OFDM radar waveform optimization for low grazing angle scenarios using mismatched filter via convex solution
摘要
In this paper, a novel approach to design a mismatched filter orthogonal frequency division multiplexing waveform, with the design criteria of peak sidelobe level is presented, which is used in a low grazing angle scenario. Our objective is to enhance the monostatic radar resolution, and improve tracking accuracy and angle measurement accuracy in low grazing angle scenarios. Additionally, a methodology to optimize the performance of system in the presence of multipath interference and the autocorrelation function of the target dispersion coefficients is proposed. The aforementioned problem is formulated in a convex problem, yielding an optimal response. The problem is initially formulated as the minimization of the peak sidelobe level of the mismatched filter’s output during the waveform design. It is then transformed into a second-order cone programming problem, which is a convex problem with a globally optimal solution. This problem allows the incorporation of convex constraints, such as the constraint of loss in processing gain and the constraint of Doppler effects. Accordingly, the loss in processing gain constraint, which has practical significance, has been reformulated as a convex constraint and included in the problem formulation. Furthermore, to mitigate the adverse impact of targets’ Doppler on the output of the mismatched filter, we express these effects as convex constraints and incorporate them in the optimization problem. Simulation results demonstrate the effectiveness of our proposed strategy for designing mismatched filter’s with optimal peak sidelobe level. Furthermore, the proposed generalized chirp-like-orthogonal frequency division multiplexing waveform demonstrates reduced peak-to-average power ratio, enhanced channel estimation accuracy, lower sidelobe levels, improved detection performance under noisy and occluded low grazing angle conditions. These results verify that the proposed design offers a computationally efficient and practically viable solution for real-time radar systems operating in dynamic and challenging environments.