A Modified Double-Threshold Spectrum Sensing Algorithm Based on Adaptive-Threshold Mean Energy Detection

The spectrum scarcity has become an important issue for the new generation of mobile communications. The cognitive radio (CR) approach represents a practical solution for this problem by inserting spectrum-aware terminals or secondary users (SU) that aim to sense efficiently the presence and absence of the licensed users or primary users (PU). Therefore, the spectrum sensing is the paramount SU function for CR systems. In this paper, we modify a double-threshold spectrum sensing algorithm based on mean energy detection (MED) that was introduced recently. In fact, we introduce an adaptive threshold for the MED test, which minimizes the overall decision error probability (DEP). The simulation results show that the novel adaptive spectrum sensing algorithm outperforms the conventional double-threshold algorithm for any value of the duty cycle. For example, a signal-to-noise ratio (SNR) gain of 3 dB was noticed for low duty cycle values (i.e., 30%). However, the proposed algorithm performs even better at high duty cycle and low SNR values.