By Barry Manz for Mouser Electronics
Most people involved with the auto industry agree that it will be many years before every vehicle on the road does not need a driver, but you’d never guess that from the frenetic pace of automakers, the electronics industry, Tesla, Uber, Waymo, and many other participants. Regardless of how fast they run the race towards full vehicle autonomy, they’ll never get there without a standard protocol for wireless connectivity, and right now there is no formally-acknowledged, government-approved “winner.” And that’s a problem.
But wait, wasn’t that settled 20 years ago after the US government created the dedicated short-range communication solution that allocated spectrum at 5.9GHz just for this purpose? Well it was, at least until recently when an entire wireless industry entered the picture, proposing its networks and infrastructure as a better solution. No doubt, the industry waited until now—as wireless technology has only just become formidable enough to accept the challenge. So, what once seemed a fait accompli is now a fierce contest between the champions of each approach with the government waffling in the middle. What’s at stake is a market that will generate hundreds of billions of dollars for electronics manufacturers, the auto industry, and wireless service providers. To understand this situation, it’s important to know how it came to be.
Autonomous Vehicle Communication
In their latest models, all major vehicle manufacturers have loaded their products with safety features such as active braking, lane departure warnings, and blind spot detecting, but these features serve only to enhance the safety of the vehicle they inhabit. For the world to become fully autonomous, all vehicles will need to communicate with each other as well as with roadside infrastructures that specifically serve to detect pedestrians, bicycles, motorcycles, deer, and anything else that gets in their way. That is, this endeavor will have to be the largest Internet of Things (IoT) application there is, with the transmission of data traveling to and from sensors, satellites, and whatever wireless solutions emerge.
Every bit of this technology is necessary to ensure that vehicle autonomy delivers what it has promised, which is: Dramatically increased traffic safety, reduced congestion, higher efficiency, and potentially fewer cars on the road. For this to happen, vehicles must have the precise situational awareness to make comprehensive and accurate decisions in real time (Figure 1). This ability, in turn, requires wireless communications between vehicles and communication to and from roadside infrastructure such as sensors mounted on traffic signs, signals, and so forth. Consider this for a moment, and it should become fairly obvious that without connectivity, global vehicle autonomy will never see the light of day.
A Rare Moment of Clairvoyance
Federal agencies aren’t known for their ability to react quickly to change nor for their skill at predicting the distant future, but more than 20 years ago smart minds in the auto industry and the US government realized that people would be a lot safer if vehicles could drive themselves. Perhaps these thinkers got their cue from The Jetsons, but whatever drove them to the decision led the Federal Communications Commission (FCC) to allocate 75MHz near the unlicensed spectrum of Wi-Fi range for exclusive use by intelligent transportation systems.
The result was the dedicated short-range communication (DSRC) solution, which ultimately became an IEEE standard called 802.11p and a higher-layer standard within it called IEEE 1609, which together documented the mechanisms necessary for such a system to work. It quickly became the foundation of a European standard for a type of vehicular communication known as ETSI ITS-G5. Since that time, the auto industry, academia, and other interested parties have assumed that DSRC would be the standard everywhere.
Wireless connectivity must address all vehicular scenarios, collectively called V2X (i.e., vehicle-to-everything), in which there are the following components, all having different requirements (Figure 2):
- V2I (vehicle-to-infrastructure)
- V2V (vehicle-to-vehicle)
- V2P (vehicle-to-pedestrian)
- V2D (vehicle-to-device)
- V2G (vehicle-to-grid)
- V2N (vehicle-to-network)
It’s important to note that when DSRC was created, the cellular industry was in its second generation, and the only automotive application for wireless technologies was used for automatic toll collections, and the required technologies to serve vehicle autonomy did not exist. Today, cellular technologies can perform not only the functions prescribed by DSRC but in the eyes of the cellular industry at a cost less expensive, faster, with greater performance, and with a guaranteed roadmap for future improvements. Nevertheless, there are differences of opinion about this—the result of which has become the contested environment concerning vehicle connectivity.
The DSRC Perspective
Proponents of DSRC say that, because this standard has been around for two decades and the auto industry and government have spent tens of thousands of hours developing it, there is no logical reason to abandon it now. Manufacturers of chipsets have even developed solutions for DSRC, which will become useless if a cellular approach takes its place. Many also believe that, although much work is still undone, a DSRC solution is nearer to fruition than the cellular approach that has only recently just thrown its hat into the ring.
