Week-Long Galileo Outage Is a Cautionary Tale about Redundancy

On July 11, 2019, in the early European afternoon, users of the Galileo Global Navigation Satellite System (GNSS) constellation were met with a notice of service degradation in all Galileo satellites “until further notice.” Galileo satellites had stopped providing accurate time and orbital data, without which correct location computations cannot be done. The problem persisted for an entire week, until the service was restored on July 18. The European GNSS Agency (GSA) later pointed to an equipment malfunction in the Galileo ground infrastructure as the cause of the outage. Galileo is now in the final stages of its initial services, and thus preparing to become fully operational next year, when a number of critical timing and positioning systems will become dependent on the availability and reliability of the constellation. While it did not wreak havoc in GNSS-dependent systems, this outage event serves as a cautionary tale to companies and entire countries that lack a back-up plan for when our trusty satellite systems fail us.

There currently exist four GNSS constellations that have global availability or will achieve it shortly: the American Global Positioning System (GPS), the Russian Global Navigation Satellite System (GLONASS), the European Galileo, and the Chinese BeiDou. The reason why Galileo’s outage did not cause chaos in GNSS-dependent systems is because most Galileo-compatible GNSS chipsets can also read signals from at least one other constellation, usually GPS. Hence, while the GSA was working to fix the outage, GNSS systems ignored Galileo signals and relied only on navigation and timing signals from other constellations.

Nevertheless, this situation invites one to ponder on certain possibilities that might have previously seemed too remote. What if this outage had happened to the GPS system instead? ABI Research estimates that to this day about a quarter of global GNSS Integrated Circuit (IC) shipments are GPS-only, so a failure in the GPS system would render all of these devices temporarily useless. What if several of the GNSS constellations went down at the same time, be that by a strike of bad luck or as a result of a coordinated attack?

Unbeknownst to the average citizen, modern human societies are extremely reliant on accurate GNSS signals, which support everything from the positioning of cars in the road to Search and Rescue (SAR) operations in remote locations, the timing data needed to run the Internet to timing data supporting cellular networks, stock exchanges, and the electrical grid, and more. In a 2017 report, London Economics evaluated the impact a five-day disruption to GNSS systems would have on the United Kingdom’s economy. This impact was estimated to be £5.2 billion (about US$6.7 billion), mostly as a result of disruptions to its transportation network (37%), emergency and justice services (30%), and maritime sector (21%). Having several global GNSS constellations is an important step toward decentralizing responsibility for these systems and offering back-up options in case a constellation fails. It is no coincidence that a major trend can be identified today in GNSS chipset manufacturers, which are looking to launch multi-constellation solutions the extra accuracy, availability, and reliability they provide.

Not all applications can rely on multi-constellation GNSS, however. For instance, as explained in the ABI Insight Low-Power GNSS Solutions to Open New Possibilities in IoT (IN-5300), Low-Power GNSS(LP-GNSS) devices are designed to keep running for years without their batteries needing to be changed. Achieving such power efficiency involves a number of decisions, including which constellations to work with. BeiDou signals can consume up to five times the energy of the measuring device that GPS and GLONASS do. Galileo can consume eight times as much. Given that LP-GNSS connections are set to reach 320 million in 2022, it is clear that a considerable segment of the GNSS market may not have the option to rely on several constellations.

Other than ensuring that GNSS constellations be run with the gravitas that they require, the first line of defense against disruption from outages like the one Galileo suffered last week lies in implementing multi-constellation capabilities in GNSS chipsets. ABI Research’s Outdoor/Wide Area Location Technologies (MD-OSC-102) Market Data forecasts show that in 2023 about 87% of all GNSS ICs shipped globally will be multi-constellation and, specifically, 39% will be able to read signals from all four global constellations.

Nevertheless, especially with the growth of the Internet of Things (IoT) and the LP-GNSS applications that come with it, it is also important to have back-ups for Positioning, Navigation, and Timing (PNT) that are not GNSS-based. The United States government recognized this many years ago, but took a long time to act on it. The U.S. Department of Transportation (DOT) mandated in 2017 that a back-up PNT system be created in case GPS signals become unavailable. It is expected that the DOT will present its solution, which might include technology offered by the likes of Locata, NextNav, Skyhook, Qualcomm, and Globalstar, by the end of this calendar year.

GNSS is a victim of its own success. Due to being reliable, cheap, and available almost everywhere, it managed to centralize too many critical systems under its wings. The world’s dependence on PNT systems is only set to increase in the coming years, and it would be unwise to not have redundancies in place for unforeseen circumstances. As the Galileo outage shows, a long disruption to GNSS is not far-fetched, and businesses and governments must work together to create contingency plans for such scenarios.