5G: Automotive challenges for the network of the future
Applications within the automotive space play an important role in the 5G networks that are supposed to appear around 2020. But besides the technical challenges, the telco and automotive industries may also experience a clash of cultures.
When the first 3G or “UMTS” mobile networks emerged, the telco industry had a powerful technology at their disposal, but it was still desperately looking for a “killer application” to boost demand for their new, fast and data-centric mobile network. This missing killer app then arrived in the form of smartphones and so subsequently intensified even the demand for the successive cellular network generation called 4G/LTE. But the telco suppliers and network operators had learned their lesson: When they started discussing the outlines of the 5th mobile network generation, their approach was almost reciprocal – first they would define the intended applications and then derive the required technology.
Nowadays, there is a broad consensus that 5G will be all about machine-type communications – particularly the Internet of Things and self-driving cars. In the automotive sector, 5G is a strong candidate for future car-to-car and car-to-infrastructure communications. Cars connected via 5G are supposed to offer functions like cooperative cruise control, high-density platooning, assistants for lane merging, overtaking and intersections, emergency vehicle warnings, traffic light advisories and many more. There is definitely no shortage of ideas how future cars could benefit from a high-performance, low-latency, ultra-reliable mobile network.
Technical versus organisational challenges
The technological implications of these plans are impressive. In order to facilitate on-the-spot reactions from the network (researchers are speaking of the “tactile internet”), latency will have to come down by a factor of 25 to 40 to a mere one millisecond – making even the speed of light a limiting factor and thus necessitating a whole new network architecture. Other basic points on the 5G target specification are data rates that may exceed 1 Gigabit per second depending on the respective application type, and minimal energy consumption with self-sustaining connected sensor modules achieving a battery life of up to ten years. These requirements are definitely bold and will pose severe challenges to the scientists and engineers developing the designated technologies. But we can safely expect that all these requirements will be met and all technical challenges can be solved.
What might however turn out to be a bigger challenge is the need to adjust and coordinate the fundamentals of the two very different industries involved. Much has been written about the deviating product and development cycles – with new smartphones appearing every six months while car makers take up to seven years to develop a new model. With the entry of mobile phones and connectivity into the cars’ cockpits, there has however been some adjustment – the telcos nowadays better understand the differing needs of car manufacturers, and automotive engineers have learned to speed up their processes when it comes to integrating the latest generation of mobile technology into an existing car line.
Business models still unclear
But the challenges involved with 5G go much further than that and touch vital aspects like the core business models. For example, between telcos and automotives there are very different views about the acceptable costs for future 5G communications. Although the telco players promise that the cost per bit is expected to decrease considerably, the deployment of their new networks will be very expensive. And if the licenses for the required radio spectrum will again turn out to be as costly as they were for the previous mobile network generations, a 5G contract can be expected to carry a considerable premium surcharge. This doesn’t go along well with the intention of an ubiquitous, fast growing and widely deployed 5G network that would be required to provide safety- and mission-critical connectivity to as many cars as possible.
Also, telco network operators speak about rolling out 5G first in metropolitan areas, as they expect to reach more paying customers there. Autonomous cars on the other hand will clearly not debut in the densely populated cities, but more likely have their first assignments on freeways and motorways with their straightforward traffic situations and comparably low traffic density most of the times.
The holy grail of functional safety
When 5G coverage is likely be scattered within the first years, this of course raises questions about network availability and reliability when it comes to vital car functions. And when a car leaves the coverage area of one provider and luckily meets coverage from another vendor – do telcos really expect an autonomous car to undergo a time-consuming roaming and handover procedure in these situations? This might bear some difficulties when for the according applications latencies within the magnitude of milliseconds are a given.
While these questions are already tricky in the starting phase of the first 5G networks, they might turn out to be even harder to answer a few years later, when the new technology will have matured to some degree. In an almost perfect world the 5G coverage might meet 99 percent on roads and motorways. What about the remaining one percent of dead spots?
“From my point of view, the telco industry is not really thinking about the functional safety, that is a key requirement in the automotive world”, says Dr. Osvaldo Gonsa, Director Mobile Communications ITS at Continental AG. “For example, all elements involved in the communications chain must be audited against the requirements of the ISO 26262 specification.” But how could such objectives be warranted in any kind of cellular network? Should cars fall back to peer-to-peer communications when they can find no network connectivity? If yes – why should they bother to reconnect to an extensive network later? This however poses telco providers with the ugly question whether they are needed at all as far as connected cars are concerned.
There still lies a great and promising future ahead when it comes to connecting cars with 5G mobile technology. But there is definitely a lot of homework to be done before. It is a good thing however, that the concerned parties have realized now that they might have more points on their agendas than originally expected – and that they have started discussing these somewhat different perspectives.