Abstract
â??The Bursty data applications become popular in 
 mobile networks, and the quality of end user experience of 
 these applications is vital for mobile telecom operators. 
 Mobile telecom operators and equipment providers begin to 
 design new scheduling strategies aiming at improving 
 quality of user experience. These scheduling strategies can 
 identify particular applications and assign corresponding 
 scheme to them. But traditional test approaches only 
 consider the common performance indicators, so they 
 cannot evaluate the new scheduling strategies. This paper 
 proposes an approach of replaying user behaviors to 
 evaluate the communication system. The approach uses user 
 behavior model to build test scenarios, and evaluate the 
 performance of each application via corresponding 
 particular indicators. Finally, a remote test system is built, 
 and the experiment on this system proves the effectiveness 
 of this approach
Highlights
The bursty data applications include Social Network Service (SNS), Instant Message (IM), Email, and so on
These bursty data applications become more and more popular in mobile networks. This trend motivates telecom operators and equipment providers to design new scheduling strategies for guaranteeing the quality of user experience (QoE) and saving resources [1,2,3]. These smart scheduling strategies can recognize particular applications depending on deep packet inspection (DPI), and assign particular schemes to corresponding applications to improve QoE
We survey the character of user behavior of busty data applications, and find that most data flows only belong to a few typical user actions
Summary
The bursty data applications include Social Network Service (SNS), Instant Message (IM), Email, and so on. Applications with continuous large data flow, such as video and audio applications, are excluded These bursty data applications become more and more popular in mobile networks. This trend motivates telecom operators and equipment providers to design new scheduling strategies for guaranteeing the quality of user experience (QoE) and saving resources [1,2,3]. These smart scheduling strategies can recognize particular applications depending on deep packet inspection (DPI), and assign particular schemes to corresponding applications to improve QoE. New smart scheduling strategies distinguish different applications so that each application should have its own QoE indicators and traffic model
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call