Adaptive threshold based combined energy and spectrum-width detection for RF channel sensing in cognitive networks using USRP B200 GNU radios: An experimental study

Spectrum sensing is the process of determining whether a wireless channel is active or idle. It is mainly implemented in dynamic spectrum access in cognitive radio networks which is a potential solution for spectrum scarcity created by static allocation of RF spectrum for exclusive use. In this paper, we investigate adaptive threshold based spectrum sensing using USRP B200 GNU radios. We use adaptive threshold based method for width and energy detection to find active channels which are avoided by the cognitive radio network users. Energy detection based approach may result in high false alarm since it does not consider the width of the signal spectra. Thus to avoid this, we consider both width of the signal spectra and energy level to detect the signal in a given band. Once cognitive radio identifies active channels, it avoids those active channels if it chooses to use them for opportunistic communications and/or reports those idle channels to database for other unlicensed secondary users. We evaluate the performance of the proposed adaptive approach using B200 GNU radios through false alarm and misdetection probabilities. We found that proposed method help to get low false alarm and misdetection probabilities.