frankraDIUM/Spatial-Optimization-of-Telecom-Network-Coverage-and-Customer-Demand-in-Greater-Accra

This project develops a geospatial framework to identify coverage gaps, quantify underserved population, and recommend optimal new cell tower locations in Greater Accra, Ghana. # 📡 Spatial Optimization of Telecom Network Coverage and Customer Demand in Greater Accra Final Map Preview --- Dashboard --- ## Project Overview End-to-end geospatial project to identify coverage gaps and recommend optimal new cell tower locations in Greater Accra, Ghana. The solution combines telecom tower data, high-resolution population density, roads, terrain, and land use to generate ranked candidate sites using Multi-Criteria Decision Analysis (MCDA). **Objective**: Support data-driven network expansion decisions by combining telecom tower data, high-resolution population density, roads, slope, and land use constraints. ## Key Features - Dynamic coverage modeling using OpenCelliD tower range data - Underserved population quantification (~565,000 people identified) - DBSCAN clustering of high-density gap zones - Full MCDA scoring: Population served, distance to existing towers, road accessibility, slope, and land use constraints (ESA WorldCover 2021) - Ranked top 20 candidate tower locations - Interactive Folium map and Power BI dashboard ## Technologies Used - **Python**: GeoPandas, Rasterio, Scikit-learn, SciPy, Folium, Contextily, Pandas, NumPy - **GIS**: QGIS (validation & clipping) - **Visualization**: Power BI, Matplotlib, Folium (HTML) - **Data**: OpenCelliD, WorldPop 2020, ESA WorldCover 2021, OSM Roads, SRTM DEM ## Results - **Population coverage rate**: 89.3% (dynamic buffers) - **Underserved population**: ~564,832 (10.7% of 5.295 million) - **Clusters identified**: 8 (after filtering) - **Top candidates**: Ranked by full MCDA score (see `top_tower_candidates_full_mcda.geojson`) ## Interactive Map Open: `top_candidates_interactive_full.html` (zoom, click markers for pop served, distances, slope, land use, score)