Sensing management scheme analysis for cognitive radio networks

Abstract

To support various applications with high demand through wireless communications, most of the radio frequency spectrum is allocated exclusively, but out of which much is known to be underutilized. Recent measurements by the Federal Communications Commission (FCC) show that 70% of the allocated frequency spectrum in the U.S. is not utilized. Furthermore, the spectrum occupancy reveals temporal and spatial variations. In order for spectrum efficiency, the service of unlicensed secondary users (SUs) can be utilized opportunistically when the primary users (PUs) do not occupy the band. The main constraint of the opportunistic SUs is that their operation should not impose interference to PUs. Hence, whenever a spectrum band is used by PUs, SUs have to vacate the band within a certain amount of time. To vacate the spectrum band in real-time, the cognitive radio (CR) network may spend much time on detection for PUs. Due to the fact that the SUs do not utilize the spectrum band during the detection, it can degrade the SUs`` own throughput. Thus, spectrum sensing, the process of periodically and dynamically monitoring a given spectrum band, is not only a mandatory function in any CR-based wireless system, but also requires an efficient coordination. Throughout the thesis, we propose an efficient sensing management scheme using the energy detector with adjustable sensing interval. While the false alarm probability causes SU``s throughput degradation, the miss-detection probability yields interfering duration on PUs. With these probabilities related to sensing duration, the sensing interval (i.e., time between adjacent sensing durations) and the PU``s arrival are analyzed. We can examine the sensing management scheme involved in a compromise between the SU``s perspective and the PU``s perspective. Consequently, we formulate throughput maximization and interfering duration minimization problems from SU``s and PU``s perspective respectively, and propose an analytic mode...