Measuring position has been a dramatic success for outdoor navigation systems, and there is a strong push to repeat that indoors. Being able to locate a person inside a building can help in many different ways, from finding the right office in a tower block to locating a required department or even a specific product in a shop, or providing highly targeted deals as you walk around the supermarket.
Unfortunately, the technology of satellite positioning through the GPS, Glonass or the forthcoming Galileo system cannot meet the requirements of indoor positioning even with the nominal enhanced measurement accuracy down to 1 meter as receivers struggle with multipath signals and the inability to see the satellites.
Tests by the European GNSS Agency (GSA) and Rx Networks measured the performance of Galileo when used in various combinations with GPS and GLONASS in real-world environments, including urban canyons and indoors, using a multi-constellation GNSS receiver. While using Galileo, or multiple satellites, helps in urban canyons outdoors, the results show that indoor performance is still poor (Figure 1).
|Urban Canyon #1||Urban Canyon #2|
|GPS||331.9 m||76.2 m|
|GPS+GLONASS||42.9 (13%)||7.6 m (10%)|
|GPS+Galileo||10.7 (3%)||5.4 m (7%)|
|GPS+GLONASS+Galileo||43.0 (13%)||24.7 m (32%)|
|Positive numbers indicate improvement over GPS.|
|Indoor #1||Indoor #2|
|GPS+GLONASS||68.4 m (25%)|
|GPS+Galileo||24.6 m (9%)||10.1 m (14%)|
|GPS+GLONASS+Galileo||64.0 m (23%)||15.8 m (23%)|
|Positive numbers indicate improvement over GPS.|
However, satellite coverage is one key technology, and the MAX2769 from Maxim also covers GPS, GLONASS, and Galileo navigation satellite systems on a single chip that can be integrated into wearable and portable designs. This single-conversion, low-IF GNSS receiver uses a low-power SiGe BiCMOS process technology to provide high performance and integration at a low cost.
Incorporated on the chip is the complete receiver chain, including a dual-input LNA and mixer, followed by the image-rejected filter, PGA, VCO, fractional-N frequency synthesizer, crystal oscillator, and a multibit ADC. The total cascaded noise figure of this receiver is as low as 1.4 dB that can help provide increased sensitivity for indoor use.
The MAX2769 also eliminates the need for external IF filters by implementing on-chip monolithic filters and requires only a few external components to form a complete low-cost GPS receiver solution in a small form-factor for wearable designs. The integrated delta-sigma fractional-N frequency synthesizer allows programming of the IF frequency within a ±40 Hz accuracy while operating with any reference or crystal frequencies that are available in the host system, and the data is output at CMOS logic levels or at the limited differential logic levels.
To improve the performance of such devices for use for indoor positioning, it is possible to add other technologies to enhance the positioning accuracy indoors. One approach is to use a surface-mount 3-axis accelerometer that is already present in many smartphones, such as the Freescale Semiconductor MMA8653, to determine the orientation of the terminal. By starting with a satellite position, any twists and turns can be detected to provide an inertial measurement of position. Unfortunately this requires a regular satellite measurement, which can drain the battery of the phone, and has shown to struggle with accuracy. This also requires maps of the indoor environments, which can be a problem.
Another approach is to use local Wi-Fi wireless signals to determine location. This presents a challenge to antenna manufacturers to combine the different sensitivity requirements of GPS and Wi-Fi. Indeed, modules such as the Antenova M10478 are specifically designed to reject the 2.4 GHz band to prevent interference and enhance the accuracy of the GPS reception.
The RADIONOVA M10478 RF Antenna Module is an ultra-compact single package that combines the RF and antenna on the same module for both L1-band GPS and Assisted-GPS systems.
It is based on CSR’s SiRFstarIV GPS architecture, but the key is that it is combined with Antenova’s high-efficiency antenna technology to provide an optimal radiation pattern for GPS reception. All front-end and receiver components are contained in a single-package laminate base module, providing a complete GPS receiver for optimum performance.
The M10478 operates on a single 1.8 V positive supply with low power consumption and several low-power modes for further power savings. An accurate 0.5 ppm TCXO ensures a short time to first fix (TTFF) which is vital for an inertial navigation combination. The M10478 is supported by SiRF’s software and links to the controller via UART, SPI or I²C host interfaces.
Similarly the JF2 from Telit is a 1.8 V module based on the SiRF IV GPS chip. This has the same UART, SPI or I²C host interfaces to an external controller, but is also optimized to link to a Telit cellular phone module. This provides the assisted-GPS capability that uses some data from the satellites and links it with data from cell phone masts to give a quicker time to fix. However, indoors this can suffer from a lack of penetration, particularly for the 1800 MHz signals.
As a result there are several new companies vying to provide location information with several different approaches, although there is also the challenge of meeting the ISO/IEC 24730 standard on real-time locating systems (RTLS).
Following its acquisition of WiFiSlam last year, Apple has been granted a patent application for a system that combines GPS, Wi-Fi access points and on-board location databases to provide indoor location information. This makes use of multiple Wi-Fi access points to narrow down the location of the terminal by sending a code to a server-based location system. The system then estimates the "presence areas" of other devices within range of the access point. Other access points nearby are then used to refine the location information, especially those in the presence area.
Q-Track of Alabama is taking a different approach, using a 1 MHz wireless signal to provide location information. Using a low frequency gives more penetration through floors and walls and is less susceptible to multipath interference.
However, the Q-track technology does not use signal strength to measure the distance between transmitter and receiver or time of flight like GPS. Instead, it measures the phase of the signal and uses the near-field properties to determine the position of the receiver and the distance from the transmitter. Outdoors the system is accurate to 15 cm says the company, rising to several meters indoors. However, by mapping out the RF environment of the building, this can be reduced to 40 cm, allowing the Q-track tags to be located accurately.
With the Internet of Things (IoT) adding more wireless connectivity, there are other opportunities for locating the position of people and tags indoors without relying on GPS.
Decawave in Dublin uses time of flight measurements from low power, spread spectrum GHz pulses to provide an indoor accuracy down to 10 cm. This is mainly used for locating equipment rather than wearable systems, although it is used for monitoring health equipment.
The DW1000 ScenSor (Seek Control Execute Network Sense Obey Respond) uses the same ultra-wideband techniques of the IEEE802.15.4 2011 standard used by ZigBee with data rates up to 6.8 Mbit/s and a range of up to 300 m from a coherent receiver design. The approach is immune to multipath fading and so allows reliable communications in high fading environments in indoors situations.
Adding indoor positioning technology to wearable devices is a challenge still to be solved. The combination of wireless technologies such as GPS, cellular and Wi-Fi offers a route forward that is present for many, but not all, terminals. This also brings a challenge of form-factor and power consumption. Combining the different technologies and ensuring interoperability and non-interference is also a challenge that designers have to face.