This chapter presents a near-field source localization method based on subarray partitioning and Vandermonde coupled canonical polyadic decomposition (VC-CPD). We consider large-scale antenna arrays, where the source is more likely to be in the near-field region. However, as the number of antennas increases, the computational complexity also rises. To address this, we partition the array into multiple subarrays, with or without overlap, ensuring that the subarray apertures satisfy the far-field assumptions. This partitioning step effectively reduces the number of antennas in each subarray, which in turn alleviates the computational burden. Additionally, the coupling of the subarrays’ received data is then decomposed within the upscaled tensor, enabling target localization based on direction-of-arrival (DOA) estimation and spatial-geometrical relationships among the subarrays. As a result of reducing the number of subarray antennas, we achieve a significant reduction in computational complexity compared to traditional methods like MUSIC and ESPRIT algorithms. Furthermore, by employing Vandermonde structure-based coupled decomposition of the subarrays, the proposed method improves localization accuracy over the CPD algorithm. Experiments are conducted to compare the performance of various algorithms.