To accurately locate a person, asset, or device within a building It has become crucial for hospitals, factories, shopping centers, offices, and warehouses. The problem is that indoors, GPS is unreliable: walls, ceilings, glass, and metal structures attenuate the signal so much that the receiver can barely use it. That's why, for more than 15 years, there has been intensive research into how to achieve accurate indoor positioning by combining technologies such as Wi-Fi RTT, Bluetooth Low Energy (BLE) beacons, UWB, inertial sensors, and even cameras or light.
Today there are many different solutions, from centimeter-level systems with UWB to hybrid approaches that mix Wi-Fi RTT, BLE, mobile sensors and advanced algorithms (trilateration, fingerprinting, Kalman filters, SLAM…). At the same time, the industry is pushing hard with new standards: Wi-Fi 802.11mc for RTT, Bluetooth 5.1 and 5.3/6.0 for direction finding and Channel Sounding, BLE chips with Time of Flight (ToF) distance measurement, or complete platforms that combine BLE and LoRaWAN to send positions to the cloud with very low power consumption.
What is indoor positioning and why isn't GPS enough?
When we talk about Indoor Positioning System (IPS) We're referring to any system that allows us to locate people or objects within buildings, industrial plants, hospitals, airports, parking lots, etc. Unlike GPS, here it's not enough to know what street we're on: often we need to know if the patient is in the correct room, if the freight elevator is at the right dock, or if a worker has entered a restricted area.
A typical IPS consists of anchors and tagsAnchors are fixed devices (BLE beacons, Wi-Fi access points, UWB nodes, Bluetooth gateways, etc.) installed in known locations. Tags are the mobile elements: these can be smartphones, ID cards, wristbands, asset tags, or small trackers. The system calculates the tag's position based on the signals it exchanges with the anchors and information from the device's own sensors.
The accuracy that can be achieved depends heavily on the technology: from errors of 30-50 cm with UWB to several meters with Wi-Fi or BLEOther factors include anchor density, ambient noise (reflections, people moving, machinery), update frequency, infrastructure cost, and energy consumption.
Main indoor positioning technologies
Today, several types of IPS coexist, each with its own advantages and disadvantages. The most common technologies can be grouped into radiofrequency, ultrasound, light and inertial sensorsas well as hybrid approaches that mix everything to get the best out of each one.
RF technologies: Wi-Fi, Bluetooth, RFID, Zigbee and UWB
Radio frequency (RF) technologies are the most widespread in indoor positioning because they take advantage of existing infrastructure or cheap hardwareAmong the most important are Wi-Fi, Bluetooth, RFID, Zigbee and UWB, each with a different range, accuracy and cost.
Wi-Fi for indoor positioning: RSSI, fingerprinting and RTT
El Wi-Fi positioning It is based on using access points (APs) as anchors. There are two main classic approaches: using received signal strength (RSSI) with trilateration, or building fingerprint maps that collect what RSSI is obtained at each point in the building.
Wi-Fi trilateration estimates the distance to each AP from the received power And, with at least three access points, it calculates the position. It's simple, but very sensitive to the environment: walls, people, furniture, and multipath can generate very large errors, especially if the propagation loss model isn't properly calibrated.
Wi-Fi fingerprinting, on the other hand, consists of a preliminary calibration phase in which the building is traversed, measuring RSSI of all APs on a grid of pointsThen, when the device is in an unknown position, it compares the current RSSI vector with those stored in the database to find the best match. This method is usually more accurate than pure trilateration, but it requires maintenance and recalibrations when the APs or the environment change.
In recent years, Wi-Fi RTT (Round Trip Time, IEEE 802.11mc standard) has emerged, which measures the round trip time of packets between the device and the APSince the propagation speed is that of light, measuring this time allows for a much more reliable distance estimate than with RSSI. Under good conditions, accuracies on the order of 1-2 meters can be achieved. Android 9 and later support Wi-Fi RTT, enabling the use of this technique without additional hardware beyond compatible access points.
Bluetooth Low Energy (BLE) with beacons, AoA/AoD, ToF and Channel Sounding
Bluetooth, and specifically Bluetooth Low Energy (BLE)Today, it's one of the leading IPS systems due to its low power consumption, reduced cost, and widespread support across smartphones, tablets, wearables, and all types of IoT devices. BLE positioning can be implemented passively or actively, using either beacons or gateways.
