High time-resolution spectrum occupancy model for testing of cognitive radio devices

The ubiquitous problem of radio spectrum scarcity caused by ever-increasing traffic, e.g. in mobile communications, demands measures. One approach would be secondary use of spectrum by radio devices with cognitive abilities employing dynamic spectrum access. In order to avoid interference with the primary use of the spectrum, such cognitive devices must be thoroughly tested. In this paper, we present a high time-resolution model of spectrum occupancy, stemming from the activity of primary users, for testing purposes. The proposed model is based on a few statistical parameters and includes a novel parameter adaptation procedure that improves the accuracy of current models significantly. The intended use of the model is over-the-air testing of cognitive devices, for instance for prototyping purposes, e.g. evaluation of the achievable performance, or for conformance testing. The model parameters are extracted from measurement data in the GSM 900 downlink band obtained at the Ilmenau University of Technology campus. However, the modeling approach is not restricted to GSM 900 but lends itself to other bands and radio technologies, too.