Olasılıksal çıkarsama kullanarak işbirlikli konumlandırma

Accurate localization is crucial in today's interconnected world for enhancing the functionality and efficiency of various localization-based services. Traditional localization techniques such as triangulation, trilateration, and fingerprinting offer acceptable accuracy under ideal conditions but face significant challenges in complex environments due to factors like multipath interference, signal attenuation and limited connectivity. These challenges highlight the necessity for more robust and flexible localization methods. Cooperative localization, which utilizes collaborative information sharing among multiple targets and reference points, emerges as a promising solution to enhance localization accuracy and reliability, particularly in environments where traditional systems fall short. This thesis adopts a probabilistic approach to address the limitations of traditional localization approaches through utilizing collaboration among agents. In this study, formulation of localization problem using a Markov Random Field graphical model is utilized and Nonparametric Belief Propagation to solve it is used. The methodology involves simulating Received Signal Strength measurements as input data, which are then used to approximate the belief distributions of target locations using Particle Filtering and Kernel Density Estimation. Through extensive simulations, the thesis demonstrates that probabilistic framework not only enhances the accuracy and robustness of localization but also adapts effectively to varying environmental conditions and noise levels. The results reveal that probabilistic inference methods significantly improve localization performance by effectively modeling uncertainties and incorporating prior knowledge, making them well-suited for dynamic and complex scenarios where traditional methods may fail. This comprehensive exploration of cooperative localization using probabilistic inference provides valuable insights into its potential applications and effectiveness, offering a robust alternative to conventional localization techniques.