In classic beacon mode, small BLE devices are deployed that They periodically issue advertising packages. with its identifier (for example, iBeacon, AltBeacon, or Eddystone protocols). Any smartphone or BLE gateway within range can read these packets, measure the RSSI, and estimate the distance based on a reference value (TX power at 1 m) and a propagation loss model. With multiple beacons visible, trilateration or proximity and zone positioning techniques can be applied.
BLE beacons have several advantages: low power consumption (years of battery life), small size and very low costThey use coin cell or AA lithium batteries and, with low transmission power, can last up to 3 years or more. Furthermore, they do not require an internet connection: simply transmitting their identifier is enough to enable navigation services, contextual notifications, proximity messaging, or asset tracking.
Its accuracy using only RSSI is usually around 3-4 meters under typical conditionsHowever, it is highly dependent on the environment. To stabilize the signal, many systems apply filtering (e.g., Kalman filter) that smooths out RSSI fluctuations. Even so, random noise and multipath still limit accuracy, hence the use of more sophisticated techniques such as BLE fingerprinting.
Bluetooth has evolved to improve this situation: versions 5.1 and later introduce the address searchThis allows estimation of the angle of arrival (AoA) or angle of departure (AoD) of the signal using antenna arrays. This opens the door to angle-based triangulation, with much smaller errors and accuracy approaching one meter or even sub-meter in controlled environments.
More recently, the Bluetooth specification adds techniques for Channel Sounding and Time of Flight (ToF)Similar to Wi-Fi RTT, these technologies allow for much more accurate distance measurements than RSSI. Some manufacturers, such as Texas Instruments, have already released BLE chips capable of Time-of-Flight (ToF)-based distance measurements, bringing BLE closer to the realm of precise time-of-flight positioning.
In addition to the beacon approach, there is the model based on Bluetooth gateways For passive positioning of people or active positioning using BLE wristbands or tags. In this case, gateways continuously scan the environment for nearby BLE devices (e.g., wristbands in prisons or nursing homes), report the RSSI of each tag seen to the server, and the central engine calculates the position in real time. Typical accuracy is also around 3-4 meters, improving upon Wi-Fi in terms of stability and power consumption.
UWB: the centimeter-based option
The UWB system usually works by using ToF-based trilateration, measuring the time it takes for the signal to go and return or the arrival time between different anchors. The large bandwidth provides high temporal resolution and an improved ability to distinguish direct paths from reflections, which enhances robustness against obstacles and building materials.
In return, UWB requires specific infrastructure deploymentIt consumes more frequency bandwidth and is subject to regulatory restrictions (typical bands between 3,1 and 10,6 GHz with limited power). The practical range is usually tens of meters, and the cost per anchor and tag is higher than in BLE, so it is reserved for applications where centimeter-level accuracy is truly critical (industrial automation, robotics, high-security access control, automotive).
RFID and Zigbee
In addition to Wi-Fi, BLE and UWB, other technologies have also been used RFID and Zigbee RFID is used to locate objects indoors. It employs electromagnetic fields to identify passive, semi-passive, or active tags, with ranges from centimeters to about 100 meters in the case of active tags. It is ideal for identification and inventory control, but not so much for continuous positioning, as it does not provide precise coordinates or real-time tracking on its own.
Zigbee, on the other hand, is a standard for low-consumption mesh network Widely used in control and monitoring (home automation, smart metering, etc.). Although it can be used for positioning using RSSI or mesh techniques, in practice its role has been overshadowed by BLE, which offers a much larger installed base and better support in mobile and consumer devices.
Ultrasound, infrared, and light
Outside of radio frequency, there are IPS systems based on ultrasound, infrared, or illuminationUltrasound measures the flight time of acoustic waves between transmitters and receivers, similar to sonar. It can achieve sub-meter accuracy, but is sensitive to temperature, ambient noise, and solid obstacles, and requires a significant number of anchors and maintaining acoustic visibility.
Infrared systems require direct line of sight between labels and anchors. They have been used as room detectors and in virtual reality systems, where various light sources and reflective elements allow for highly accurate user location. The problem is that any obstacle blocking the beam disrupts the measurement, so coverage can be fragile.
Finally, some lighting manufacturers have developed solutions for visible light-based positioningIn these systems, each light emits a unique blinking pattern that the mobile phone's camera can detect. This allows for highly accurate user location, but requires replacing the existing lighting and maintaining a specific hardware and software provider.
IMU and inertial positioning
All modern smartphones integrate a inertial measurement unit (IMU) with accelerometers, gyroscopes, and magnetometers. By combining these signals, the relative movement of the device in 3D space can be reconstructed: how far it has moved, in what direction, how many times it has rotated, whether it has changed floors, etc.
