Demo: recommendation system for dynamic spectrum access through spectrum mining and prediction

Abstract

Inspired by the spectrum database in the white spaces, we propose a recommendation whose function is similar to that of the spectrum database. We aim at predicting the spectrum availabilities and recommending the part that is more likely to be available for the usage of SUs. The prediction on the spectrum availabilities is through running the spectrum mining algorithm on the collected spectrum data. Therefore, the two main components for our demo are the data collection and spectrum mining. After the mining process is finished, our system would give the predictions and recommendations on the spectrum that is more likely to be available.