A novel bioradar is developed in our lab to obtain speech signal from human beings. Due to the special attribute of the millimeter wave bioradar, this speech detecting method may provide some exciting possibilities of wide applications. However, the resulting speech is of less intelligible and poor audibility since the present of the combined and colored additive noise. This paper, therefore, investigates the problem of the bioradar speech enhancement by corporates the wavelet packet entropy based voice/unvoiced radar speech adaptive detection method in a wavelet packet adaptation speech enhancement process. The results from both acoustic and listening evaluations suggest that the background noise can be reduced efficiently while the distortion of bioradar speech remains acceptable, suggesting that the proposed algorithm achieved a better performance of noise reduction over other traditional speech enhancement algorithms.