Hybrid Neural Network-Based Indoor Localisation System for Mobile Robots Using CSI Data in a Robotics Simulator

2025-11-03

AI Summary

To address the challenge of high-precision localization for mobile robots in complex indoor environments, this paper proposes a hybrid neural network (HyNN) localization method leveraging massive MIMO channel state information (CSI). The raw CSI sequences are first transformed into time-frequency images using the TINTO toolkit; subsequently, a CNN extracts spatial features while an MLP models nonlinear mappings, jointly forming an end-to-end position regression model. The HyNN is integrated into a ROS simulation platform and fused with a Kalman filter to enhance robustness in state estimation. Extensive experiments across multiple heterogeneous indoor scenarios demonstrate significantly lower localization error compared to baseline methods and confirm strong cross-scenario generalization capability. To the best of our knowledge, this work represents the first systematic realization of CSI-based image representation and HyNN architecture in a closed-loop, real-time robotic localization framework—establishing a scalable, low-cost, GPS-free paradigm for high-accuracy indoor navigation.

Technology Category

Intelligent Robots: Localization, Mapping, and NavigationMachine Learning: Neuro-Symbolic LearningPlanning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applicationsUser Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendationSearch and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

We present a hybrid neural network model for inferring the position of mobile robots using Channel State Information (CSI) data from a Massive MIMO system. By leveraging an existing CSI dataset, our approach integrates a Convolutional Neural Network (CNN) with a Multilayer Perceptron (MLP) to form a Hybrid Neural Network (HyNN) that estimates 2D robot positions. CSI readings are converted into synthetic images using the TINTO tool. The localisation solution is integrated with a robotics simulator, and the Robot Operating System (ROS), which facilitates its evaluation through heterogeneous test cases, and the adoption of state estimators like Kalman filters. Our contributions illustrate the potential of our HyNN model in achieving precise indoor localisation and navigation for mobile robots in complex environments. The study follows, and proposes, a generalisable procedure applicable beyond the specific use case studied, making it adaptable to different scenarios and datasets.

Problem

• Develops hybrid neural network for robot indoor localization
• Uses CSI data converted to images for position estimation
• Integrates localization with robotics simulator and ROS

Innovation

• Hybrid neural network combines CNN and MLP
• Converts CSI data into images using TINTO
• Integrates localization with ROS and robotics simulator