Multi-Modal Cooperative Spectrum Sensing Based on Dempster-Shafer Fusion in 5G-Based Cognitive Radio

In 5G-based cognitive radio, the primary user signal is more active due to the broad frequency band. The traditional cooperative spectrum sensing only detects one characteristic of PU using one kind of detector, which may decrease the sensing performance when the wideband PU is in severe fading channel. In this paper, a multi-modal cooperative spectrum sensing is proposed to make an accurate decision through combining multi-modal sensing data of the PU signal, such as energy, power spectrum, and signal waveform. Each secondary user (SU) deploys multiple kinds of detectors, such as energy detector, spectral detector and waveform detector. The multi-modal sensing data from different detectors are sent to a fusion center. In the fusion center, the local decision is achieved through the Bayesian fusion, while the global decision is determined by the DS fusion. The sensing credibility of each detector can be fully considered in the DS fusion, in order to avoid the performance difference of different detectors. Weight DS fusion is also proposed to improve the decision performance through decreasing the sensing impact of malicious SU while increasing the fusion proportion of dominant SU. The simulation results have shown that the proposed multi-modal cooperative spectrum sensing can achieve better sensing performance in fading channel.