5G or not 5G?

Just a few years ago, there was only one solution in terms of communications protocols for delivering vehicle connectivity. Now, road operators and vehicle manufacturers face choices – including a moral choice, perhaps. Jason Barnes looks at the current state of play

There is a debate raging in the ITS world over future communications protocols. 750 Asfinag, Austria’s national strategic road operator, has announced it will from 2020 be using ITS-G5 to support cooperative ITS (C-ITS) applications (‘First things first’, 1846 ITS International January/February).

Asfinag’s Manfred Harrer and Peter Meckel highlighted several reasons for their organisation’s protocol choice. These included: the continued absence of a defined standard for 5G which will support vehicle connectivity; the time it will take from the current point to realise fully tested and deployable systems based on 5G; and a desire to get on immediately and improve roads’ safety using an already-matured solution.

But 5G enjoys a considerable groundswell of support. Whether it be for machine-to-machine connectivity in Industry 4.0 manufacturing applications, or cellular Vehicle to Everything (C-V2X), a moniker coined to provide differentiation from non-cellular vehicle connectivity solutions), articles in both the mainstream and specialist media declare it to be the protocol to use.

A powerful coalition of telecoms providers (telcos), automotive companies and chipset manufacturers is the prime mover behind this - but there is the contention that 5G does not yet exist in any substantive form. As Harrer and Meckel noted, a published standard which will support vehicle connectivity is not due until the end of this year. From there, we will still need testing, certification and products.

The ITS sector has spent many years developing a bespoke solution for communication between vehicles, infrastructure and nomadic smart devices. This public sector-driven undertaking has involved all manner of stakeholders and cooperation from the regional up to the international levels. This solution — ITS-G5, also known variously as 802.11p or dedicated short-range communications (DSRC) — is the solution adopted by Asfinag.

Critics of the exhortations to wait for 5G point to some commercial organisations’ desire to capture market share. As well as causing delay, they say, cornering the market effectively puts a price on safety.

The reality may be rather less clear-cut.

Defining 5G’s future

In Europe, significant effort is going into fostering 5G to 1816 European Union member states’ strategic advantage. The EU’s 5G Action Plan (5GAP) has applications across many vertical markets, including mobility.

Through 5GAP, the European Commission (EC) has set ambitious connectivity targets, starting with a coordinated launch of 5G in all EU member states by the end of 2020 and then a rapid ramp-up of infrastructure deployment by 2025 to ensure full urban coverage as well as uninterrupted 5G coverage along main roads and railways.

The 374 Ertico-coordinated 5G MOBIX project, launched last November, is integral to 5GAP and brings together 46 partners in cross-border and urban corridors across EU countries. It uses technological innovations to evaluate the benefits of 5G and will “define deployment scenarios, identify and respond to standardisation and spectrum gaps as well as understand the critical scenarios that require the advanced connectivity provided by 5G, and the features to enable advanced Cooperative, Connected and Automated Mobility [CCAM] use cases”. 5G MOBIX will include cooperation with China and Korea.

Also in November 2018, Ertico made submissions for funding under the Horizon2020 programme which will look at the needs of member states in the field of connectivity.

“This is because member states are already invested — ITS-G5 is already being deployed in some places in support of some Day 1 and Day 1.5 C-ITS applications but there is a sense that it is not enough in terms of what is required,” says Ertico’s platforms coordinator, Eusebiu Catana.

“ITS-G5 development was triggered by the public sector especially for road safety applications. 5G development for mobility applications is being triggered by industry to address identified shortcomings. With ITS-G5, there are latency and cybersecurity issues and some cross-border issues are not resolved. Legislative progress has been slow and there are missing guidelines. This affects interoperability between stakeholders. There are also net neutrality concerns.”

Catana does not expect 5G to replace all connectivity technologies immediately. “There has to be a guarantee that 5G can answer all users’ needs, be they cities, drivers or service providers,” he states. “Less than 30% have yet launched 5G strategies. Nevertheless, I think that we can expect to see investments made by frontrunners such as Germany, Sweden, Holland, Belgium, Spain and France bear fruit by 2020, with further developments across the EU by 2025.”

The 324 US Department of Transportation (DoT), meanwhile, is adopting a technology-agnostic, market-driven approach to vehicle connectivity. There, DSRC is a commonly-understood entity and already deployed in 26 states. US DoT has conducted a survey which invited comment from state DoTs and other relevant stakeholders as to statuses and needs in terms of vehicle connectivity. This was due to be concluded by late January 2019, but the longest federal government shutdown in history has caused delays.

