Group Splitting And Merging Mobility Model For Mobile Ad Hoc Networks And Performance Evaluation

The mobility model is one of the most important factors in the performance evalua-tion of a mobile ad hoc network (MANET). Traditionally, the Reference Point Group Mo-bility (RPGM) model has been used to model the group mobility, where nodes in the same group always stay together throughout the simulation process. Such models fail in model-ling scenarios where groups may be partitioned and merged. These kinds of application scenarios can be found in wireless ad hoc network applications, such as outdoor teaching, scenes of a fire, flood, earthquake, battlefield, conference, seminar sessions, and conven-tional events.In this dissertation, it is aimed to develop an approach to mobility modelling for mo-bile ad hoc networks. The proposed mobility model is called Group Splitting and Merging Mobility (GSMM) model for mobile ad hoc networks and performance evaluation. In this model, group members may sometimes move together and sometimes separately. They have similar mobility patterns and can directly communicate with each other. Indeed, the model presents the trip chain of individuals belonging to a single home. During daily ac-tivities, they move from home to some locations and return back after completing daily tasks. Alternatively they partition and merge at some reference points placed on the trip chain. Therefore, at reference points the group members dynamically re-configure them-selves triggering group partitioning and merging.We study the impact of our mobility model on DSDV, AODV and DSR routing pro-tocols; by simulation.We compare the routing protocols performances in terms of packet delivery ratio, throughput, loss packet, and routing packet overhead. The simulation is written by the network simulator 2 (NS-2) and the graphs are generated using MATLAB. Simulation results show that different routing protocols have different'reactions'to the mobility model.