Cloud computing is a rapidly growing technology in recent years since it can provide users and enterprises with powerful computing capabilities to store and process data in its datacenters at low prices. However, as there are limited number of cloud datacenters across the world, accessing a distant cloud will cause long response latency and as a result the cloud computing performance will be adversely affected. Therefore, cloudlet has been proposed as a promising paradigm to solve this problem. Since a cloudlet is located at the edge of the Internet, it can be viewed as a small-scale cloud brought closer to users. Therefore, it can provide powerful computing resource to users with reduced response latency. However, since a cloudlet is a small-scale cloud, it is important to decide which content to be cached to its datacenter. In this paper, a new algorithm for cloudlet content caching based on web mining is developed. The new algorithm first exploits the k-means clustering to discover different accessing patterns and the most frequently accessed web objects to be cached. Then the association rule is utilized to prefetch web objects which will be potentially accessed. The algorithm also takes the sizes of web objects and cloudlet into consideration. The experiment implemented with the new algorithm was conducted and the result shows that compared to traditional caching algorithms, the proposed new algorithm will increase the cloudlet content cache hit ratio and reduce the response latency.
Software And Hardware
• Hardware: Processor: i3 ,i5 RAM: 4GB Hard disk: 16 GB • Software: operating System : Windws2000/XP/7/8/10 Anaconda,jupyter,spyder,flask Frontend :-python Backend:- MYSQL