Improving Low-Altitude UAS Navigation

Global navigation satellite system (GNSS) signals are constantly changing.

A startup is testing a new drone-based delivery application for when the aircraft loses its global positioning system (GPS) signal. The vehicle flies off course. Soon, it loses communication, cannot return home, and can't achieve its mission.

An Air Force unit conducts a risk assessment for an upcoming unmanned aerial system (UAS) mission. They know, based on past experience, that global navigation satellite system (GNSS) signals are constantly changing, but don't have a baseline of what is occurring due to terrain or buildings, and what is being caused by the enemy.

An eVTOL aircraft is flying cargo and personnel in urban and suburban settings but can't operate with automation because the GPS signals seem to be ok and then lost at different times and locations. Even the downtown vertiport has occasional issues delaying departures and arrivals. Could it be jamming? Or is something else going on?

All these scenarios share some common themes: GPS/GNSS is the most common source of aircraft navigation and at low altitudes it can be unreliable, Is it possible to make GPS/GNSS more reliable?

As many amazing things as civil and military organizations now do with GNSS, users continue to bump up against its most frustrating limitation: at lower altitudes, you just can't rely on it. Engineers have been working on this issue for years, but the truth is, mitigating GNSS interference is an extremely hard problem to solve — certainly at a cost that's viable for drone applications at scale.

Forecast Risk Analysis best- and worst-case scenarios for downtown Indianapolis.

What if instead of “fixing” GNSS reliability by using other sensors and signals, we could just predict exactly where and when it will be degraded? Wouldn't that mitigate most of these issues? Or even enable performance improvements? That's exactly what new GNSS prediction technologies are doing. And they're about to upend the status quo for many of today's most important aviation use cases, as well as tomorrow's.

GNSS Limitations

GNSS has come a long way since the early days of Navy Transit testing in the 1960s. But the technology still has some baked-in challenges — like the fact that constellations of satellites transmitting navigation signals are constantly in motion. Those satellites orbit the Earth at about one degree per minute, transmitting some 22,000 kilometers down to a given aircraft's (or vehicle's, or device's) receiver. So, the strength and reliability of GPS or other GNSS signals can change literally every second.

For aircraft flying at high altitudes, that's not a problem; they'll typically communicate with up to a dozen satellites at once. Closer to the ground though, terrain and buildings can deteriorate GNSS signals or block them altogether. The problem becomes particularly thorny:

  • At altitudes below 400 feet: Here, aircraft must navigate far more dynamic environments, especially in urban and suburban areas, with clear views of navigation satellites frequently obstructed by buildings and terrain.

  • When flying beyond visual line of site (BVLOS): The Federal Aviation Administration (FAA) has deemed GPS navigation so critical to flying BVLOS, most FAA BVLOS waivers require drone operators to immediately land aircraft if GNSS degrades or is lost. The policy is understandable, but it's become a significant handcuff for technology leaders trying to test new drone business models, much less actually put them into production.

  • When trying to manage unmanned air traffic: With no awareness of terrain or buildings, GNSS as currently implemented cannot be relied on for the safe management of a dynamic automated airspace — certainly not one where thousands of drones navigate densely populated areas.

It's not that these problems are entirely insurmountable. Drones (and other autonomous vehicles and devices) can augment navigation with better inertial systems, additional sensors and cameras, radar, and other workarounds. But these add significant cost and weight, making them impractical for drone use cases at any kind of scale. None offer a 100% reliable solution either. Bottom line: if we're going to unleash the new civil and military drone applications that technologists envision, we must find a practical way to get more reliable navigation. If this can be done by leveraging existing, inexpensive technologies — all the better.

Enter GNSS Prediction

Military and civilian engineers have a long history of solving the most difficult technology problems. But sometimes, the cleverest innovations arise when engineers recognize that even if they can't solve a problem directly, they can get to results that are almost as useful. We might not be able to guarantee reliable GNSS signals everywhere that aircraft need them with current technology. One thing we can do, however, is predict exactly where and when GPS/ GNSS data will be unreliable or unavailable. That's exactly what technology leaders in this space are now doing.

