A novel method of multi-band spectrum sensing exploiting dynamic compressive sensing

摘要

Multi-band spectrum sensing is now facing challenges of high sampling rate and random access of primary users. However, these problems can be solved in terms of dynamic compressive sensing (DCS). In this paper, we propose a novel method of multi-band spectrum sensing using DCS, which can sample broadband signals at sub-Nyquist rate. The proposed method is explicitly intended for sparse streaming signal in frequency domain with time-varying support. The dynamic sparse signal is first modeled. And then, the measurement matrix is constructed based on Analog-to-Information Converter (AIC) to ensure down-sampling of dynamic sparse signals. Finally, based-Kalman filter reconstruction algorithm is employed to directly reconstruct the frequency domain information of a dynamic sparse signal. The simulation results demonstrate that the proposed method performs well.