a schedulingalgorithm, referred to as Mobile Resource Aware (MRA) streamprocessing scheduling, for mobile edge environment. Comparedwith other existing scheduling algorithms, our MRA algorithmcan optimally schedule resources for stream processing tasksthrough adapting to the mobile edge environment with limitednode resources. We implement MRA scheduling algorithm inStorm through a custom scheduler and we evaluate the performance of MRA in an emulation mobile edge environment. Ourexperimental results have demonstrated that our MRA algorithmcan achieve significantly higher system performance than theother two existing scheduling algorithms.