Adaptive measurement adjustment for sparse streaming signal

In traditional Compressive Sensing (CS) algorithms, the sparsity of a signal is used to determine a proper number of compressive measurements M according to empiricism. However, the sparsity is impossible to be known ahead of time. In order to ensure the reconstruction accuracy of the signal, M is set to be as large as possible. In this paper, we propose an algorithm, termed Adaptive Measurement Adjustment (AMA) which is used to adaptively adjust the number of compressive measurements until it reaches a proper value. Unlike the existing similar algorithm, AMA develops a specific criteria for the signal judgment, and it can change corresponding parameters to adapt different categories of sparse signals, that makes AMA more reliable and flexible. Furthermore, AMA uses a modified bisection method, which has guarantee of convergence speed. Experiment results show that AMA is superior to the similar algorithm with appropriate conditions.