Multi-band Wideband Signal Technology

High-resolution imaging has been pursued in radar imaging. Improving bandwidth is the most usual approach to achieve the goal, but this method often brings more difficults to the accomplishment in hardware. This paper focus on the algorithm which can obtain equivalent ultra wide band signal by coherently combining multi-subbands measurements. We implement this algorithm through fitting an all-pole model to the signal. The whole process can be divided into four steps:singular value decomposition, coefficients estimation of the all-pole model by using least-square method, adjust the all-whole model and global all-pole signal model estimation through non-linear least-square iterative.After these four steps, the radar subbands can be mutually cohered and be prepared for the following processing. For the purpose of research, the paper first introduces the concepts of singular value decomposition. and the algorithm of total least square. Then, all-pole model is mentioned. Methods of coefficients estimation in all-pole model is also introuduced. On this basis, we discuss the mutual coherent processing between subbands and we verify the algorithmf by simulating in each corresponding step.In order to verify the effect of improving resolution, three conditions are compared, including single band, multibands without bandwidth extrapolation and multibands with bandwidth extrapolation. Besides this, the paper also studys basic principles and algorithms of ISAR imaging. We combine these basic algorithms with the measurements of radar subbands, using the coherent processing and all-pole model estimation to get high resolution ISAR images. After the comparation between traditionally algorithms and the method proposed in this paper, we get the conclusion that mutual coherent processing and all pole model estimation between radar subbands can improve imaging resolution obviously.