Comparison of Indoor Location Determination Methods That Use Wi-Fi Fingerprinting

Navigation systems have become an indispensable part of our daily lives. In order for this to be achieved, it is necessary to determine the local location. The most commonly used location determination method is the Global Positioning System (GPS). GPS signals often cannot penetrate closed areas. For this reason, using GPS for navigation in closed areas is not efficient. For this reason, different methods have been developed for position determination for navigation in closed areas. The primary method of these is position estimation via Wi-Fi fingerprint. There are many studies done on this subject. In this article, the performance of different machine learning methods in indoor location estimation based on wireless network signal strength is examined. Using a dataset, methods such as artificial neural networks, k-NN, linear regression, support vector machines, decision trees, and random forests were applied, and the results were compared. The database used is explained in detail. It is explained how this database is applied to machine learning algorithms. In the results, the most successful method and the factors affecting success were evaluated.