SISTEMI PER LA MOBILITÀ DEGLI UTENTI E DEGLI APPLICATIVI IN RETI WIRED E WIRELESS

The words mobility and network are found together in many contexts. The issue alone of modeling geographical user mobility in wireless networks has countless applications. Depending on one’s background, the concept is investigated with very different tools and aims. Moreover, the last decade saw also a growing interest in code mobility, i.e. the possibility for soft-ware applications (or parts thereof) to migrate and keeps working in different devices and environ-ments. A notable real-life and successful application is distributed computing, which under certain hypothesis can void the need of expensive supercomputers. The general rationale is splitting a very demanding computing task into a large number of independent sub-problems, each addressable by limited-power machines, weakly connected (typically through the Internet, the quintessence of a wired network). Following this lines of thought, we organized this thesis in two distinct and independent parts: Part I It deals with audio fingerprinting, and a special emphasis is put on the application of broadcast mon-itoring and on the implementation aspects. Although the problem is tackled from many sides, one of the most prominent difficulties is the high computing power required for the task. We thus devised and operated a distributed-computing solution, which is described in detail. Tests were conducted on the computing cluster available at the Department of Engineering of the University of Ferrara. Part II It focuses instead on wireless networks. Even if the approach is quite general, the stress is on WiFi networks. More specifically, we tried to evaluate how mobile-users’ experience can be improved. Two tools are considered. In the first place, we wrote a packet-level simulator and used it to esti-mate the impact of pricing strategies in allocating the bandwidth resource, finding out the need for such solutions. Secondly, we developed a high-level simulator that strongly advises to deepen the topic of user cooperation for the selection of the “best” point of access, when many are available. We also propose one such policy.