In this paper, a new Rake receiver based on Compressed Sensing (CS) was proposed for impulse radio Ultra-Wideband (IR-UWB) wireless communication systems. Considering of the sparsity of UWB channel, CS reconstruction model was transformed from channel estimation model. Sparse vectors are reconstructed by solving an optimization problem, and this completed the channel estimation. According to the analysis of simulation results compared with the receiver based on LS algorithm, the proposed receiver can take full advantage of the sparsity of UWB channel, so it provides better performance in tracking time-varying fading channels and also helps obtain lower Bit Error Rate (BER) for IR-UWB systems.