Non-binary choice

From a technical aspect, Knut Evensen, 108 Q-Free’s chief technologist and chair of CEN/TC 278’s Working Group 17 (Urban ITS), challenges assertions that ITS-G5 is incomplete.

“There’s no doubt that 5G will eventually be faster and better than ITS-G5 — once it’s working. And it will work, given the resources being thrown at it, but that’s no reason to stop progressing ITS-G5. It’ll be at least three years until we have something 5G-based which is usable on-road, and around four to six years before chipsets are available in sufficient quantities — and by then I predict that new actors will be up on the stage claiming we should wait for the upcoming 6G technology. That, they will claim, is almost ready to roll out, much cheaper, faster and better.”

An issue which needs to be addressed is what people mean by ‘5G’. “There are three different technological concepts and that colours the argument,” Evensen says. “The first would use mobile telephony base stations to set up and control operations, and such a concept is already here as part of the PC5 interface for C-V2X connectivity. The second would do away with roadside stations and cellular towers. That not only doesn’t yet exist, it’s also damned difficult to achieve as there are timing issues when switching between different modes; access to the GPS clock for all stakeholders may be needed to address this. The third solution would use additional roadside systems and would need the ability to switch between centralised and local control. That adds complexity when trying to allocate time slots to different applications.”

ITS-G5, by contrast, is a broadcast-type system which uses the Aloha protocol, he explains. “This is ‘old’ technology but simple, cheap and robust — it was designed to survive a nuclear war. The solution is not to use either ITS-G5 or G5 but to have both, and this can be done without problems.”

In many respects, the real issues are non-technical, he continues. “The telcos want to get their hands on the 5.9GHz spectrum space currently reserved for safety-related ITS applications. Frequency spectrum is very expensive at auction and the telcos want that part to remain unused until they’re ready. They’ve been offered a huge chunk of spectrum at 3.4-3.8GHz but they’re claiming that they need 5.9. The underlying reason? It’s free.”

However, there is another matter to consider. “US university research states that if we install ITS-G5-based systems now, instead of waiting another three years or so for 5G, 40,000 lives will be saved in the US alone,” Evensen adds. “A similar figure would be saved in western Europe and more still around the world. The calculations involved are complex but they boil down to a simple moral question: are income streams and patent revenues worth more than saving 100,000 lives?”

Migration paths

Evensen also strongly objects to assertions that ITS-G5/5G is a binary choice. Like Harrer and Meckel, he notes that making a technology choice now does not preclude change in the future — indeed, in Australia his parent company has already demonstrated a hybrid solution which uses 4G.

Ertico’s Catana highlights EC mobility projects which will provide a migration path to 5G. Concorda, for instance, is addressing the issues relating to hybrid solutions where long-range cellular can complement short-range communication technology-based systems, and C-Roads Platform is a joint initiative of EU member states and road operators for testing and implementing C-ITS services in light of cross-border harmonisation and interoperability.

So, are roads operators faced with a stark choice and possible technology orphans? As Evensen notes, technology is a sliding window and perceptions of what exactly constitutes ‘5G’ vary.

The latter point is picked up by Maged Zaki, director, technology and product marketing with 213 Qualcomm — one of the companies which, some suggest, is attempting to cause delay in order to gain market share.

“Qualcomm takes a pragmatic approach aimed at getting V2X safety applications in vehicles and on the road as soon as possible — and in a size, weight and cost form which is acceptable to the auto manufacturers,” Zaki insists. “That helps to address the question of who pays for deployments, as DSRC/ITS-G5 is not the most cost-efficient way to add V2X into vehicles and road infrastructure.”

Qualcomm was first to market when in 2017 it announced a chipset based on the Release 14 C-V2X specifications. “This is 4G LTE-based — not, as some assume, 5G but it does provide a strong evolutionary route to 5G. And we’re not attempting to corner the market — since our announcement at least four other companies have produced or signalled their intention to produce chipsets,” he continues.

The C-V2X solution uses the latest LTE wireless coding and waveform to establish peer-to-peer dialogues. According to Zaki, it does not require communications towers and works perfectly well along road corridors without network coverage or SIM subscription.

