Abstract: It is well known that the traditional signal detection methods are based on the Nyquist theorem. However, it will lead to an extremely high sampling rate for ultra-wideband signal detection, making it difficult to implement in practical hardwares. In order to solve this problem, impulse radio UWB (IR-UWB) signal detection problem based on compressive sensing (CS) is studied in this paper. By exploiting the sparsity of the IR-UWB signal in the time domain, we design a CS-based IR-UWB signal detection system and propose an adaptive weighted orthogonal matching pursuit (AWOMP) algorithm.Simulation results show that the proposed algorithm can detect the IR-UWB signal with a sampling rate much lower than the Nyquist rate. Moreover, it can maintain higher performance gains than the existing matching pursuit (MP) based detection algorithm, especially at low signal-to-noise ratio (SNR) egimes.