Research On Time Delay Estimation Of Polynomial Phase Signal

With the fast developing of modern communication industry, wireless positioning technology has been applied to every aspect of human life. Communication service, position tracking and information security etc have become important developing direction, influencing human life more and more.The wireless positioning is based on abstracting characteristic parameters of signals transmitted between base stations and moving stations which makes the estimating of target position possible. AOA, TOA, SOA, TDOA establish their own positioning system, among which, the TDOA has been put more emphasis. In many application area, the signal with continuous phase will always be not stationary, thus the GCC method which is based on stationary signal model won`t be applicative anymore.In modern signal processing, the signal of non-linear, non-stationary property can be approximated with polynomial phase signal in many cases. In this thesis, the work mainly focuses on the TDOA of polynomial phase signal and research on this based on different order of polynomial phase signal. Main work is as follow:Firstly, Research on the cyclic cross correlation method, of which the amplitude and phase information is used for TDOA estimation of LFM signal and compares with the GCC method, proving GCC not applicative for non-stationary signal TODA estimating.Secondly, methods of single-component/multi-component LFM signal TDOA estimating based on Fractional Fourier Transform(FrFT) are researched and its correctness and robustness is proved in low SNR environment through simulation.Thirdly, a method of single-component/multi-component LFM signal TDOA estimating based on Simplified Fractional Fourier Transform(SFFT) is proposed, which will eliminate the pseudo peak interference in FrFT. Compared to conventional GCC method, this method show much better performance and robustness in relatively low SNR.Then, a fast TDOA estimating method based on SFFT is proposed which uses pre-estimating method to minimize the search scale, decreasing the computing amount, which will also achieve competitive estimating performance compared to ordinary SFFT method and far better performance than conventional GCC method in relatively low SNR.At last, a method is proposed where instantaneous cross terms is used to transform cubic phase signal to LFM signal model and uses SFFT to do ordinary/ fast TDOA estimating. Simulation validates this algorithm for cubic phase signal TDOA estimating problem and proved its robustness in various signal environment compared to conventional GCC method.