LoC-WiFi-SLAM: Low-Complexity of WiFi-Enhanced SLAM

In this paper, we present the strengths of WiFi-enhanced SLAM over traditional SLAM systems. We minimize the Root Mean Square Error (RMSE) in localization and mapping by incorporating WiFi signal information. The work elegantly embeds WiFi signals in the well-studied SLAM optimization architecture without degrading the efficiency on the computational side, providing improved accuracy and reliability. We utilize a deep neural network (DNN) to learn the optimal fusion of WiFi and traditional sensors, which enhances system performance significantly. The presented LoC-WiFi-SLAM system is of moderate computational as well as spatial complexity. Simulation results show that the Loc-WiFi-SLAM system is able to use available WiFi infrastructure to complement traditional sensors, thus achieving an improvement in mapping and location accuracy. It not only increases localization accuracy but also makes it suitable for robotic systems, thus providing a robust and scalable solution for environments.