Spectrum Resource Optimization for NOMA-Based Cognitive Radio in 5G Communications

In traditional cognitive radio (CR), secondary user (SU) can only access the idle spectrum when primary user (PU) is absent, which has to vacate the spectrum when detecting the presence of the PU. Hence, spectrum utilization of the traditional scheme is very low. Recently, nonorthogonal multiple access (NOMA) has been proposed to improve spectrum efficiency of 5G communications. In this paper, NOMA-based CR has been proposed to allow the SU to access multiple subchannels both at the absence and presence of the PU. PU-first-decoding mode (PFDM) and SU-first-decoding mode (SFDM) are proposed at the receiver to decode the NOMA signals, respectively. In the PFDM, the perfect SU throughput can be achieved, but the subchannel power has to be controlled to guarantee the PU throughput. However, in the SFDM, the SU throughput can be decreased due to the interference caused by the PU. Aiming at PFDM and SFDM, we have proposed two optimization problems, respectively, which seek to maximize normalized throughput of the SU by jointly optimizing spectrum resource including number of subchannels and subchannel transmission power. A joint optimization algorithm is proposed to solve the proposed optimization problems. The lower bound of sensing time for energy detection is then achieved to ensure spectrum sensing performance including false alarm probability and detection probability. The simulation results have shown the predominant transmission performance of the NOMA-based CR.