Bluetooth is a widespread solution for personal area networks (PAN). However, it currently lacks in spontaneous and energy
efficient device communication which could make it more successful. To allow spontaneous networking, communication nodes should
have the ability to timely discover each other. Moreover, in order to improve energy efficiency and increase battery lifetime
the energy spent during discovery procedures should be minimized while guaranteeing high probability of successfully completing
the discovery and the information transfer procedure. In this paper we derive performance considerations on the behavior of
the system through experiments in a physical testbed where devices move while discovering nodes in their neighborhood and
then exchange information. These considerations can be used for accurately model neighbor discovery process and information
transfer in Bluetooth networks.
Keywords Bluetooth - Markov model - Analysis - Experimental Results