The pro-DSRC camp also makes the case that while the cellular industry says it can implement V2X today, the truth is that the cellular approach has been subjected to far less testing, and its full certification would require years of effort that the auto industry does not have. Although they agree that cellular technology can currently provide non-safety-related features, there is no proof that the technology will perform well in congested traffic environments. Likewise, they counter that the cellular industry’s approach has no funding mechanism for its ongoing operations, while DSRC apparently has no subscriber fees—which seems unrealistic as someone will have to pay for all these capabilities, and consumers usually wind up writing the checks.
DSRC was also designed to have low latency, which is a crucial metric for vehicle autonomy scenarios in which success or failure to deliver data and make decisions are a measurement of milliseconds—although even DSRC stalwarts concede that the fifth generation (5G) cellular network solution coming in the next two or three years should solve latency problems. Finally, there is the issue of cellular network reliability during catastrophes such as hurricanes, which hasn’t proven to be universally robust. However, one could probably say the same of DSRC, but considering it is not yet in the deployment phase, no one currently knows.
The Cellular Perspective
In the other corner populated by the cellular industry, some device manufacturers (most notably Qualcomm), and an increasing number of automakers believe that even current cellular technology and networks can meet most of the requirements for vehicle autonomy. And unlike DSRC, the cellular high-speed-data infrastructure is already massively deployed. The cellular industry finds claims of its absence of a clear roadmap laughable, as cellular technology has advanced consistently since the late 1980s and, by necessity, wireless carriers must always operate with continual innovation in mind.
The cellular industry also believes its solution will be far less expensive to implement than DSRC, as the latter requires massive amounts of new infrastructure and (of which cellular supporters claim) is incapable of scaling to accommodate increasing traffic demand. And considering DSRC was created decades ago, it has become an archaic relic compared to the cellular technology that has passed through four generations and is about to unleash massive new capabilities through its 5G advancements and innovations.
Perhaps most worrisome, the cellular industry claims, is that DSRC is a government-sponsored effort—meaning that there’s no assurance of consistent funding. Although this is an excellent point, it applies equally to cellular as well, because they need to make money, and whether that comes from fees or taxes, the consumer will still be on the hook for at least some of the operational costs.
But possibly the most lethal weapon in the cellular industry’s arsenal is the fact that DSRC requires the use of roadside units (RSUs), none of which have been deployed, and such a widespread deployment would cost hundreds of millions, if not billions, of dollars and would by most estimates take up to 15 years. RSUs are one of the two main components in DSRC networks, and the other component is on-board units (OBUs) within vehicles that have integral transceivers. An RSU communicates with OBUs to acquire traffic information such as time, speeds, vehicle locations, and other metrics. Placed in an area of congestion or an intersection, for example, an RSU calculates vehicle travel times and congestion origination times and sends this data to all OBUs within its range.
As for the issue of service outages, the cellular industry says it can effectively deal with them using device-to-device (D2D) technology that allows communication to take place between users without having to first pass through a base station. Although data rates drop precipitously when this is enabled, this shouldn’t be a concern, as almost all core, safety-centric V2X capabilities need only low speeds to function.
As their parting shot, champions of the cellular approach note that while the US is vacillating about standards, the rest of the world has moved on, and in Europe and Asia, the choice is cellular. For example, a consortium including all major German automakers as well as major infrastructure providers, such as Vodafone, Ericsson (now part of Flex Power Modules), Intel, Huawei, and Nokia, believe cellular technology is the best route. If the US chose one version of vehicle connectivity and Europe, Asia, and elsewhere adopted another, this alternative would place a burden on all automakers that would need to build country-specific connectivity solutions into their vehicles.
Cellular V2X Defined
Cellular V2X (C-V2X) embeds into vehicular environments the latest technologies in cellular communications, which include downlink speeds of about 100Mbps, low latencies (i.e., fast round-trip signal-response times), carrier aggregation to increase channel bandwidths (and thus speed), multiple-input multiple-output (MIMO) and beamforming, and many other features. It will operate in the 75MHz of spectrum with 5.9GHz dedicated to intelligent transportation systems and using other frequencies to complement network performance. This range of spectrum is essential because signal propagation at 5.9GHz covers shorter ranges and is more affected by radio-frequency (RF) inhibiting obstructions (as are those at 3GHz and below).
