Middleware for Context-Aware Opportunistic Networks

Abstract: Mobile devices such as palmtops and cell phones are continuously increasing in capabilities and popularity. At the same time, due to their decreasing price they are becoming more and more attractive and available to the average customer. This has lead to the development of many new applications for such portable electronic devices. Mobile devices tend to have increasing local resources in terms of memory/storage and CPU [2]. Despite these improvements in hardware attributes, there are still limitations that characterize these portable devices and which have not improved as quickly as the increase of the local computational power. These limitations mostly concern the network resources and battery power that are both still rather poor. Thus the main barriers for mobile nodes are network resources and limiting the power consumption of the device itself. Today, wireless networks provide limited reliability and less bandwidth than fixed networks. Moreover, all mobile nodes are highly energy dependent as they use batteries with a limited capacity. Additionally, roaming is a feature that increasingly must be supported for such wireless devices, as their physical portability leads to users to use them even as they move about. This may require the utilization of different wireless networks while the node is on the move. Therefore, for all the above reasons there is a demand for the development of intelligent mechanisms and techniques for optimizing the management of these limited resources, while exploiting the local resources, thus providing users with the best possible performance within the available resources. At the present, there are operating systems, such as the Symbian OS [22], specially designed for supporting advanced features in mobile computing. However, there is still development to be done. Moreover, although there are many new applications for mobile computing, they are not yet sophisticated enough to cope with changes in the wireless environment, these changes occur due to the node’s change in context. This creates a gap that should be filled by software between the applications and the operating system; this is frequently called middleware. This middleware provides a collaborative partnership between the operating system and the applications, assisting and making both more sophisticated, in terms of scheduling and managing traffic in a wireless environment. The focus of this project is how to utilize such middleware to best serve the needs of the mobile user.