En ciudades y otros entornos urbanos, existen dos desaf\u00edos comunes para el rendimiento del GNSS. El primero es la interferencia multitrayecto, que se produce cuando las se\u00f1ales rebotan en edificios, fachadas de cristal e incluso coches aparcados antes de llegar al receptor. La oclusi\u00f3n de la se\u00f1al es otro problema que ocurre cuando edificios y estructuras altas bloquean f\u00edsicamente la visi\u00f3n de algunas se\u00f1ales satelitales. Altos edificios de cristal, calles estrechas, puentes de hormig\u00f3n y pasos elevados forman ca\u00f1ones urbanos y pueden ser una barrera incluso para los sistemas de navegaci\u00f3n m\u00e1s sofisticados.<\/p>\n
The term \u201curban canyon\u201d was inspired by New York\u2019s Canyon of Heroes \u2014 a stretch of Lower Broadway where tall buildings line the streets similar to a canyoenn. These human-built canyons can confuse GNSS receivers making it hard to accurately calculate a vehicle\u2019s position. For autonomous cars, that\u2019s not just inconvenient \u2014 it\u2019s a major safety issue. However, with the right technology, the automotive world can \u201cclose\u2019\u201d these urban canyons, explains Manuel Del Castillo, vice president of business development at\u00a0Focal Point Positioning<\/a>.<\/p>\n On open roads with a clear view of the sky, satellite navigation can be remarkably accurate. Signals from multiple GNSS constellations reach the vehicle\u2019s receiver unimpeded, helping calculate position with impressive accuracy. However, this often isn\u2019t the case in dense urban areas.<\/p>\n Tall glass buildings, narrow streets, concrete bridges and overpasses all form urban canyons \u2014 and can be a barrier to even the most sophisticated navigation systems.<\/p>\n The Challenge<\/strong><\/p>\n In cities and other urban environments, there are two common challenges for GNSS performance. The first is multipath interference, which occurs when signals bounce off buildings, glass fa\u00e7ades and even parked cars before reaching the receiver. Rather than receiving one clean signal from the satellite, the receiver gets a clean signal and several delayed copies, leading to erroneous positioning estimates.<\/p>\n Signal occlusion is another issue, which occurs when tall buildings and structures physically block some satellite signals from view. The signals that are actually received from that satellite are reflections. This makes it difficult for the receiver to lock onto a stable fix.<\/p>\n In practice, both issues can cause sudden anomalies \u2014 enough to place a car on the wrong street entirely. For drivers, this is frustrating. For autonomous systems, it\u2019s a safety risk.<\/p>\n The Road to Autonomy<\/strong><\/p>\n Urban GNSS challenges aren\u2019t new \u2014 taxi drivers in London and New York have long experienced their navigation systems getting \u201clost\u201d among the towers. However, positioning accuracy is now more important than ever as automotive technology evolves and we hand over more control to our vehicles.<\/p>\n Advanced driver assistance systems (ADAS) are now pushing the limits of conventional GNSS. Features such as lane-keeping, automated lane changes and intelligent speed adaptation all rely on knowing the vehicle\u2019s exact position \u2013 not just the road it\u2019s on, but which lane.<\/p>\n As we move further towards autonomous driving, the stakes will be even higher. If GNSS references are unreliable, this could cause serious errors on the road. A sudden position jump in the middle of a complex urban manoeuvre is more than inconvenient \u2014 it\u2019s dangerous.<\/p>\n Closing the Canyon<\/strong><\/p>\n If autonomous cars are to drive safely and reliably in urban environments, GNSS must evolve. The answer lies in rethinking how satellite signals are processed \u2014 and in tackling the root causes of error. Traditional receivers rely heavily on hardware-based processing, meaning they integrate new technologies at a slow pace.<\/p>\n To help overcome this challenge, we developed S-GNSS Auto \u2014 software that enhances GNSS receiver reliability and accuracy in autonomous vehicles. Delivered as a simple firmware upgrade, it transforms GNSS into a more powerful component of the ADAS stack in areas where traditional solutions fall short.<\/p>\n We recently integrated S-GNSS Auto\u00a0onto STMicroelectronics\u2019 Teseo GNSS devices<\/a>, and tested the impact of the joint solution in some of the most challenging urban environments: Shinjuku in Tokyo, and Frankfurt and the Black Forest in Germany. The combined solution demonstrated an improvement in measurement accuracy by up to four times and position accuracy by up to three times in the challenging sections of these environments. By ignoring reflected or non-line-of-sight signals, S-GNSS Auto can also reject potential spoofing attacks, enhancing the security of the GNSS receiver.<\/p>\n McKinsey reports that<\/a>\u00a012% to 20% of cars could have advanced autonomous driving capabilities by 2030. For automakers, this means expanding the roads and environments that can safely support these capabilities. S-GNSS\u00ae Auto helps make that possible by improving GNSS reliability and laying the foundation for advanced vehicle-to-everything (V2X) and ADAS technologies needed to support autonomous vehicle safety in challenging urban areas. Working directly from the chip, it provides a cost-effective and accessible way for automotive OEMs to upgrade their technology via a firmware upgrade.<\/p>\n To see the impact of the integrated S-GNSS Auto and Teseo solution, download the latest data from our trials in Japan and Germany\u00a0here<\/a>.<\/p>\n