Network connection selection : solving a new multiobjective optimization problem

Abstract

Mobile terminals are nowadays able to use several different kinds of wireless network connections to transfer data. The users of mobile terminals can select the network connection that is the most suitable for their purposes. In this thesis, we consider a more advanced case in which a mobile terminal can use several different connections simultaneously to transfer data. When the data is composed of separate components, the components can be transferred using different connections. Then the user needs to make a selection which network connection to use for each component. This kind of selection may be very difficult to make and most of the users are not interested in making such selections. Therefore, an automatic selection method is needed. The network connection selection is modelled as a multiobjective optimization problem in which the aim is to minimize the time used in transferring and the costs of the transfer. To solve the problem, we have developed a fast heuristic solution method that uses only little computational capacity so that it is possible to implement it in a mobile terminal. In addition, the method takes into account possibly changing network environment during the transmission and improves, when possible, the transmission schedule formed by selecting a network connection for each component of the data.