A Behavior Pattern Based Mobility Simulation Framework for Office Environments

Modeling movements in office is useful for smart indoor ad hoc networks. People carrying PDA or cell phones can encounter others and in some cases are able to establish connections and transfer data between them. Currently, commonly used mobility models, such as random walk and random waypoint model, do not capture the real movements in real life scenarios, especially in office environments where three typical patterns of heterogeneous behavior, i.e., entity movements, group movements and regular movements, often occur. In this paper we propose a novel mobility simulation framework based on behavior patterns for office environments. The base part is simulation time controller, on which we model the structure of offices and define behavior patterns. In this paper we define three typical patterns of behavior to simulate the heterogeneous movements mentioned above.To simulate more real movements, people can add more patterns of behavior to this framework, which is the main motivation of our framework. We also derive theoretic results of hitting time, which determines the packet delivery delay in delay tolerant networks. Simulation studies show our expressions have error always under 10%. And the staying ratio of our simulation, i.e., the ratio between the time people spend in main place and the total time, is close to the MIT real traces.