DT-ECO: Digital Twin-Empowered Coverage Optimization for Indoor Wireless Networks

Wireless coverage maximization in indoor environments, such as underground mines and industrial sites, is a non-trivial task due to the complex propagation conditions caused by blockages, scattering, diverse materials, and nonuniform RF signal attenuation. Standard statistical models often fail to capture these inherent complexities, frequently overestimating or underestimating RF attenuation. Digital twin (DT) technology offers a promising alternative by accurately replicating RF propagation in real-world environments through its awareness of geometries, material properties, and positions of 3D scattering objects. This paper proposes DT-ECO (Digital Twin-Empowered Coverage Optimization), a site-specific and context-aware framework to enhance indoor wireless coverage. DT-ECO operates in two main stages: (i) high-fidelity virtual RF environment modeling and (ii) access point (AP) placement optimization. The digital twin is constructed using open-source tools, namely LiDAR for capturing 3D environment, Blender and Mitsuba for creating geometrically accurate 3D models, and NVIDIA Sionna™ for computationally efficient, high-fidelity ray tracing. Leveraging this virtual RF environment, DT-ECO formulates and solves an optimization problem to determine optimal AP placements that maximize coverage. Owing to the problem's computational complexity, a particle swarm optimization (PSO) algorithm guided by DT-generated coverage maps is employed. Extensive simulations demonstrate that DT-ECO significantly outperforms conventional RF propagation models in terms of signal strength distribution and coverage uniformity in indoor environment, confirming its potential as a low-cost, effective solution for optimizing indoor wireless infrastructure.