Service Discovery and Device Identification in Cognitive Radio Networks

Cognitive radios (CR) will be able to communicate adaptively in an effort to optimize spectral efficiency. An integral step towards this goal involves obtaining a representative view of the various services operating in a local area. Although it is possible to load different software modules to identify each potential service, such an approach is needlessly inefficient. Instead, rather than use a collection of complete protocols on a CR, we believe that it is essential to have a separate identification module that is capable of reliably identifying services and devices while minimizing the code needed. In particular, by effectively leveraging protocol-specific properties, we show that it is possible to utilize data from narrowband spectral sampling in order to identify broader band services and individual devices. We demonstrate the feasibility of such service and device identification using GNU radio and the universal software radio peripheral (USRP) platform by identifying radio services in the industrial, scientific, and medical (ISM) radio band. Further, we show that physical layer signatures may be used to reliably identify devices, thereby allowing CRs to exploit physical layer information in support of basic authentication functionality.