The Analysis Of Three Key Issues In Ad Hoc Network Based On Probabilistic Model Checking

As an expansion and supplement of Internet and traditional networks, Ad Hoc network has been becoming a hot research field in networks. In Ad Hoc network, mobile node with limited bandwidth moves frequently and topology changes dynamically. Portable mobile node with limited resource, multihop communication and distributed control in the network also make some key issues more complex, such as routing protocols, power management and so on. Therefore, aforemented problems bring new challenges for its research and application.Different from simulation and test, this thesis based on Probabilistic Model Checking studies three key issues in Ad Hoc network, including mobility model, power management and broadcasting algorithm. The main results obtained are as follows:(1)Mobility model dramatically affects the performance of protocols and networks. City Area mobility model is studied and analyzed based on Probabilistic Model Checking. According to the mobility and communication in city street, City Street mobility model is proposed based on Manhattan mobility model and Street Unit mobility model. Then the influence of nodes location and mobility on the neighbor probability and communication between nodes are analyzed with Probabilistic Model Checking tool-PRISM.(2)In Ad Hoc network, mobile node is battery-operated and power-constrained. It is important to prolong the lifetime of battery and optimize energy efficiency. Dynamic Power Management (DPM) model of mobile node is proposed based on energy-hungry components. A stochastic optimum control scheme is applied in this model to improve node energy efficiency. Compared with two familiar time out schemes belonging to predictive schemes, the energy efficiency of this scheme is better as a whole.(3)Broadcasting is a common operation in Ad Hoc network. The efficiency of broadcasting algorithm has significant impact on the performance of the entire network. Based on Dynamic Probability broadcasting (DP) algorithm, Border-Node-first Dynamic Probability broadcasting (BN-DP) algorithm is presented. Nodes located at the radio border of the sender are set with higher rebroadcast probability. Analyses reveal that with the same reachability, BN-DP algorithm saves more rebroadcasts, overheads and operates more efficiently than FP (Fixed Probability) algorithm and DP algorithm do.