Properties of random direction models

摘要: A number of mobility models have been proposed for the purpose either analyzing or simulating movement users in a mobile wireless network. Two more popular are random waypoint and direction models. The model is physically appealing but difficult to understand. Although less physically, it much easier User speeds easily calculated, unlike model, and, as we observe, user positions directions uniformly distributed. contribution this paper establish last property rich class that allow future movements depend on past movements. To end, consider finite oneand two-dimensional spaces. We two variations, with wrap around reflection. simple relationship between these both, show distributed Markov regardless initial position. In addition, sample path both models, namely any piecewise linear applied preserves uniform distribution position provided were initially throughout space equal likelihood being pointed direction.