Wideband Spectrum Sensing Using Adaptive Neuro Fuzzy Inference System in Cognitive Radio Networks

Radio Frequency (RF) spectrum is anexpensive and limited natural resource for wireless communication systems. In recent times, Cognitive Radio (CR)has come out as one of the most competent candidates for enhancing the spectral exploitation effectiveness. Spectrum sensing is one of the most decisive elements in a CR system facilitating CR to access the licensed spectrum when it is not exploited by Primary Users (PUs).Conventional spectrum sensing approaches such as waveform based sensing algorithm, matched filter algorithm and energy detection algorithm are employed for recognizing the spectrums holes in the band. In actual fact, existing wideband spectrum sensing approaches in a distributed CR network is complicated to recognize, owing to huge implementation/computational complication and huge economic/energy costs. In order to overcome these concerns, a novel spectrum sensing method based on the ANFIS algorithm which is principally exploited to identify the borders of the subband and recognize the spectrum holes in specified input band. ANFIS is employed for effectively sensing the spectrum and considerably reducing the sensing error throughout the process spectrum sensing. The parameters such as power spectral density, bandwidth efficiency, SNR and channel capacity is used for identifying the condition of the spectrum. The experimental results shows that the sensing the spectrum using the proposed method is better than the other techniques.