Predictive Dynamic Spectrum Access

This paper explores the idea of predictive dynamic spectrum access (PDSA). Modern spectrum resource allocation research typically divides users into two classes: primary users and secondary users. Primary users own licenses to particular frequency bands and are allowed to use it whenever they wish. Secondary users can reuse the frequencies when they are not being used by a primary user. The goal of PDSA is to gather statistical information about a primary user in an effort to predict when the channel will be idle. This allows us to better plan secondary use of the spectrum without the cooperation of the primary user. We explore two approaches to PDSA in this paper. The first uses cyclostationary detection on the primary users’ channel access pattern to determine expected channel idle times. These techniques are simulated with both TDMA and CSMA networks. The second briefly examines the use of Hidden Markov Models (HMMs) for use in PDSA.