Abstract-This paper investigates double-threshold based energy detector for cooperative spectrum sensing mechanisms in cognitive wireless radio networks. We first propose a hybrid double-threshold based energy detector (HDTED) to improve the sensing performance at secondary users (SUs) by exploiting both the local binary/energy decisions and global binary decisions feedback from the fusion centre (FC). Significantly, we derive closed-form expressions and bounds for the probabilities of missed detection and false alarm considering a practical scenario where all channel links suffer from Rayleigh fading and background noise. The derived expressions not only show the improved performance achieved with the HDTED scheme but also enable us to analyse the impacts of the number of the SUs and the fading channels on the cooperative spectrum sensing performance. Furthermore, based on the derived bounds, we propose an optimal SU selection algorithm for forwarding the local decisions to the FC, which helps reduce the number of forwarding bits for a lower-complexity signaling. Finally, numerical results are provided to demonstrate the validity of the analytical findings.