Networks are becoming more and more popular and the traffic
over the networks is growing rapidly. Because the network infrastructure is
expanding at a quick pace, especially the growth of the Internet, both in size
and in the types of applications, the performance issue and dimensioning issue
of networks are becoming more critical. It is important to provide a robust and
flexible networking environment.
To design a system that can deal with these issues well at a reasonable cost, the first step is to characterize the network traffic. In order to get traffic properties that are general it is necessary to study traces of networks in various environments and of different types. A detailed analysis and characterization of network traffic is fundamental to obtain deeper insights into the network system, to fully realize the potential improvements in network performance, and to optimize management of the resources.
A lot of work has been done in traffic analysis. The major part is on the Internet traffic. Recently, wire less communication has moved from voice service to data service. Wireless Application Protocol (WAP) makes it possible for commonly used devices such as cell phones, PCs and PDAs to access the Intern et. These new wireless data applications produce data that may have different characteristics from those of wired data applications and wireless voice data studied before.
In this thesis, we study the data generated by Mobile Browser applications in a cellular network of Bell Mobility in Quebec and Ontario, and compare our results with other results of Internet traffic characterization. We also do some simple performance prediction of the cellular network using Layered Queuing Models (LQM).