“Testing in around 20 locations in the US, Europe and the Far East has proven the effectiveness of our offering,” he says. “Essentially, we’ve kept the upper layers of the DSRC solution but replaced the 802.11p-based radio with a much more effective communications technology. This achieves higher reliability and two to four times the communications range, including non-line-of-sight.” That means, for instance, that cadence braking can occur over greater distances, including round corners and in congestion.

“From our standpoint, there’s nothing to stop deployment of both ITS-G5 and 5G,” Zaki continues. “278 Ford committed at CES 2019 to have C-V2X fitted in its vehicles by 2022. We encourage maintaining technology neutrality and keeping the market open. What’s important is saving lives and the ability to develop the paths to future applications such as autonomous driving.”

Standards progress

C-V2X, says Maxime Flament, chief technology officer with the 5G Automotive Association (5GAA), uses a series of fundamental principles embedded within 4G LTE to provide forwards-compatibility with 5G. He echoes Zaki’s comment with relation to taking out the ‘old’ radio and inserting the ‘new’.

“There’s a definite DSRC legacy, in that almost all the upper layers are being re-used,” Flament says. “A lot of work has gone on over the last year with ETSI to make those layers technology-agnostic. That’s included a lot of input from the telcos. 5G is very much in the making. 5G NR [new radio] was standardised in Release 15 in June 2018 and it means that already we can do lots of the things promised by the availability of massive broadband — the Internet of Things [IoT], and so on. With Release 16 at the end of this year, additions will include NR V2X. That will give us ultra-reliable low-latency connections. NR V2X will eventually bring new capabilities such as advanced and automated driving but it is just one part of the technological puzzle.”

Other parts of the jigsaw, he notes, include C-V2X, as standardised within Release 14, and narrowband-IoT (NB-IoT). The latter is a low-power wide-area network (LPWAN) radio technology standardised by the 3GPP. It is realised by upgrading 4G LTE base stations to enable them to provide a large number of narrowband communications channels which can be used to support a wide variety of smart city-type applications.

“Backwards compatibility with DSRC would mean the addition of more technology, so we advocate interoperability at the service level and relying upon the mobile networks’ zero latency to bridge the gap,” Flament continues.

In terms of timescales for deployment, he sees Ford’s announcement as a landmark but adds some caveats: “Ford has said 2022 but this is a stepwise approach as the company will equip all of its vehicles by 2020. This is huge: its US market is about 2.4 million new vehicles per year. We can expect to see similar announcements to Ford’s and a well-recognised trickle-down of technology from high-end vehicles. But there’s a big difference between who’s first and who’s having an effect — mass is critical to making all this work.”

Looking forward

Just a few years ago, the idea that the solution to vehicle connectivity could be anything other than something DSRC-based would have seemed absurd. However, prolonged gestation had already seen much evolution within the ‘DSRC-only’ world itself. The technology solutions had become far leaner and, realistically, that had to happen as the costs of deploying roadside beacons every few hundred metres along every road in the world made themselves more apparent.

Using mass-market protocols and infrastructure is another natural progression, as we are already seeing with the use of LPWAN protocols such as NB-IoT for mobility-related applications such as on-street parking. Being able to use cellular networks for vehicle connectivity is altogether neater from a number of perspectives.

The reality is that we are still not there yet but what we are starting to see is the emergence of true choice. ITS-G5 proponents may stand accused of trying to jealously guard market share, in the sense that they are quick to point out that 5G is still some years away - but at the same time they recognise the need for migration paths and are actively working towards these. 5G development, meanwhile, continues apace.

A point is worth making about market penetration. Flament describes Ford’s announcement as ‘huge’ but the figures bear further scrutiny. Ford enjoys 14.2% of the US market (Statista, 2017 figures) but it is equal second with 1686 Toyota — which, like GM, has put its weight behind ITS-G5 — and 2.4 million vehicles per year is less than 1% of the US’s total fleet of around 276 million.

Support for 5G will surely grow — BMW, for instance, has aligned itself with the cellular technology — but divisions remain. 2453 Renault has chosen to run with ITS-G5, while other marques are hedging their bets — 994 Volkswagen has gone with 5G in the US, and ITS-G5 in Europe.

This means that ubiquity is going to take time whichever connectivity solution is chosen, and having the technology ready - or even starting to be deployed - does not mean that it is ‘working’ at a level which can make an immediate and appreciable difference or signify market dominance.

The main point which emerges is the consensus that deploying connectivity now does not close the door on the future, and any fears over whether that may be so are unfounded. The question for roads operators, whether public or private, is when they want to get on and use technology to further improve their roads’ safety and environmental performance.