Imroatul Nasiyah
"Cities are brutal for satellite navigation," explains Ardeshir Mohamadi, a doctoral fellow at the Norwegian University of Science and Technology (NTNU). His research is at the forefront of a critical quest: to dramatically enhance the accuracy of affordable GPS receivers, those ubiquitous components found in our smartphones and fitness watches, without resorting to costly external correction services. This pursuit of pinpoint precision is not merely an academic exercise; it holds paramount importance for the future of transportation, especially for autonomous vehicles that depend on unfaltering positional awareness to navigate safely.
The primary culprit behind GPS’s urban struggles lies in what are commonly known as "urban canyons." These are the dense city environments characterized by towering skyscrapers and an abundance of reflective surfaces. "In cities, glass and concrete make satellite signals bounce back and forth," Mohamadi elaborates. "Tall buildings block the view, and what works perfectly on an open motorway is not so good when you enter a built-up area." When GPS signals, which are essentially radio waves, encounter these obstacles, they don’t travel in a straight line to the receiver. Instead, they reflect off the facades of buildings, taking longer and more circuitous routes. This delay in signal travel time is interpreted by the GPS receiver as a greater distance to the satellite, leading to significant inaccuracies in the calculated position. It’s akin to being at the bottom of a deep gorge, where direct sunlight is scarce, and illumination only reaches you after multiple bounces off the canyon walls.
This phenomenon has profound implications for autonomous vehicles. For these sophisticated machines, the difference between confident, safe behavior and hesitant, unreliable driving can be the margin of error introduced by a city’s urban canyons. "That is why we developed SmartNav, a type of positioning technology designed for ‘urban canyons’," Mohamadi states, highlighting the critical need for a robust solution.
Beyond the issue of signal reflection, even the satellite signals that do manage to reach a receiver in a city without excessive bouncing may lack the necessary precision. Standard GPS, while adequate for general navigation, is not always accurate enough for tasks demanding centimeter-level precision, such as those required by self-driving cars.
To tackle this multifaceted problem, Mohamadi and his team at NTNU embarked on an ambitious project to combine several different technologies into a cohesive system. Their goal was to create a computer program that could be seamlessly integrated into the navigation systems of autonomous vehicles, significantly improving their positional accuracy. Before delving into their innovative approach, it’s beneficial to understand the fundamental principles of how GPS operates.
The Global Positioning System (GPS) relies on a constellation of satellites orbiting the Earth. These satellites continuously transmit radio wave signals containing vital information: the satellite’s precise location in orbit and the exact time the signal was sent. A GPS receiver, such as the one in your smartphone, picks up these signals. By receiving signals from at least four satellites, the receiver can triangulate its position on Earth by calculating the distance to each satellite. The time stamp within the signal is crucial for this distance calculation; the longer the signal takes to arrive, the further away the satellite is.
In the urban canyon scenario, the integrity of this time stamp is compromised. As mentioned, signals bouncing off buildings arrive at the receiver later than they should. This discrepancy throws off the distance calculations, leading to the erroneous position readings. The NTNU researchers explored a novel approach to circumvent this issue: discarding the problematic time-coded data altogether and instead focusing on other characteristics of the radio wave itself.
One such characteristic they investigated is the "carrier phase" of the radio wave. This refers to the wave’s directional orientation – whether it’s traveling upwards or downwards when it reaches the receiver. "Using only the carrier phase can provide very high accuracy, but it takes time, which is not very practical when the receiver is moving," Mohamadi explains. The challenge here is that to achieve precise measurements based on the carrier phase, the receiver needs to remain stationary for an extended period – not just a microsecond, but potentially several minutes – to allow the signal to stabilize and be accurately analyzed. This is clearly not feasible for a vehicle in motion.
While the direct use of carrier phase proved impractical for dynamic navigation, the researchers explored other established methods for enhancing GPS accuracy. One such method is Real-Time Kinematic (RTK). RTK systems utilize a network of ground-based reference stations that transmit correction data to GPS receivers. These stations know their exact location and can detect discrepancies in the satellite signals, then broadcast corrections to nearby receivers. RTK can achieve remarkably high accuracy, down to a few centimeters. However, its widespread adoption for general consumers is hindered by its cost. RTK systems and their associated subscription services are typically intended for professional users like surveyors and engineers who require this level of precision for their work.
A more advanced, yet still potentially accessible, technique is Precise Point Positioning – Real-Time Kinematic (PPP-RTK). This approach combines the benefits of precise point positioning (which uses a single receiver to achieve high accuracy) with real-time kinematic corrections. Crucially, the European Galileo satellite navigation system has begun broadcasting its correction data free of charge, making PPP-RTK a more viable option for broader applications.
But the innovation didn’t stop there. As the NTNU researchers were refining their algorithms in Trondheim, Google was simultaneously launching a new service that would prove to be a crucial piece of the puzzle. This new service leverages the vast amount of data Google has collected about urban environments. Imagine planning a trip to a city like London. You open Google Maps on your device, input your hotel’s address, and can then zoom in to get a detailed view of the streetscape, including the facades of buildings and their relative heights. Google has meticulously created 3D models of buildings in thousands of cities worldwide.
The genius of Google’s contribution lies in its ability to use these 3D building models to predict how satellite signals will reflect and refract within the urban canyons. By understanding the geometry of the city and the likely paths of reflected signals, Google’s service can help mitigate the errors caused by signal multipath. This is particularly effective in solving what’s known as the "wrong-side-of-the-street" problem, where GPS inaccuracies can make it appear as if you’re navigating on the opposite side of the road from your actual location. "They combine data from sensors, Wi-Fi, mobile networks and 3D building models to produce smooth position estimates that can withstand errors caused by reflections," Mohamadi explains, underscoring the power of this integrated approach.
The NTNU researchers were then able to weave together all these disparate threads – the carrier phase information, the PPP-RTK corrections, and Google’s sophisticated 3D city models – with their own custom-developed algorithms. The result was a powerful positioning system, aptly named SmartNav, designed specifically to overcome the challenges of urban navigation.
When they put their combined system to the test on the streets of Trondheim, the results were nothing short of remarkable. They achieved an accuracy that surpassed ten centimeters 90 percent of the time. This level of precision is a significant leap forward, providing a degree of reliability that is essential for autonomous vehicles and other applications demanding highly accurate location data in complex urban settings.
Furthermore, the integration of PPP-RTK technology, which is becoming increasingly accessible due to its free correction data, promises to make this advanced positioning technology available to the general public. "PPP-RTK reduces the need for dense networks of local base stations and expensive subscriptions, enabling cheap, large-scale implementation on mass-market receivers," Mohamadi concludes. This development signals a future where the frustrating GPS inaccuracies of urban navigation are a thing of the past, paving the way for safer, more efficient, and more intuitive journeys through our cities.
