부산대학교 도서관

Multiple-input–multiple-output (MIMO) sparse electromagnetic vector sensor (EMVS) arrays have brought new perspectives to signal processing due to their flexibility and higher resolution. In this study, we focus on angle estimation in a monostatic MIMO system using arbitrary geometry EMVS arrays. An improved parallel factor (PARAFAC)-based algorithm is introduced. By harnessing the natural multidimensional structure of the array output, we rearrange it into a PARAFAC model. The factor matrices are obtained using the complex parallel factor analysis (COMFAC), followed by applying vector cross-product (VCP)/phase compensation for rough/refined estimation. This approach achieves superresolution estimation while automatically pairing angles and showing low computational complexity. As a result, it outperforms existing algorithms, and numerical simulation experiments validate the improvements achieved by the proposed method.