GNSS is like opium. Highly addictive, pleasing to the user, but laced with peril when you’re hooked. GPS World readers are well aware of the vulnerabilities, and probably aware of the serious concerns governmental bodies have about our reliance upon it. Here, we consider not so much how it can fail as what the impact of failure is on society, and what mitigations exist to minimize those impacts.
Impact on society
The most commonly cited causes of GNSS failure are jamming (unintentional and intentional) and spoofing. Less well appreciated are solar weather, satellite system errors, receiver system errors and, most importantly, cybersecurity-related errors. Any of these can cause a significant disruption to how society functions today.
The U.S. Department of Homeland Security identified that 15 out of the 18 Critical National Infrastructure sectors were vulnerable to GNSS failure: communications, emergency services, information technology, banking and finance, healthcare and public health, energy (electric, oil and gas), nuclear, dams, chemical, critical manufacturing, defense industrial base, postal and shipping, transportation, government facilities, and commercial facilities.
The threat is real and present. Conflicted areas are routinely jammed and spoofed. Even in peace, GNSS is fragile. In the past year, GNSS interference led to a runway at DFW airport being closed, a 33-hour GNSS outage in the Denver metro area, and even a recent Melbourne Formula One race had to be stopped for 40 minutes due to GNSS data problems.
Mitigations
In most cases, no alternate references are in place, and without them, it is difficult to know that the GNSS data being received is wrong. The ship Stena Impero, for example, was seized by Iran for being in its territorial waters. It is thought to have been victim to spoofing that led it there, but proving it is difficult.
Alternate references exist at low cost, but they generally divide “PNT” into “PN” and “T”, and whether on land, at sea or in the air. On land, most requirements are related to “T” because most applications, such as broadcast facilities and data centers, don’t move, and even when they do, there are enough landmarks available to at least make a sanity check. At sea and in the air, by contrast, motion is the reason we are there, there are few landmarks to confirm location, hence “PN” dominates.
For “T”, armageddon clocks can provide holdover for brief interruptions. However, they must be sufficiently tested to ensure that they succeed. In the Denver incident, radio systems had rubidium clocks for backup, but they drifted too far during the outage to be useable, whereas cellular networks, with alternate terrestrial timing sources, continued to work. Terrestrial time distribution systems over existing IP infrastructure, which Hoptroff supplies, work well globally, but are restricted to land-based time synchronization applications.
For “PN” and “T”, in air, sea and land, low-Earth orbit satellite services such as Satelles, Starlink and OneWeb provide global solutions. Their signals are much stronger than GNSS and therefore are much harder to jam or spoof, but they remain susceptible to interferences such as space weather, which destroyed 40 Starlink satellites on launch last year. However, as the Denver incident shows, until you test the solution, you never really know what might go wrong. eLoran terrestrial wireless solutions are very effective but limited in reach to within a few thousand miles of terrestrial transmitters. Such systems are being installed in South Korea and Saudi Arabia due to their proximity to hostile neighbors. They are likely to be installed in North America and Europe within the next decade.
What all these solutions have in common is that, while they are not expensive, they are not free, and are only available under license. Sovereign GNSS providers have, to date, provided PNT signals at no cost to the consumer. It must be accepted in the new landscape, not just that PNT is no longer free, but also that the supplier can choose its customers. Sovereign GNSS access could be restricted at the whim of a president. Private services are already selective — Starlink chose to provide the Ukraine with service during the current conflict, but it has no obligation to continue to do so.
Get on with the risk register
One of the biggest problems with preparing alternatives to GNSS is that the risk is ignored until it has started to have an impact on business and society. This can be addressed by ensuring that resilient PNT is on corporate and institutional risk registers. It is starting to happen now that governments have started to raise concerns. However, we have a way to go before inclusion on risk registers is a foregone conclusion.
