基于Hadoop一种移动云计算本地化调度算法的研究

RESEARCH ON A HADOOP-BASED LOCALISED SCHEDULING ALGORITHM FOR MOBILE CLOUD COMPUTING

移动云计算将云计算的概念引入到移动互联网中。在对移动终端和云计算进行研究的基础上,针对移动终端请求计算量小、并发性高、用户实时性要求高的特点,提出一种基于Hadoop的移动云计算本地化调度算法。同时针对实际应用场景可能出现的热点数据问题,给出改进方案。实验结果表明改进后的本地化调度算法能够缩短系统响应时间,提高系统的吞吐率。

Mobile cloud computing introduces the concept of cloud computing to mobile intemet. Based on studying mobile terminals and cloud computing, and considering the characteristics of mobile terminals in low requested computation workload, high concurrency and high real-time property demand, we propose in this paper a Hadoop-based localised scheduling algorithm for mobile cloud computing. At the same time we come up with an improvement solution against the hot data problem in actual application environment that may appear. Experimental results show that the improved localised scheduling algorithm can shorten the system response time and improve the system throughput ratio.