Mini-Batch Gradient Ascent for Enhanced Indoor Localization in IoT Networks with Rayleigh Fading

This paper presents an innovative algorithm that combines mini-batch gradient descent with adaptive techniques to enhance the accuracy and efficiency of localization in complex environments. The suggested approach utilizes the gradient ascent technique to refine the localization of IoT devices by iteratively refining their estimated positions based on received signal strength (RSS) measurements. By integrating mini-batch processing, the algorithm improves computational efficiency and convergence speed, making it well-suited for real-time applications. Additionally, the adaptive mechanism dynamically modifies the learning rate and batch size based on the current localization error, guaranteeing robust performance even in different signal conditions. The simulation outcomes indicate that the adaptive mini-batch gradient ascent established localization algorithm surpasses conventional techniques in terms of localization precision and convergence rate, especially in complex and rapidly changing indoor IoT networks. The suggested method provides a flexible and effective way to determine precise locations, which is essential for various IoT applications such as indoor, asset tracking, as well as smart home automation.