A fast gradient-based sensing matrix optimization approach for compressive sensing

Sensing matrix design is among the essential keys for compressive sensing to efficiently reconstruct sparse signals. It has been demonstrated that sensing matrices, with improved mutual coherence property, have good performance. In this paper, we propose a fast approach to sensing matrix optimization based on fast gradient method. Simulation results show that our approach provides good performance compared to conventional methods. Moreover, it provides a significant gain in terms of computing time.