I. Introduction
Non-orthogonal multiple access (NOMA) has been proposed as a candidate technique for future wireless networks [1]. The key idea of NOMA is to allocate multiple users to an orthogonal resource block, e.g., a time slot, a frequency band, or a spreading code, but with different power levels [2], [3]. It has been demonstrated that NOMA outperforms orthogonal multiple access (OMA) from the aspects of spectral efficiency, connection density, and user fairness [4].