Equitable internet coverage has emerged as a key global priority, which is essential for promoting inclusive and sustainable development. The Indonesian government aims to provide universal internet access by 2024, particularly in remote regions. This study introduces a novel machine-learning-based approach to identify the priority areas for deploying Base Transceiver Station (BTS) towers, which are crucial for achieving the internet access targets of the government. A BTS Network Priority Index was developed by integrating the internet demand estimates with a BTS suitability index derived from key predictors: proximity to fiber optic stations, physical–environmental suitability, and infrastructure–economic readiness. The model identified areas with high internet demand and high BTS suitability as the most critical for immediate development, covering 20 km2. Additionally, future BTS development should target areas with high demand but medium suitability (900 km2) and medium demand but high suitability (280 km2). To validate the methodology, the Random Forest model was employed, which achieved an area under the curve value of 0.7315, indicating strong predictive performance. For the BTS Deployment Suitability parameter, the median was 0.65, with the lower and upper quartiles at 0.44 and 0.85, respectively, confirming that most proposed locations are highly suitable for development. This systematic approach provides data-driven insights for the equitable distribution of BTS towers to ensure efficient internet infrastructure expansion across Indonesia. Furthermore, the study offers a framework that can be adapted by other countries aiming to improve their digital infrastructure and achieve comprehensive, equitable internet access.

中文翻译：

---

用于 BTS 基础设施规划的地理空间智能框架，以实现印度尼西亚的普遍互联网接入目标


公平的互联网覆盖已成为一项关键的全球优先事项，这对于促进包容性和可持续发展至关重要。印度尼西亚政府的目标是到 2024 年提供普遍的互联网接入，尤其是在偏远地区。本研究引入了一种基于机器学习的新型方法，以确定部署基站收发信台 （BTS） 塔的优先领域，这对于实现政府的互联网接入目标至关重要。BTS 网络优先级指数是通过将互联网需求估计值与从关键预测因子（与光纤站的距离、物理-环境适宜性和基础设施-经济就绪性）得出的 BTS 适宜性指数相结合而开发的。该模型将互联网需求高和 BTS 适宜性高的区域确定为对立即开发最关键的区域，占地 20 平方公里。此外，未来的 BTS 开发应针对高需求但适宜性中等 （900 km2） 和中等需求但适宜性高 （280 km2） 的区域。为了验证该方法，采用了随机森林模型，该模型实现了 0.7315 的曲线下面积值，表明具有很强的预测性能。对于 BTS 部署适宜性参数，中位数为 0.65，下四分位数和上四分位数分别为 0.44 和 0.85，这证实了大多数拟建位置都非常适合开发。这种系统的方法为 BTS 塔的公平分配提供了数据驱动的见解，以确保在印度尼西亚高效扩展互联网基础设施。此外，该研究提供了一个框架，其他国家可以采用该框架，旨在改善其数字基础设施并实现全面、公平的互联网接入。