Abstract

Mobility is a key characteristic of ad hoc networks. Yet the influence of mobility on interaction pattern evolution has not been analytically investigated for mobile networks. Based on the observation that future interaction patterns are usually dependent on the current interaction status, we model the mobility-induced evolving network structure in mobile sensor networks as a continuous-time Markovian process, which conforms to the continuous position variation due to sensor node movements. Employing spatial coverage theory, we quantitatively characterize how the velocity of sensor movements affects the parameters of the continuous-time Markovian process, and hence characterize how the velocity of sensor movement affects the evolution of interaction patterns among mobile sensors. These results enable us to theoretically predict how the velocity of sensor movements affects the speed of time synchronization, which is crucial for wireless sensor network synchronization algorithms since it directly affects energy-efficiency. QualNet experimental results confirm the theoretical predictions.