This approach, known as dead reckoning Dead reckoning, or sunk navigation, doesn't require anchors, but its accuracy degrades over time because errors accumulate. In a matter of seconds or a few minutes, the estimated position can shift by several meters. Therefore, the IMU is typically used in conjunction with other technologies (Wi-Fi, BLE, magnetometer, barometer, digital maps) to correct and readjust the course.
Some systems based on inertial sensors have gone a step further and propose the so-called “Indoor GPS without beacons or app”The initial position is obtained, for example, by scanning a QR code that opens a web app with the building's floor plan. From there, the mobile phone's IMU updates the position as the user moves. It's a very interesting solution because it doesn't require installing hardware or forcing the user to download a native app, although for now it's limited: it doesn't allow geomarketing, background notifications, or tracking with the screen off without beacon or Wi-Fi support.
Computer vision, light and SLAM
La computer vision It's another key component in some advanced IPS systems. The principle is simple: the user points the camera around them, and the system compares the images with a database or a 3D model of the building to determine the vantage point from which they were taken, or even to leverage other possibilities. live view to supplement location tracking. It can also detect QR codes or other visual markers to pinpoint the location.
In addition, many indoor navigation systems employ techniques of SLAM (Simultaneous Localization and Mapping)These systems use sensor data (IMU, camera, etc.) to build a map while simultaneously locating the user within it. These approaches are very powerful in robotics and autonomous vehicles, and are beginning to be adopted for mobile devices, but they require significant computing power and are not always practical for large-scale deployments.
Position calculation methods: RSSI, trilateration, triangulation, and fingerprinting
Beyond the physical technology, the heart of an IPS lies in the location methods that are applied to the received signals. Among the most common are RSSI, trilateration, triangulation, fingerprinting, AoA/AoD, dead reckoning, and even filtering algorithms like Kalman.
The use of RSSI It is the simplest and most widespread: the power of the signal received from several anchors is measured (for example, the signal strength in dBmand is translated into an approximate distance using a propagation model. This method is cheap and easy to implement, but very sensitive to obstacles and changes in the environment. For this reason, it is usually combined with filtering, maps, or more robust techniques.
La trilateration It takes those distances and calculates the point where the circles centered on each anchor intersect. This is the typical method in GPS, Wi-Fi, and BLE when the positions of the transmitters are well known. Triangulation, on the other hand, uses angles of arrival (AoA/AoD) Instead of distances: with antenna arrays and good electronics, a fairly accurate direction to the device can be obtained and, using several anchors, its position pinpointed.
El fingerprinting This is especially interesting for precise positioning with BLE and Wi-Fi. In the first phase, the environment is scanned, taking RSSI (or even magnetic field) samples on a grid of points. In the second phase, when the device needs to be located, the current signal vector is compared with the database using classifier or regression algorithms. This approach is usually much more robust than trilateration based solely on theoretical models, and can achieve errors of around 2 meters or less in well-mapped environments, at the cost of that initial calibration effort.
To improve stability, many systems incorporate Kalman filters or other Bayesian filters These systems combine sensor data (RSSI, IMU, barometer, etc.) with a motion model. This reduces abrupt changes in the estimated position and makes the route feel much more natural to the user.
Comparison: BLE beacons versus Wi-Fi and other systems
When an organization considers deploying an indoor positioning system, the usual approach is to compare BLE, Wi-Fi, UWB beacons and purely inertial or visual solutionsThere is no universal answer, but there are a number of clear criteria: accuracy, coverage, cost, consumption, maintenance, and user experience.
BLE beacons They tend to win in terms of cost and power consumption: they are very inexpensive, easy to install, require no power wiring if battery-powered, and are supported by most smartphones. Their accuracy with RSSI and trilateration is around 3-4 meters, which can be improved to around 1-2 meters with advanced algorithms and a good beacon density (for example, 3-4 devices per 200 m² or even more in complex areas).
Wi-Fi, on the other hand, takes advantage of a infrastructure that is almost always already deployedThis reduces the incremental cost. However, it consumes more power, the signal is less stable, and on iOS, access to Wi-Fi scanning is very limited, so many iPhone solutions rely on BLE. Wi-Fi RTT can greatly improve accuracy, but it requires compatible access points and relatively modern mobile devices.
UWB is the option to choose when Centimeter-level precision is needed And the investment is justified: warehouse robots, AGVs, hands-free vehicle access control, ultra-fine tracking of critical tools, etc. The cost per node and the need for specific infrastructure mean it's not the first choice for visitor guidance or basic geomarketing.
