Bluetooth automatic network recognition - the AIR-AWARE approach

Automatic network recognition and classification may prove to be an important concept in the framework of cognitive radio and networks. For practical implementations, these operations must be carried out in a simple way by using simple devices and algorithms that require low computational load. The AIR-AWARE approach proposes to use MAC sub-layer features for technology recognition purposes where a rudimentary device like an energy detector is used for technology-specific feature extraction. The aim of this work is automatic Bluetooth classification. To this purpose, two MAC features reflecting properties, related to the time-varying pattern of MAC packet exchanges, are proposed. Experimental data obtained by using the Universal Software Radio Peripheral as energy detector show that the two proposed features are capable of highlighting MAC sub-layer behaviour peculiar to Bluetooth. These features may therefore lead to successful Bluetooth recognition and the results obtained provide support to the validity of the AIR-AWARE approach.