Achieving adaptive compressive spectrum sensing for cognitive radio

The spectrum is sparse in the real world, and it has enormous advantages when applying compressed sensing (CS) technology to wideband spectrum sensing. However, the sparsity is often unknown in practice, so a large number of measurements have to be chosen, which will lead to the performance degradation. To solve this problem, an adaptive compressive spectrum sensing method is proposed. The coarse sparsity estimation can be obtained by analyzing the relationship between the second derivative of compressive measurements and sparsity. Then by increasing the number of measurements and continuing iterations step by step in both training subset and test subset, the accurate sparsity estimation can be obtained while the halting criterion can be met. Simulation shows that the performance of our method is better than other traditional CS methods, which is very important for decreasing the complexity and memory. Moreover, the effectiveness of the proposed method is also verified in noise environment. © 2020, Editorial Office of Systems Engineering and Electronics. All right reserved.