Análise de algoritmos de sensoriamento espectral para canais com desvanecimento generalizado.

Abstract

Currently, allocation of services to frequency bands is done in an static way and gene- rally restricts the use of each band to a single service. This causes the transmission capacity of the frequency band to be inefficiently used when the service allocated to it is idle. It is necessary to develop more efficient strategies to use the electromagnetic spectrum to increase the trans- mission rate. Cognitive Radio is proposed as part of the solution. A cognitive radio is a device that senses the spectrum in search of gaps in the transmission from users which hold priority in the use of certain frequency bands. These blanks are then used by the cognitive radio for its own transmissions. The environment in which the device is inserted imposes fading effects on the transmission, that must be considered when employing cognitive radio. The most cited spectrum sensing technique in the literature is the energy detector, due to its broad application and simple implementation. In this dissertation, the performance of the energy detector, while subject to several different fading models, is analyzed. The effects of relatively simple fading models are studied, in particular the Rayleigh, Rice and Nakagami-m models, followed by the generalized fading models κ − μ, η − μ and α − μ. The performance of the detector is exa- mined by observing the Probability of Detection (PD) for several levels of the Signal-to-Noise Ratio (SNR), while varying each parameter of the models to determine its influence. A determi- nant influence is found of the parameter m (related to the variance of the channel gain) over the energy detector probability of detection. This information can be applied to the performance es- timation of the energy detector in a certain environment, given that the designer has knowledge of the physical characteristics of the location where the cognitive radio device operates.