Research On Range Estimation Algorithms With High Resolution By Utilized The Technique Of Virtual Frequency Extension

Recently, MIMO radar has attracted a large number of scholars’ attention due to its superior performance in target detection as compared with traditional phased array radar. It has been widely used in wireless communication, national defense, biomedical engineering, seismic surveying and so on. As an important task of MIMO radar system, the range estimation detects the distance between the target and MIMO radar by utilizing the reflected signal from the target. The existing algorithms of the range estimation include the matched filtering method, the CS(compressive sensing) based method and the SF(step frequency) based method. For the matched filtering method, there are much data required to guarantee the range resolution. For the CS based method, it can reduce the amount of sampling data to some extent, but the measurement resolution is limited if there are only few number of pulses utilized. For the SF based method, it occupies wider band, which puts forward new demand for spectrum resources, and the range resolution is also limited by the number of pulses utilized.In order to solve the problem mentioned above, this thesis proposed a new range estimation algorithm based on the technique of virtual frequency extension. The proposed algorithm significantly improves the range resolution by utilized fewer transmission pulses as compared with traditional methods, which saves the transmission power and reduces the signal bandwidth required. However, unfortunately, the cross terms are introduced inevitably by the proposed method, which degrades its performance obviously. The impact of the cross terms on the performance is mainly determined by the autocorrelation property of the transmit waveforms. In order to suppress the cross terms, P4 sequence which has better autocorrelation property was employed in the proposed algorithm and the final performance is further improved, which is very close to the optimal performance produced by the real frequency points.As a result of the shortage of spectrum resource, the frequency interference becomes more and more serious, which degrades the performance of traditional range estimation methods in MIMO radar. This thesis proposed an adaptive range estimation algorithm with avoiding the frequency interference based on the technique of frequency extension mentioned above. The main idea of the proposed adaptive algorithm is that, with the help of periodic update mechanism at the transmitter, a clean frequency will be selected from the predefined frequency band and the signal without frequency interference will be received by utilized the autocorrelation property. Then the technique of frequency extension is utilized to improve the final performance and suppress the bad impact of the frequency interference.