51 results about "Keystone transform" patented technology

The invention discloses an improved fractional Fourier transform maneuver weak target detection method. The present invention first establishes the uniform acceleration motion signal model of the high-speed maneuvering weak target; then performs Fourier transform on the original echo data in the fast time domain, and after the transformed data undergoes Keystone transformation, the problem of distance unit walking caused by the primary phase is corrected, Then IFFT is performed in the fast time domain, and the FRFT of all orders in p[0,2] is calculated for each range unit in the final echo data obtained to form the signal energy in the fractional order p and fractional order transformation domain u constitutes the two-dimensional distribution of the two-dimensional parameter plane (p, u), and the two-dimensional search of the peak point on this plane can realize the detection of accelerated targets. The invention can realize the detection of the weak maneuvering target in the background environment of low signal-to-noise ratio, and has good reliability, feasibility and real-time performance at the same time.

The invention discloses a broadband DS / FH (Direct Spreading) / FH (Frequency Hopping) signal parameter estimation method and system based on Keystone transform, and aims to provide a high-precision andlow-complexity broadband DS / FH signal parameter fast estimation method. The technical scheme is as follows: using a differential correlation-based DS / FH signal estimation method to obtain a delay time estimation value of a received signal, dispreading the received signal to obtain a dispreading signal by referring to the delay time estimation value; judging the frequency hopping and a symbol turnover position of the received signal according to the delay time estimation value, performing segmented processing on the dispreading signal, and forming a segmented signal matrix by segmented signals; respectively performing Keystone transform on each row of the segmented signal matrix, and respectively carrying out Fourier transform on each row of the signal matrix after the Keystone transform to obtain a frequency domain matrix, taking module square for the frequency domain, and performing non-coherent accumulation according to columns to obtain accumulated vectors; and finding the frequency fm corresponding to a maximum value from the accumulated vectors, and carrying out speed estimation and completing parameter estimation.

The invention provides a coherent accumulation detection method applied to a third-order maneuvering target, which belongs to the technical field of radar signals. The invention firstly performs pulse compression in the fast time frequency domain on the echo signal, and then uses the third-order Keystone transform to correct the third-order distance migration of the target. The frequency-shifted cross-correlation and scaled Fourier transform are then used to derive a value that defines the velocity. The first-order distance walking is eliminated by using the fast-time frequency-domain matched filter equation constructed by the defined velocity, and the acceleration and jerk of the target are jointly searched by using the two-dimensional matching filter equation in the fast-time-frequency domain, and the second-order distance walking and multi-dimensional walking of the target are compensated and eliminated. Puller diffusion effect. Finally, the coherent accumulation detection of target energy is carried out by slow-time fast Fourier transform. The invention simultaneously utilizes the amplitude and phase information of the target echo to achieve coherent accumulation, eliminates the range migration and Doppler migration effects of the third-order maneuvering target, can significantly improve the detection performance of the radar, and has strong practicability.