Cooperative spectrum sensing optimization based adaptive neuro-fuzzy inference system (ANFIS) in cognitive radio networks

The tremendous growth of the wireless communications and their applications stimulate the urgent need to keep on the available radio spectrum. As a result, cognitive radio (CR) technologies are proposed and developed to manage the limitation of the available spectrum by methods of sensing and sharing the free channels. Wideband spectrum sensing algorithms have a great impact of detecting the vacant channels of the whole spectrum simultaneously. Cooperative sensing techniques are introduced based on sharing users' sensing outcomes among other users. Therefore, it represents an efficient method to overcome signal shadowing and fading problems. Recently, artificial intelligence (AI) techniques are considered to improve the quality of service (QoS) parameters in cognitive radio networks. In this paper, an adaptive Neuro-Fuzzy interference system (ANFIS) algorithm is proposed in the process of decision-making to detect the optimal and accurate free channels. ANFIS model is trained with some pertinent features over a Music-like channel power level (P-MU(k)), channel identity number (k), and channel repetition number. Consequently, the second stage is introduced by applying ANFIS technique on the adaptive blind cooperative wideband spectrum sensing basis to select the optimum required number of cooperative users with increasing performance based on the detected signal to noise ratio (SNR) level per secondary user. Simulation is based on Simulink of five users with different SNR due to fading and shadowing problems. Simulation results proved that, the proposed technique based on cooperative spectrum sensing algorithm with ANFIS model for detection outperformed other traditional detection techniques.