On the Modeling and Performance of Three Opportunistic Spectrum Access Schemes

Abstract

We model and study the amount of supported secondary system traffic for three opportunistic spectrum access schemes. We assume that the spectrum band is partitioned into <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</i> frequency bins, and whenever all are occupied, spectrum access by secondary radio (SR) will be blocked. On the other hand, an SR transmission is forced to terminate whenever a primary radio (PR) transmission is detected in the same frequency bin. The grade of service (GoS) of the secondary system is quantified by the SR service dropped and SR service blocked probabilities. We model each scheme using a continuous-time Markov chain and derive the expressions to compute the two GoS performance metrics. In the first scheme, the PRs randomly access the frequency bins with no knowledge about the presence of SRs. Therefore, the existing SR transmission has to drop whenever a PR chooses to transmit in the same frequency bin, although there are other unoccupied frequency bins. For the second scheme, the newly arrived PR will avoid colliding with existing SRs by occupying the vacant frequency bin. In the third scheme, the newly arrived SR is limited to access <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</i> - <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">r</i> frequency bins only, where <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">r</i> can flexibly be selected to optimize the amount of supported SR traffic. This paper suggests that more SR traffic can be supported when spectrum access is coordinated between primary and secondary systems. We finally compare the effective spectrum utilization. The simulated results verify the correctness of the derived expressions.