[Paper Review] A Comprehensive Survey of Machine Learning Based Localization with Wireless Signals

Daoud Burghal, Ashwin T. Ravi+3 more|arXiv (Cornell University)|Dec 21, 2020

Indoor and Outdoor Localization Technologies364 references42 citations

TL;DR

This paper surveys ML-based localization using RF signals, detailing system architectures, input features, ML methods, datasets, and open challenges.

ABSTRACT

The last few decades have witnessed a growing interest in location-based services. Using localization systems based on Radio Frequency (RF) signals has proven its efficacy for both indoor and outdoor applications. However, challenges remain with respect to both complexity and accuracy of such systems. Machine Learning (ML) is one of the most promising methods for mitigating these problems, as ML (especially deep learning) offers powerful practical data-driven tools that can be integrated into localization systems. In this paper, we provide a comprehensive survey of ML-based localization solutions that use RF signals. The survey spans different aspects, ranging from the system architectures, to the input features, the ML methods, and the datasets. A main point of the paper is the interaction between the domain knowledge arising from the physics of localization systems, and the various ML approaches. Besides the ML methods, the utilized input features play a major role in shaping the localization solution; we present a detailed discussion of the different features and what could influence them, be it the underlying wireless technology or standards or the preprocessing techniques. A detailed discussion is dedicated to the different ML methods that have been applied to localization problems, discussing the underlying problem and the solution structure. Furthermore, we summarize the different ways the datasets were acquired, and then list the publicly available ones. Overall, the survey categorizes and partly summarizes insights from almost 400 papers in this field. This survey is self-contained, as we provide a concise review of the main ML and wireless propagation concepts, which shall help the researchers in either field navigate through the surveyed solutions, and suggested open problems.

Motivation & Objective

• Assess how ML can mitigate RF localization challenges across indoor and outdoor environments.
• Summarize input features, wireless technologies, and standards that influence ML-based localization.
• Categorize ML frameworks, data availability scenarios, and DL-dominated approaches in localization.
• Highlight publicly available datasets and practical challenges to guide future research.

Proposed method

• Review ML fundamentals and wireless propagation concepts to establish a common baseline.
• Categorize localization approaches (trilateration/ ToA, proximity, fingerprinting, direct methods) and how ML integrates with them.
• Survey input features and their relation to wireless technologies and standards.
• Examine ML frameworks, data availability (supervised, unsupervised, transfer learning), and DL-focused solutions.
• Summarize datasets and provide open problems and directions for future work.

Experimental results

Research questions

• RQ1 What ML techniques are applied to RF-based localization across different system architectures?
• RQ2 What input features and wireless standards most impact ML-based localization performance?
• RQ3 How do data availability and learning frameworks (supervised/unsupervised/transfer) shape ML localization solutions?
• RQ4 What public datasets exist and what are the main open challenges and research directions?

Key findings

• ML-based localization is categorized around four main classes (trilateration/ ToA, proximity, fingerprinting, direct methods) with ML augmenting or replacing traditional steps.
• Feature choices and preprocessing substantially influence ML performance, and ML can leverage heterogeneous data sources.
• Deep learning has driven recent growth in ML-based localization, with many recent solutions leveraging DL architectures.
• Public datasets and data acquisition methods are summarized to support reproducibility and benchmarking.
• The paper identifies key challenges (training data availability, robustness, real-time computation on devices, feature selection) and proposes future research directions.