Artificial intelligence-assisted indoor signal propagation and modem placement optimization

Abstract Accurate modeling of structural obstacles affecting electromagnetic signal propagation in indoor spaces is important in planning wireless communication systems. This study proposes a hybrid and data-driven method that integrates 3D point cloud data with wireless signal strength measurements (RSSI) to model the effect of structural elements on electromagnetic signal propagation in indoor spaces. Unlike traditional approaches, the proposed method can estimate the location, geometry, and signal attenuation effect of walls without requiring any prior knowledge of building materials. In the first stage, surface meshing is performed on the point cloud data, and then density-based wall segmentation is performed using the DBSCAN algorithm. Each wall cluster obtained is matched with the RSSI data measured on-site, signal deviations are analyzed, and the WAI is calculated for each wall. These indices provide an indirect estimation of physical parameters that cannot be measured directly, such as electromagnetic permeability and are used as decision-support tools for the optimum positioning of access points in indoor spaces. Experimental results show that the proposed approach provides significant increases in signal strength and that access points can be determined without requiring knowledge of the building materials of wireless communication systems. Experimental results show that the proposed method has a generalizable structure for smart building applications and indoor digital twin systems.