New GPS/GNSS prediction techniques can analyze line-of-sight to each individual satellite over a service area, across every square meter of that area, every second, in four dimensions. These cloud-based solutions ray-trace up to individual satellites to identify every Systems location and altitude where GNSS signals will be blocked by buildings or terrain at a given time. And they can provide that insight back to users, on demand, to enable safer and more reliable navigation in civil contexts, and better planning and situational awareness in military operations. While similar approaches have been used in simulation, a cloud-based scale platform can be used to provide real-time and future predictions across entire cities, territories, and regions.

With GNSS prediction intelligence, users can:

  • Know ahead of time where and when GPS/GNSS will perform best.

  • Plan flights using routes they know to be reliable.

  • Move flight or mission times — sometimes by just a few minutes backwards or forwards — to avoid times when interference is worse.

  • Better evaluate the integrity of the navigation information a system receives.

  • Know the baseline GNSS coverage based on building and terrain interference to determine where bad actors may be creating interference or false signals.

  • Make better use of resources inside the receiver and satellite selection to improve performance.

GNSS Foresight Live view of the same route through downtown Indianapolis, over an hour apart. 8

Possibilities for Predictive GNSS

With these GNSS predictive capabilities, we can continue using mature existing standards and ecosystems around GNSS, but now extend them to lower altitudes. Existing frameworks for GPS certification for IFR flight, for example, certify that systems will work with certain acceptable amounts of interference. With GNSS prediction intelligence, we can now pinpoint exactly how much interference will affect a given flight path at a given time, and devise flight plans to ensure aircraft remain within those thresholds.

For the first time, we can:

  • Predict GNSS reliability over dynamic areas: New prediction technologies provide instant access to GNSS reliability forecasts — down to 1-meter resolution, second by second, for up to 72 hours into the future. This can be useful for determining precisely where and when to fly electric vertical takeoff and landing (eVTOL) multi-rotor, and other types of UAS.

  • Forecast risk: Mission planners can use GNSS prediction data to identify best- and worst-case coverage when piloting unmanned aircraft. For example, they could designate “fly anytime” areas, “no fly” areas, and “fly sometimes” areas, where specific paths are designated good at specific times. These insights can be used for concept of operations (CON-OPS) activities, when selecting sites for vertiports, defining fixed routes through certain geographies, and more. Just knowing the quality of the data navigation systems will have to work with allows for much more accurate risk analysis.

  • Optimize mission planning: In a military context, predictive insights can be used in both offensive and defensive operations that rely on GNSS. It's now possible to pinpoint with a high degree of accuracy where an aircraft can operate, for how long, where it will have redundancy, and what the consequences will be if it needs to change routes. Those characteristics can inform decisions such as at which altitude to fly, or whether to proceed with a mission now or wait 30 minutes for more favorable conditions. In an adversarial environment, knowing the expected reliability of GNSS signals in specific times and places makes it easier to determine which signals are authentic and which might be spoofed.

Unleashing the Aviation Models of Tomorrow

In many ways, low altitude aviation is in a similar place as the early days of flight, when we had no way to predict weather. In those days, aircraft losses were a regular occurrence, especially when flying through storms at night, because we just didn't understand how they would perform in bad weather and had no way to predict it. Fast forward a century, and pilots can access worldwide weather within seconds from anywhere, and simply navigate away from storms.

The progress of low-altitude aircraft and drones is following a similar path. Thanks to novel predictive approaches, we can forecast GNSS performance with confidence anywhere around the globe. By knowing exactly what to expect, we can mitigate interference, plan around it, maximize situational awareness, and even improve performance, unleashing the future of unmanned civilian and military aircraft that we've been waiting for.

This article was written by Jeremy Bennington, VP of PNT Assurance, Spirent Communications (San Jose, CA). For more information, visit here .