A Fast 3-D Near-Field Imaging Method Using MMW MIMO Antenna Array System

Abstract

Large-scale millimeter-wave (MMW) multiple-input multiple-output (MIMO) arrays provide high-resolution near-field 3-D imaging but suffer from severe computational burdens due to the massive volume of raw echo data. To address this critical bottleneck, this article proposes a highly efficient two-stage imaging framework. In the first stage, an equivalent phase center (EPC)-based multistatic-to-monostatic conversion (MMC) compressively reduces the data dimensionality, enabling a 3-D range migration algorithm (RMA) to rapidly generate a coarse global image. In the second stage, the scene is dynamically partitioned, and predefined thresholds are applied to extract target blocks. Subsequently, a lightweight 2-D holographic slice reconstruction (HSR) algorithm is executed exclusively on these localized blocks to achieve refined high-resolution imaging. Furthermore, the impact of transceiver separation on imaging quality is analytically evaluated, and the optimal block size is derived to maximize the well-focused region while minimizing computational overhead. Validated through simulations and a custom-integrated $512 ~\times ~512$ MMW MIMO hardware prototype, the proposed framework achieves comparable imaging quality to conventional algorithms yet delivers orders-of-magnitude reduction in execution time, thereby offering a highly practical solution for real-time near-field MMW sensing applications.