The Key Technology Research On Optimizing Delivery Ratio In DTNs

In the recent years, as a new kind of challenging network, delay tolerant networks have received extensive attention and research. In DTNs, end-to-end connection from source to destination is not guaranteed due to the node’s uncertain mobility, the easy-interrupt connections, and the high transmission delay. As a result, many researchers suggest that, the messages are routed in a store-carry-forward paradigm, in order to improve the delivery ratio. However, how to decide which node to forward is the first question to be solved, which needs a routing protocol. Secondly, a reasonable message scheduling and dropping strategy is also needed. And the third question is how to control energy consumption, which is used in beaconing and forwarding messages.Thus, this thesis focuses on maximizing the message delivery ratio in DTNs, and proposes a serious of novel models and methods from three perspectives: how to design the routing protocol, how to establish the buffer management scheme, and how to propose the efficient energy control strategy. The main contributions of this thesis are shown as follows:(1) Routing protocol. This thesis proposes a clustering routing method based on semi-markov model through analyzing and comparing with the existing routing methods, which first clusters nodes according to path similarity, then predicts nodes’ contact probabilities at the specific time by applying semi-markov model and determines a collection of nodes containing the source and the destination, which are used in the routing table. Finally, dynamic routing tables related to the current time are obtained by executing path-finding algorithm.(2) Buffer management. First of all, a buffer management strategy based on the game of life is proposed for delay tolerant networks in this thesis. This strategy is designed for Epidemic routing protocol and determines the specific operations of a message stored in the node’s local buffer according to the proportion of the message holders in all the node’s neighbors. Finally, the messages are sorted and dropped according to the utilities. Moreover, in order to cope with the different message sizes, a knapsack-based message scheduling and dropping strategy is proposed based on Epidemic routing protocol. To improve the delivery ratio, this strategy sorts the messages by utility values and decides which message to drop based on the knapsack problem in the case of overflowing.(3) Energy control. In this thesis, in order to maximize delivery ratio, we propose a dynamic beaconing control strategy for Epidemic routing in energy-constrained DTN based on time-continuous markov model. Then, aiming to find a practical energy control strategy, it is proposed that the Wi Fi hotspot mode of a mobile phone is applied in the PSN, in order to realize the efficient message transmission. To maximize the message dissemination scope within the limited energy constraint of each phone, an energy efficient phone-to-phone communication method based on Wi Fi hotspot is presented to schedule the phone’s switch between hotspot mode and client mode.In summary, this thesis not only proposes a series of models and methods on routing protocol, buffer management and energy control, but also provides the theoretical analysis and extensive simulations to verify the correctness of them, which provides necessary theoretical and technical support for the subsequent applications of DTNs.