Indoor Navigation with Ambient Radio Signals

INDOORS is a research project which studies indoor localization based on ambient radio signals broadcasted by FM, TV and cellular stations (also called “infrastructure-free localization”).


Indoor navigation remains an open challenge despite the success of satellite-based positioning systems, such as GPS, Galileo, and GLONASS. While these global systems aim to provide accurate world-wide positioning, they are unavailable in areas without a clear sky view, such as urban canyons and indoor environments. Existing indoor positioning systems require installation of dedicated in-building infrastructure.

The aim of the project is to explore indoor positioning based on ambient radio signals, such as FM and TV broadcasts, cellular network signals. While not initially designed for localization, ambient radio signals are transmitted with high power, in different frequency bands and from multiple locations, and thus provide coverage and indoor reception even in less populated areas. Pioneering works have already demonstrated feasibility of indoor localization with FM, TV and GSM signals. However, they only proved the concept and more research is required to evaluate practical benefits and limitations of indoor localization based on ambient radio signals.

The following research questions will be addressed: 1) What is the localization performance of ambient radio based systems over a long time span, in terms of accuracy, time stability and robustness to environment dynamics? 2) Which signals properties apart from signal strength can be used for localization? 3) Which radio types/bands, signal features and localization methods, or their combinations, can provide best performance, stability and robustness?

The project focuses on real-world experimental approach. Firstly, a multi-band radio signal acquisition and localization platform will be created and employed to systematically collect raw multi-band signal samples in multiple locations across several indoor testbeds, over the course of 12 months. In parallel with data collection, the project will explore and develop relevant signal processing methods and localization algorithms; the latter will include both basic and advanced methods derived from state-of-the-art indoor localization systems. Analysis of the collected data with developed algorithms will provide insights to the research questions.

As a result, the project will provide understanding of practical bounds of ambient radio based indoor localization. The collected data will be released to the research community, thus providing a common reference for evaluation of novel localization algorithms. All of the above will facilitate further research of this relatively young approach to indoor localization, potentially leading to widely available and cost-efficient indoor positioning which will in turn boost the development of truly ubiquitous location-based services.

Existing solutions

A wide variety of technologies and sensing modalities have been explored to date, including video, audio, light, pressure sensors, radio (such as Wi-Fi, Bluetooth, ultrawideband signals, pseudolites transmitting GPS-like signals).

Unfortunately, current systems still have two common limitations:

  • their availability is limited to instrumented buildings/areas only;
  • they have considerable hardware, installation and maintenance costs.

Some facts

  • Only 10% of wireless 911 calls in Washington D.C. had accurate location data in the first half of 2013. [Find Me 911]
  • Around 64% of wireless 911 calls are made from indoors and do not have accurate location information. [The Washington Post]
  • Disoriented firefighters lost their lives in a burning building because they could not find the way back. [IEEE Spectrum]

Why this matters

The lack of a cheap and widely available indoor localization technology blocks wide adoption of indoor location-based services, including context-aware applications, location-based gaming, wheelchair and blind navigation, shopping assistants, vehicular navigation in warehouses and underground parking, and emergency rescue navigation.


Questions? Ideas? Proposals? Comments? Please contact the principal investigator Dr. Andrei Popleteev.

Open source

The data acquistion platform (DAQ) of the project is now available as open-source software, available under the MIT license.