TSEGAYE AYELE ABERA2026-03-102022-12-17https://etd.hu.edu.et/handle/123456789/1222Cognitive radio is a robust technology that helps to overcome spectrum underutilization. In recent generations, there has been an exponential growth in bandwidth usage and arising underutilization of radio spectrum resources, due to increasing the extent of radio spectrum demands. However, the spectrum detection and decision threshold techniques are a promised wireless communication technology dedicated to the efficient use of the free spectrum bands. The main challenges in cognitive radio network (CRN) are an inefficient use of licensed spectrum bands due to noise uncertainty, the selection of the most appropriate decision threshold technique based on fading environments. Therefore, investigations on minimizing interferences to obtain accurate information from the desired sensing region of licensed bands at the fusion center during the energy detection process. In this thesis work, performance analysis of decision thresholds on dual-hop CRN’s based on energy detection (ED) to conquer the reviewed problems on kinds of literature. Thus, the comparatives of fixed and adaptive threshold techniques receiver operating characteristic (ROC) curve plots based on the effect of noise uncertainty (NU) parameters and different SNR Environments to obtain outperformed decision thresholds. In addition, evaluating the effect of the number of signal samples size on detection probability during the ED process, the performance analysis of adaptive ED at different SNR environments based on ROC and complementary ROC curve plots. Moreover, in the evaluation of hard decision rules performance for adaptive thresholds of ED at fusion center (FC), the obtained ROC curve plots for ‘’Logic-OR’’ fusion rule performs high false alarming probability with high detection probability than ‘’Logic-AND’’ Rules. Based on discussion results, adaptive decision thresholds performed better than fixed thresholds at low SNR conditions, while the fixed decision thresholds relatively had better performance in high SNR environments. Furthermore, when the sample size increases the detection performance also increases proportionally at perceived SNR ranges. Lastly, the performance evaluations are executed using MATLAB R2018a software and an OFDM signal with Quadrature Phase Shift Keying (QPSK) scheme is advanced based on desired values of false alarming probability (PFA ≤ 0.1) and detection probability (PD ≥ 0.9) at different SNR and number of sample size (N) values.enCRNEnergy DetectionSample SizesDecision ThresholdsFusion RulesNUThesis