IMPLEMENTATION OF ANFIS ALGORITHM FOR EFFICIENT SPECTRUM HANDS-OFFS IN COGNITIVE RADIO NETWORKS

The advancements in wireless communication in the recent times have raised the demand for wireless radio spectrum. Thus, Cognitive Radio Network (CRN) is introduced as it has the capability to sense the available channels and radio-frequency (RF) spectrum in order to minimize the interference to primary users. Basically, Cognitive Radio (CR) is the network that alters the parameters of transmission and enables more users to communicate concurrently. However, unnecessary handovers in the wireless network lead to the rise of ping pong effect. This effect reduces the efficiency of the network in terms of providing continuous call to the users. The conventional fuzzy model used two fuzzy controllers to implement the entire process. Although, it is efficient but this model is very complex as Fuzzy rules are generated manually. Thus, a novel approach is presented in this dissertation in which amalgamation of fuzzy inference system and neural network is carried out. The main purpose of designing this approach is to reduce the probability of Spectrum Hands-off (HO) which in turn will reduce the ping pong effect. The proposed approach includes two ANFIS controllers and the implementation is performed in terms of different parameters such as Signal Strength from primary to secondary user (SSps), Signal to noise ratio (SNRpu), power of secondary user (Psu), Channel hold time (HT), Velocity of secondary user (Vsu). Simulation is carried out in MATLAB and the comparative analysis is performed which ensured the effectiveness of the proposed ANFIS model. It is observed that the outcome of proposed model surpassed the efficiency of conventional fuzzy logic controller.