STUDY AND ANALYSIS OF COGNITIVE RADIO FOR WIRELESS NETWORKS

Abstract

Nowadays, the world is on the cutting edge of the innovative technological revolution in wireless communication networks. The crowding of the device is increasing day by day but the band of spectrum in which devices have to communicate with each other are fixed in nature. So, in the specified band of spectrum, the massive number of device to device connection has to be made in direct or indirect way without interfering with other devices. Cognitive radio is one of the technologies, which is an arrangement of the wireless channel in which transmitter and receiver can logically sense band of spectrum, whether the channels are occupied or not. It can be shift immediately to the available band of the spectrum without interfering the occupied one. By doing so, users can use optimum radio frequency band of spectrum, whereas dropping the interference to primary users. The most important thing to avoid interference is to first sense the particular band of spectrum. Thus, the spectrum sensing is one of the prime function of cognitive radio, which observe the unused band of the spectrum at a given time, space and geographical state. Energy detection is one of the simplest and easiest methods of spectrum sensing due to its simple circuitry and less execution time. The performance of energy detection which is extensively used to accomplish spectrum sensing in cognitive radio over different fading channels. In this study, the average probability of detection and the average AUC curve are derived using the probability density function of the received instantaneous signal to noise ratio. To solve the mathematical complexity, the different mathematical approximation are used as Gaussian-Hermite integration and Holtzman approximation. The performance analysis of energy detection based spectrum sensing in cognitive radio over multipath and multipath/shadowed fading channels like inverse Gaussian, Nakagami m/shadowed, and Weibull/log-normal have been investigated. The energy detection over inverse Gaussian fading channel with selection combining scheme is presented and the average xviii

probability of detection has been formulated with all three realistic environment conditions such as light, moderate and heavy shadowing. In a similar way, the analytical expressions of average probability of detection and average AUC curves for energy detection over composite Nakagami-m/log-normal and Weibull/log-normal fading channels with maximum ratio combining diversity schemes have been studied. The diversity schemes provide better detection of the signal at the receiver’s end in comparison to the one without diversity scheme. Finally, to increase the detection capabilities of the energy detection, threshold should be optimized in comparison to a fixed threshold. So, the optimized threshold can be obtained by minimizing the total probability of error which gives better spectrum sensing even at very low signal to noise ratio. The threshold optimization is applied over inverse Gaussian with selection combining, Nakagami-m/log-normal and Weibull/log-normal with maximum ratio combining diversity schemes.