Networked and, to an even greater extent, autonomous driving generates huge amounts of data that cars have to exchange in fractions of a second. 5G can help here.
Fewer accidents, fewer traffic jams. Sleep, watch TV or read instead of accelerating, changing gear and steering. This is what the future of driving should look like. Automobile manufacturers, technology groups and research institutes have been working for years on technologies that make automated driving possible. The car of the future will use three data sources to capture what is happening around it: built-in sensors, direct radio communication with nearby vehicles (vehicle-to-vehicle communication, V2V) and mobile radio communication with more distant vehicles. Today’s driver assistance systems, such as the emergency brake or lane departure warning system, rely exclusively on sensor data. So is a mobile communication connection actually necessary at all for safe, partially and fully automated driving?
“Self-driving cars must be able to do without communication. But it’s very helpful,” says Christian Wietfeld, who heads the Department of Communication Networks at the Technical University of Dortmund and has been researching the use of mobile communications technology in road traffic for years. According to the technological research department of Lucid Motors, self-driving cars will produce up to 19 terabytes of data per hour in the future . This will require extensive bandwidth, which can reliably transmit data for safety-critical functions in real time. The networked car communicates with its environment via C-V2X technology (Cellular Vehicle-to-Everything). With V2V, vehicles can coordinate their movements and actions with each other, warn pedestrians of possible risks (V2P, Vehicle-to-Person) and communicate with traffic signs (V2I, Vehicle-to-Infrastructure) in order to know who has right of way. Audi, for example, is networking its vehicles with the traffic light infrastructure. The “traffic light information” service shows the driver in the cockpit the speed at which he will catch the next green phase (GLOSA -Green Light Optimized Speed Advisory). If this is not possible at the applicable speed limit, a countdown is used to show when the next green phase will start (time-to-green).
Sensor technology is too shortsighted
According to Statista and Roland Berger, up to thirty percent of cars worldwide now have a connection to the mobile network, and the trend is rising. And according to Statista Digital Outlook Connected Car. the number of networked vehicles will increase drastically. According to the study, more than 350 million networked cars will be on the road worldwide by 2023. At the moment there are around 157 million. However, cameras, radar scanners and ultrasonic sensors only provide data about the vehicle’s immediate surroundings. Obstacles further away, such as the end of the traffic jam one kilometer away or ice behind the next bend, are therefore detected relatively late by the car. The reaction time is correspondingly short – especially at higher speeds. Even if cars are running fully automated or with a high degree of automation, situations are to be expected that will quickly bring algorithms to their limits. In such a case, the driver must take the wheel again.
5G has a lot of potential
Behind the term 5G lies not only a new, fast radio protocol (5G New Radio), but also new features in the mobile radio core network. “5G is a complete ecosystem consisting of many sub-components,” says Michael Einhaus, professor at the University of Applied Sciences in Leipzig. For example, Mobile Edge Computing shifts the data analysis from a distant cloud to the vicinity of the base stations, and thus to the vicinity of the vehicles. If two cars exchange data via the mobile network, the data transmission distance is significantly shortened – and thus also the transmission time (latency). Network slicing, in turn, makes it possible to create several virtual networks on the same antenna. As a result of this, data for a driver assistance system can be transmitted with a specified level of quality via a dedicated network, without being disturbed by other data streams.
More bandwidth, lower latency
Compared to LTE Advanced (4G), which is currently the fastest LTE version, 5G offers will offer significantly higher bandwidths and less delay in transmission between the mobile base station and the car. 5G also transmits data at a faster cycle rate than LTE. 5G developers save additional time by simplifying data encoding and accelerating decoding. Because of this, the 5G radio standard should be able to provide one million networked devices per square kilometer with connectivity. Cars will also benefit from this capability when networked vehicles become more numerous and the capacity of a radio cell has to be shared with other participants, such as wearables, networked bicycles and smart meters.
LTE will not disappear
However, today’s mobile networks can only be converted gradually to 5G. The current LTE Advanced radio protocol (4G) will thus remain an integral part of the 5G network. Many scenarios of partially and fully autonomous driving already work with LTE radio communication. If, for example, a vehicle is to be warned of the end of a traffic jam, transmission via LTE Advanced with latency of less than ten milliseconds is sufficient. Even if the data is still being processed in a central cloud and the transmission time increases to one second, a car traveling at 130 km/h will only travel 36 meters during this time. The situation is different, however, with direct radio communication between vehicles (V2V), as this takes place over short distances of a few hundred meters. Low latency is important here. Example, platooning: Several trucks drive behind one another with a distance of just a few meters between each vehicle in order to save fuel and space. “In this scenario, I need highly reliable communication,” says mobile communications expert Wietfeld. According to Wietfeld, if an obstacle appears, the leading truck in the platoon must pass this information on to the other vehicles in milliseconds. No problem for the 5G mobile radio communication, which is supposed to achieve latency rates of less than one millisecond.
Over the long term, the sensor data collected from cars, bicycles, pedestrians and the traffic infrastructure will be used to create a complete digital image of the road network. This means that every road user will know at all times where vehicles and people are in his vicinity, where accidents have happened and where there are traffic jams. They will even be provided with information when someone suddenly walks onto the road. Only in this way will it be possible to realize autonomous driving and “Vision Zero” – road traffic without any fatalities. The mobile radio network of the future will thus unite a colorful bunch of new technologies. “5G is like a Swiss army knife,” says automobile and mobile communications expert Johannes Springer, who heads the 5G program for the automotive sector at Deutsche Telekom. “We will use a different tool, based on what we want to do.”
Recent Comments