In application of the side-looking millimeter wave radar to monitor traffic flow,the key issue is to recognize background power spectrum.In this study,the traffic radar principles and echo power properties were analyzed.Then,a novel background power spectrum recognition(BPSR) algorithm was proposed which was based on order statistics and coherent averaging.Finally,test results in various environments were presented.Quantitative evaluation and comparison indicate that the algorithm is feasible and the detection performance has been noticeably improved.