``Analysis of Multimedia Workloads with Implications for Internet Streaming",

Lei Guo, Songqing Chen, Zhen Xiao, and Xiaodong Zhang

Proceedings of the 14th International World Wide Web Conference (WWW'2005),  
 Chiba, Japan, May 10-14, 2005. 

Abstract

In this paper, we study the media workload collected from a large number of 
commercial Web sites hosted by a major ISP and that collected from a large 
group of home users connected to the Internet via a well-known cable company. 
Unlike other studies that only focus on the characteristics of streaming or 
downloading media traffic either on a client side or on a server side, we 
present a comprehensive study of the overall media content delivery mechanisms 
on the Internet from both the server side and the client side. Some of our 
key findings are: (1) Surprisingly, the majority of media contents are still 
delivered via downloading from Web servers. (2) A substantial percentage of 
media downloading connections are aborted before completion due to the long 
waiting time and low patience of clients, resulting up to 20% pure bandwidth 
waste. (3) A hybrid approach, pseudo streaming, is used by clients to imitate 
real streaming. Compared with downloading, clients using pseudo streaming 
tend to abort more connections and abort earlier due to the early feedback 
of the media content, and result in less bandwidth waste. (4) The mismatch 
between the downloading rate and the client playback speed in pseudo streaming 
is common, which either causes frequent playback delays to the clients, or 
unnecessary traffic to the Internet. (5) Compared with streaming, downloading 
and pseudo streaming are neither bandwidth efficient nor performance 
effective. Our findings indicate that inadequate streaming support has 
caused many clients to suffer from poor quality of service and inefficient 
Internet bandwidth utilization. To address this problem, we propose the 
design of AutoStream, an innovative system that can provide additional 
previewing and streaming services automatically for media objects hosted 
on standard Web sites in server farms at the client's will. Our trace driven 
simulations show that AutoStream has the potential of greatly improving the 
quality of media delivery to clients while reducing network bandwidth usage 
substantially.