Auto-regressive spectral gap filling algorithms for photonics-based highly sparse coherent multi-band radars in complex scenarios

Auto-regressive gap filling algorithms have been demonstrated for interpolation among spectral gaps created by incoherent dual band radar systems. The required synchronized reconstruction leads to a heavy computational load. The recent use of photonics in microwaves systems, has resulted in coherent multiband radars that allow implementing these auto-regressive algorithms without the need to recover the coherence, and consequently with a reduced computational complexity. In a previous work, a modified auto-regressive algorithm with optimization control for coherent multiband radar has been presented. The algorithm has shown the ability to interpolate among two coherent sub-bands, obtaining the same performance of equivalent radar with a total band equal to the sum of the used sparse sub-bands. Therefore, the application of the modified auto-regressive algorithm to radar with sparse sub-bands overcomes the bandwidth limitations of RF front-ends and the frequency regulation restrictions. In the previous work, the algorithm was verified in simple scenarios and using only two sub-bands. Here, the algorithm using two sub-bands is applied and verified to scenarios with increased complexity, including the case of a boat as target. Moreover, for the first time we extend the algorithm to the case of several sub-bands (more than two) and we analyze the performance. We demonstrate that the benefit of using the auto-regressive gap filling algorithm instead of data fusion, increases as the number of sub-bands gets higher.