Every aspect of modern mobile wireless networks is dynamic. As radios are now attached to moving objects which may make planned, spontaneous, or random movements, the mobility of these objects governs the network state and presents diverse and highly time-varying operating conditions. With increasing density and mobility, the operating regimes of the networks exponentially widen and network connectivity may drastically change over time.
Mobile Opportunistic Networks (MON) is a new paradigm network designed to be able to adapt quickly to such a dynamic regime. It is formed by heterogeneous connectivities (e.g., Wi-Fi, Bluetooth) and opportunistic contacts driven by human mobility. The data delivery is achieved by a method named as ``store,carry and forward", where each node can store and carry the packets until forwarding them over the opportunistic contacts with the next-hop nodes. Compared to the data delivery in the existing networks such as MANET or cellular networks, MON further considers carrying as an additional transmission option.
Such transition in the delivery paradigm can provide huge improvements over the existing networks in terms of capacity, implementation cost and energy efficiency.
The performance of MON depends on to what extent the nodes utilize their opportunistic contacts. There exist three key operations in MON which are directly related to such utilization: First operation is a packet forwarding. Whenever the opportunistic contact occurs, each node should make the forwarding decisions such as selecting the packet to be forwarded or next-hop neighbor nodes. Second operation is sensing other nodes in the vicinity. The effective sensing helps each node fully utilize the randomly encountering contacts without any missing opportunity. Third, a prediction is also essential. By predicting potential contacts, nodes can exploit their future opportunistic contacts in their forwarding decisions.
In this thesis, we study forwarding, sensing and...