Resumen de Nuevo método de detección espectral para sistemas multiportadoras basado en la radio cognitiva
Nowadays, the growing demand for frequency-free bands for new emerging wireless technologies and fixed frequency allocation policies in the use of the radio spectrum have resulted in a saturation of the radio frequency spectrum.
From this point of view, it could be thought that the radio spectrum is fully occupied and there are no more frequency bands for new wireless communication systems. However, spectrum occupancy measurements reveal that the licensed radio spectrum has a very low use in certain frequency bands. Cognitive Radio (RC) is a technology that has been proposed as a possible solution to improve access to the radio spectrum and increase its efficiency of use.
RC technology has mainly three functions: detection, analysis and spectrum decision. In order for RC technology to succeed, it must perform the detection of the primary user's signals in a reliable way even if these signals are in a low Signal to Noise Ratio (SNR) scenario and ensure that they do not cause interference to primary users.
This doctoral thesis presents the design of a new radio spectrum detection technique, called Candidate Spectral Estimation (CASE), whose objective is to present a possible solution to the need for a reliable and rapid detection of the signals of primary user for systems based on Cognitive Radio. This proposed detector is based on the filter bank scheme for spectral detection and is included as the fourth alternative in spectral detection techniques within the transmitter detection category. The CASE detector is able to detect and identify primary users with specific spectral characteristics.
Additionally, this thesis work provides a study that shows the potential of this new spectral estimation technique in RC scenarios. As examples of possible applications, we consider the detector (CASE) for the detection of Digital Television signals, (DVB-T) and multi-carrier signals for filter banks. The results of numerical simulations, performed in MATLAB, are presented to evaluate the behavior of the proposed detector. The simulations were carried out under different scenarios using mainly signal levels with high and low SNR. To demonstrate the potential of the CASE detector in estimating the power level and the frequency location of the primary user's signal in the presence of an interfering signal and white noise, a Scenario with high SNR was used.
Finally, a comparison is presented between the spectral detection techniques such as the energy detector and cycle-stationary characteristics to evaluate the performance of this new detection technique specially designed for application in Cognitive Radio. We present, 1) the robustness of the CASE detector to interfering signals and white noise in scenarios with high SNR, 2) the ability of the CASE detector to identify the primary user on stage with low SNR levels and interfering signals
