Quadratic FM Signal Detection and Parameter Estimation Using Coherently Integrated Trilinear Autocorrelation Function

This paper proposes an integrated time-frequency rate distribution (TFRD) technique, i.e., coherently integrated trilinear autocorrelation function (CITAF), for the detection and parameter estimation of quadratic frequency modulated (QFM) signal embedded in white Gaussian noise. The basic idea is to design a trilinear autocorrelation function (TAF) enabling a coherent integration of the signal energy in both of the time and lag-time domains. Theoretical analysis and numerical simulations show that this new method has a lower signal to noise ratio (SNR) threshold and higher estimation accuracy in comparison with several state-of-the-art algorithms. Furthermore, the coherently integrated nonuniform TAF (CINTAF) is briefly introduced.