期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2013
卷号：4
期号：1
页码：495-496
语种：English
出版社：Ayushmaan Technologies
摘要：For data delivery, a wide range of emerging applications confronts the existing techniques to users and applications accessing data from multiple autonomous servers. We develop a framework an alternative and more flexible approach that maximizes user utility for comparing pull based solutions and present dual optimization approaches. It maximizes the user utility while satisfying constraints on the usage of system resources it is the first approach and it satisfies the utility of user profiles while minimizing the usage of system resources it is the second approach. For the latter approach, we present a static optimal solution (SUP) and formally identify sufficient conditions for SUP to be optimal for both. A limitation of static solutions to pullbased delivery is that they cannot adapt to the dynamic behavior of large area applications. Moreover, to improve user utility with only a moderate increase in resource utilization an adaptive algorithm (fbSUP) is proposed and shows how it can incorporate feedback. We can analyze the behavior of SUP and fbSUP under various update models by using real and synthetic data traces. When the estimations of SUP closely estimate the real event stream, we can achieve a high degree of satisfaction of user utility, and has the potential to save a significant amount of system resource is shown in our experiments. We further show that SUP can exploit feed back to improve user utility with only a moderate increase in resource utilization.
关键词：Online Data Delivery;Framework;Dual Optimization;Pull Based Delivery;Adaptive Algorithm
