Semi-Cognitive Radio Networks: A Novel Dynamic Spectrum Sharing Mechanism

In conventional cognitive radio networks, channels that are in use by opportunistic secondary users (SUs) can be recaptured by the network's licensed primary users (PUs) at will, thus interrupting the connectivity of the former. To compensate for this, we propose here a semi-cogntive radio network (SCRN) paradigm where PUs are constrained to first use all free channels in the network before being allowed to capture channels that are currently in use by SUs. By imposing a monetary (or other) penalty to the network's secondary spectrum owners when opportunistic channel use becomes excessive, this additional constraint only induces a slight drop in the PUs' performance while offering significant benefits to the network's SUs. In this paper, we provide a game-theoretic analysis of such systems and we derive both centralized and decentralized adaptive algorithms that allow the system control process to converge to a stable equilibrium state. Our numerical results show that, with the same channel efficiency, SCRNs provide increased profits to the primary network and significantly reduced interruption rates to the secondary network.