Two-Stage Spectrum Sensing Model for Varying SNR Conditions in Cognitive Radio Network

Cognitive Radio (CR) has come up as one of the efficient technologies to meet the requirement of radio spectrum, which is a very limited resource of present era. CR allows Secondary Users (SU) to access the spectrum dynamically without disrupting the services of Primary Users (PU). In Cognitive Radio Network (CRN) the most important and challenging process is spectrum sensing as the secondary users need efficient sensing techniques to access the spectrum opportunistically. Some spectrum sensing techniques focus on enhancing the detection efficiency; whereas, some others provide better robustness against noise uncertainty and rest reduce the time of sensing as well as complexity. A novel two-stage architecture of spectrum sensing is projected for Cognitive Radio Network in this paper. Two different stages have been described in this model to reduce the complexity and enhance the robustness against noise. After evaluating the Signal to Noise Ratio (SNR) of the incoming signal, a hybrid method associated with Improved Energy Detection and Cyclostationary Feature Detection is established to sense the spectrum in low SNR regime. Digital filtering technique is also used to trim down the effect of noise. This technique emphasizes on the effective performance of the sensing technique during the transmission of weak primary signals. Subsequently, a Binary Hypothesis Test has been formulated to perceive the resultant detection performance. Simulation results reflect that the proposed architecture outperforms existing techniques in varying SNR regime.