The IR-UWB Received Signal Reconstruction Based on Quantized Compressed Sensing

Compressed Sensing (CS) theory provides a new solution for low-rate sampling design of Impulse Radio Ultra-WideBand (IR-UWB) receiver, but the quantization process is usually idealized in existent CS based sampling architectures. In this paper, the influence of quantization noise is fully considered, and an IR-UWB signal reconstruction method with high anti-noise performance is proposed. Based on the analysis of the receiver noise distribution characteristics, the signal reconstruction optimization model is revised, and then the performance of Dantzig-Selector (DS) method is compared with the traditional signal reconstruction algorithms. Further, a joint DS-SP method which can self-adaptively select the reconstruction algorithms between DS and SP (Subspace Pursuit) is proposed. Simulation results show that the joint DS-SP method which has computational complexity trade-off between DS and SP can get the best performance under different noise regions and quantization precisions. Whats more, joint DS-SP has large performance improvement compared to the traditional reconstruction algorithms, thus provides a new strategy of CS signal reconstruction for the design of IR-UWB receivers digital back-end.