Characterizing QQ Application across Internet

摘要

Instant messenger applications have large amount of users on the Internet. Understanding the characteristics of such kinds of application is very important because it is a close reflection of human behavior to use the Internet. QQ is one of the most popular instant messenger applications in China. We implement both active and passive measurements to QQ traffic on a large-scale IP backbone in China. We find that although the average QQ traffic accounts for less than 1% of the total traffic volume on the Internet, it exhibits a clear busy time in which a much larger amount of traffic occur, leading to a threat to local networks at the servers edge. Also, we observe that user behavior can be clearly depicted by measuring such instant messenger applications.