The design and implementation of intelligent rolling bearings (IRBs) with self-powering, self-sensing, wireless transmission, and self-diagnosis capabilities are crucial for promoting intelligent upgrades in rotating machinery. Traditional sensing schemes rely on external power sources, which hinder wireless sensing. In this study, a wave-shaped-cage-based variable reluctance generator (WC-VRG) is proposed to construct an IRB with the aforementioned capabilities. With bearing rotation, both the crests and troughs of the wave-shaped cage periodically pass through a magnetic circuit composed of the permanent magnet, iron core, and coil. The variable reluctance effect induces a current in the coil, which realizes rotation energy to electric energy conversion in the noncontact mode. The WC-VRG does not require relative movement of the coil and permanent magnet, so it has little effect on the bearing structure. Theoretical research and simulation analysis verify the feasibility of the WC-VRG. Experiments are performed using different vital parameters to evaluate the output performance. A wireless transmission module integrated with the WC-VRG is constructed to transmit the self-sensing signal, which is applied for remote monitoring of the main characteristic frequencies. Additionally, typical faults are classified using the fast Fourier transform and deep learning (DL), with an accuracy of 95.4%. The proposed WC-VRG provides a practical scheme for constructing IRBs.