Reducing latency for mobile devices

Abstract

The massive development of the web has caused the servers to be overloaded and causes heavy amounts of network traffic. A Lot of users around the world are concurrently accessing the web, hence due to insufficient bandwidth bottleneck is created. To make web service better, rapid time of response is necessary to maintain the user's connection to the web, and therefore, many ways need to be created to diminish the latency of connecting to web pages. In this case, web caching o ers an e ective approach by temporarily saving files close by, minimizing time of access, yet restricting the device to a degree. Web prefetching, which is a more e ective strategy, is designed to fetch pages beforehand, additionally reducing latency. In the past, several Web prefetching methods have been proposed, incorporating di erent web and data mining techniques. To analyze proxy server data Web Use Mining has been integrated, and to get and to get a perception into user's web browsing trends, and then create regulations. Thus, we propose a Link analysis method, with a modified weighted Hits based algorithm that mainly focuses on two traits of how similar and popular the page is.These are integrated with other rules to rank the pages to be perfected. After the experimentation results reveal that the modified Hits algorithm can discover more relevant pages compared to Arc and Hits.