| Summary: | 博士 === 國立成功大學 === 資訊工程學系 === 102 === Opportunistic Cloud is a kind of mobile cloud systems used in opportunistic networks. Recently, with the development and popularization of mobile devices, the mobile cloud becomes an emerging network service model. Mobile devices can form mobile clouds by assembling and cooperating with each other to provide various network services. Moreover, exploiting opportunistic contacts of users, mobile clouds can establish an opportunistic network to disseminate data. Generally, because opportunistic clouds are disconnected, data dissemination in this system can be divided into two types, intra-cloud data dissemination and inter-cloud data dissemination. When delivering data inside a cloud, traditional single-hop or multi-hop communication is usually adopted. However, since users might be concentrated and moving fast which will lead to conflict and interruption in spectrum access, the delivery bandwidth is always limited. In this dissertation, applying the concept of Dynamic Spectrum Access, the cognitive radio network technique can utilize the licensed spectrums efficiently to alleviate the spectrum access problem. When disseminating data between clouds, the store-carry-forward-based methods are usually used due to the lacks of stable connections in opportunistic networks. This kind of methods can achieve high data delivery rate but suffers from heavy burdens. In this dissertation, the methods which consider the information of data locality and data time-sensitiveness are demonstrated having better performance in data delivery. Furthermore, network security is another concerned issue in the opportunistic clouds. In this dissertation, we would like to provide users a secure routing path to achieve secure data dissemination. Generally, protecting network security costs a lot. Since different services might demand diverse security requirements, providing proper security protection according to the demanded security requirements is essential. To this end, this dissertation first evaluates the protection capability of security algorithms. Then, an appropriate set of security algorithms can be derived to protect network security of a link. After that, the security levels of all links will be derived. Referring to the security levels of links, the proposed secure dynamic routing protocol can derive a secure routing path for users. Finally, the simulation results show that the methods proposed in the dissertation can be used to securely and efficiently disseminate data in opportunistic clouds.
|
|---|
