Recent researches on various parts of allocated spectrum have demonstratedthe inefficiency of spectrum utilisiation. On the other hand, according to the noticeable growth of wireless communication systems, the demand for bandwidth has increased. As a result, the static frequency assignment approach needs to be reconsidered. Cognitive radio offers a proper solution for efficient spectrum usage. Detection of the idle frequency bands is the first step for exploiting the spectrum.Thus, one of the main components of the cognitive radio iectrum sensing.In this thesis,after presenting a brief introduction about cognitive radio, different spectrum sensing techniques and their applications are introduced. Afterwards, considering the importance of selecting the appropriate physical layer for cognitive radio, we have considered two physical layer candidates. In order to reduce hardware complexity for spectrum sensing task, we have compared the performance of spectrum sensing in two physical layer candidates based on orthogonal frequency division multiplexing (OFDM) and filter bank multicarrier (FBMC). Performance of the detectors is investigated under two different scenarios. For this comparison the area under receiver operating characteristic (ROC) curve namely AUC is considered as a criterion. In the first scenario, it is assumed that the secondary user which does not perform the spectrum sensing is not permitted to transmit any signal (quiet sensing).In this case, it is shown that the performance of both detectors is the same. In the second scenario, the secondry user is permitted to transmit in a frequency band adjacent to the primary user (PU) channel while spectrum sensing is performed (non-quiet sensing). The results indicate that although DFT-based detector is computationally less complex than filter bank-based detector, filter-bank based detector's performance is much better than DFT-based detector in heavy interference environment. Consequently, there is a trade off between the performance of the detector and its complexity when choosing one of the above mentioned physical layer candidates. Keywords: Cognitive radio, spectrum sensing, OFDM, FBMC, quiet sensing, non-quiet sensing