However, its two most important attributes are its harmonization with more than 800 wireless carriers throughout the world and its evolution from 4G to 5G capabilities. The first attribute allows automakers and suppliers to design and build according to a single set of standards, and the second characteristic ensures that this cellular approach has a technology roadmap within the Third-Generation Partnership Project (3GPP), the global organization that manages the development of mobile communications standards. Thus, C-V2X will become essentially future-proof because of these two primary attributes. As for DSRC, on the other hand, there is no “next-generation” plan in development.
The fifth generation of cellular standards (5G) is the most inclusive, future-focused generation since the inception of the cellular industry, with the final specifications released in June 2018. Rather than simply increasing network speeds and adding a few new features, 5G spreads the use of cellular technologies to new applications, including vehicle autonomy and IoT, and uses an entirely new network architecture called 5G New Radio (NR) that makes it far easier for networks, infrastructure, and user devices to evolve with advances in technology.
5G standard capabilities can deliver greater range, better non-line-of-sight performance, a lower packet error rate (PER), a higher capacity, and a greater ability than DSRC to operate in spectrums densely populated by wireless devices. Round-trip latency over short distances may be as low as a few milliseconds compared to tens of milliseconds today, which is essential for real-time applications such as vehicle autonomy.
There are two basic transmission modes within the C-V2X approach: Direct access (also called sidelink) and vehicle-to-network (V2N). Direct access allows vehicles of any type, including motorcycles and “enabled” bicycles, and infrastructure to communicate with each other independent of a cellular network connection. The second mode, V2N or cellular vehicle-to-network (C-V2N), allows vehicles and infrastructure to communicate with each other over the traditional cellular network to provide near-real-time information about road conditions and traffic. It can also be useful for applications like infotainment that is not related to safety.
Together, these modes can eliminate the disruption of service, which is another mandatory requirement for vehicle autonomy. As many vehicles already incorporate cellular connectivity, others will surely follow. Adoption of the C-V2X approach will allow vehicles to incorporate interactive communication features as well as to stream or download high-definition (HD) media.
Both DSRC and C-V2X rely on RSUs. However, DSRC requires them while C-V2X can benefit from them. This difference is important. The small-cell base stations that wireless carriers will deploy in the upcoming years, to bolster coverage, can also be useful as RSUs deliver information from the roadside infrastructure and possibly to backhaul network traffic. So, as the cellular industry is committed to deploying them anyway, using them for vehicle autonomy purposes would potentially cost less than implementing the huge number of RSUs that DSRC will require.
In the last two years, automakers have shifted their allegiance from DSRC to the C-V2X approach en masse, taking most of the companies involved in developing vehicle autonomy with them—making the likelihood that the C-V2X approach will trump DSRC almost a certainty. All this has occurred while the US government and the remaining DSRC supporters cling to what little hope they have of maintaining the status quo.
The US Department of Transportation (DOT) and the National Highway Traffic Safety Administration (NHTSA) seemed ready to move forward last year, with a notice mandating that all vehicles integrate vehicle-to-vehicle (V2V) communications technology by 2023. Although V2V is just a piece of the vehicle autonomy puzzle, the agencies got their wings clipped by the Trump administration in November when it decided not to pursue such a rule, presumably letting the industry decide instead.
Since then, the FCC and other agencies have dialed back the intensity of their commitment to DSRC. For example, at a recent event in Washington DC, FCC Commissioner Jessica Rosenworcel recently said, “I don’t think we should strand our thinking about safety technology and spectrum ideas from 20 years ago,” while another Commissioner, Michael O’Reilly, said, he thinks issues surrounding 5.9GHz “are a mess.” In addition to statements like this, the agency has recently been veering into the woods, expressing enthusiasm for the use of spectrum sharing between DSRC, Wi-Fi, and other entities to develop innovative wireless solutions.
The takeaway from this is that the government seems to be waiting for someone else to decide for them, and the cellular industry has been more than happy to do that. In fact, it’s safe to say that C-V2X has already won this battle, as it has amassed a tsunami of marketing resources dedicated to cementing its position.
So, projecting C-V2X as the winner is less of a bold prediction than it is a statement of fact. In short, C-V2X will be the standard for V2X, and it will use the spectrum dedicated to DSRC while also using standard cellular frequencies for complementary “non-safety” applications, empowering full vehicle autonomy to accelerate forward at high speed. Once C-V2X becomes unavoidable, the government will likely then make its decision.