A new super-resolution spectral estimation technique using staggered PRFs

This paper presents a new super-resolution spectral estimation technique based upon nonuniform sampling of time series data. The focus is the application to a radar system employing nonuniform transmission of pulses or staggered pulse repetition frequencies (PRFs). This spectral estimator can detect target doppler shifts an order of magnitude greater than the largest PRF; thus it is capable of providing unambiguous range and doppler coverage simultaneously for a radar system operating on long-range targets. The mathematical realization of this estimator requires relatively less signal processing than other conventional nonlinear methods; e.g., maximum entropy spectral analysis (MESA). In addition, it appears to provide sharper target peaks with very little frequency offset even in a low signal to noise ratio environment. The narrow bandwidth of the peak may become a disadvantage when doppler detection is performed in real-time. This occurs because real time processing involves examining a large number of points on the frequency axis. However, this problem can be alleviated greatly by using an integer search algorithm which exploits the property of nonuniform sampling.