Finally, solutions based exclusively on IMUs, cameras, or lighting offer alternatives without the need for dedicated hardware, but their practical usability and its robustness They are still behind RF options for many massive use cases.
BLE + Wi-Fi RTT + LPWAN: hybrid architectures and use cases
One of the most powerful approaches that is gaining traction is that of the hybrid systemswhich combine various technologies depending on the environment and the objective. For example, there are locators that alternate between outdoor precision GPSWi-Fi RTT and BLE indoors, and low-power networks such as LTE-M, LoRa or Sigfox to send data to the server with minimal consumption.
A good example is the combination of BLE beacons for local positioning and LoRaWAN for data transmissionIn this architecture, small trackers using BLE and LoRaWAN listen to the beacons deployed throughout the building and calculate their position using trilateration or proximity. They then report the coordinates (or at least the ID of the nearest beacon) through a LoRaWAN gateway, which can cover an entire building or campus. The backend, often open source, receives the data and displays it on a web dashboard, allowing users to view assets, people, or vehicles in near real-time.
This model is very attractive because it drastically reduces the number of gateways required and Take advantage of LoRaWAN's low power consumptionIn addition, the trackers can incorporate SOS buttons for emergencies, accelerometers to detect movement, and intelligent logic to send less data when stationary, extending battery life to several months.
On the smartphone side, commercial solutions such as those from some indoor navigation providers combine Wi-Fi (where available), BLE, IMU, magnetometer and barometer for location and guidance. On Android, they can even do without beacons by using the existing Wi-Fi network; on iOS, where Wi-Fi scanning is limited, they rely more on BLE and sensor fusion, greatly reducing the number of beacons needed compared to other systems.
Bluetooth IPS in detail: operating modes and deployment
The Bluetooth-based indoor positioning system (Bluetooth IPS) has established itself as one of the most well-rounded solutions in terms of cost, consumption, accuracy, and ease of deploymentIt works with both fixed anchors (beacons or sensors) and with tags or mobile devices that act as transmitters.
In mode positioning with BLE sensorsFixed sensors (BLE gateways) are placed around the interior space. These sensors passively detect all BLE transmissions from tags, mobile devices, or wearables and measure their RSSI. The signal data is sent to a central server where a positioning engine calculates coordinates using trilateration, fingerprinting, or a combination of both. The server can then display the position on an interior map and trigger actions such as alarms, notifications, or geoanalytics reports.
In mode positioning with beaconsThe logic is reversed: the beacons are fixed, and the mobile device (phone, tag, or tracker) calculates its position based on the beacons it detects. This allows for the development of indoor navigation services (the typical "blue dot" that moves across the map), proximity messaging (coupons, offers, contextual alerts), and virtual geofences that trigger actions upon entering or leaving a designated area.
The typical deployment of a BLE IPS system requires careful planning beacon or gateway densityThe installation height (roof, walls, poles), power source (battery, PoE), and radio configuration (advertising interval, TX power, channels) are all important factors. Furthermore, it is crucial to map the environment in detail, record the coordinates of all anchors, and document MAC addresses, UUIDs, and other parameters to facilitate maintenance and troubleshooting.
The use cases are very varied: critical asset tracking in hospitals and industries, patient location, visitor guidance in shopping centers or airports, people flow analysis, workplace safety, restricted area control, and proximity marketing campaignsThe same infrastructure can serve many of these applications at the same time, which improves the return on investment.
In real-world projects, for example in a large shopping mall, BLE systems have been deployed using ESP32s or commercial beacons to collect signals, apply Kalman filters, combine trilateration and fingerprinting, and offer visitors an app or even a simple desktop interface that displays their position, routes, and points of interest. All of this is achieved by leveraging the Low BLE power consumption, compatibility with modern mobile devices, and ease of integration with cloud platforms.
Ultimately, precise indoor positioning with Wi-Fi RTT, BLE beacons and complementary technologies relies on a mix of increasingly capable hardware (chips with ToF, Bluetooth 5.x/6.0, Wi-Fi RTT AP, hybrid trackers) and intelligent software (trilateration, fingerprinting, sensor fusion, Kalman filters, SLAM).
Choosing the right combination for each project involves thoroughly analyzing the environment, the required level of precision, the budget, and the limitations of the devices, but the good news is that today it is perfectly feasible to set up reliable, scalable, and quite precise systems without having to spend a fortune on infrastructure or force the user to struggle with obscure technologies. Share this information so that more users can learn about the topic.