Every time I write about the increasingly intelligent vehicles that automakers are putting on the road, Metcalfe's Law is also in the back of my mind. Originally formulated by George Gilder a quarter-century ago, it was later attributed to (and named for) Bob Metcalfe in regard to his co-founding of Ethernet. Wikipedia's summary of the Law is as follows

The effect [editor note: i.e. the power] of a telecommunications network is proportional to the square of the number of connected users of the system (n2).

And, particularly germane to this blog post, the Wikipedia entry goes on to say:

Metcalfe's law was originally presented, c. 1980, not in terms of users, but rather of "compatible communicating devices" (for example, fax machines, telephones, etc.). Only later with the globalization of the Internet did this law carry over to users and networks as its original intent was to describe Ethernet purchases and connections. The law is also very much related to economics and business management, especially with competitive companies looking to merge with one another.

I'll claim here that vehicles also qualify as "compatible communicating devices" and, as such, are candidates for enhancement courtesy of a wireless network interconnecting them. Imagine, for example, the benefits that would accrue from vehicles telling other vehicles around them that they've been delayed by a just-occurred accident (either their own or one in proximity to them), and that those other vehicles should accordingly re-route themselves around the bottleneck. Or imagine, for fully autonomous vehicles, the ability to completely eliminate traffic lights in favor of inter-vehicle communication that enables safe and smooth transit through intersections for all vehicles approaching from all directions. I'm sure you can come up with plenty of other candidate scenarios.

Admittedly, such functions could alternatively be implemented, at least to some degree, via cameras mounted at intersections (which will likely exist nonetheless), for example. But I'd argue that the source data is both more accurate and more rapidly perceived and disseminated when it's sourced from affected vehicles themselves. And of course, the data won't be sent solely to other vehicles, as the Wikipedia entry for "Vehicle-to-everything" notes:

Vehicle-to-everything (V2X) communication is the passing of information from a vehicle to any entity that may affect the vehicle, and vice versa. It is a vehicular communication system that incorporates other more specific types of communication as V2I (Vehicle-to-Infrastructure), V2V (Vehicle-to-vehicle), V2P (Vehicle-to-Pedestrian), V2D (Vehicle-to-device) and V2G (Vehicle-to-grid). The main motivations for V2X are safety and energy savings. The main obstacles to its adoption are legal issues and the fact that, unless almost all vehicles adopt it, its effectiveness is limited.

Personally, I'd add "privacy" to the main-obstacle list, since V2X-transmitted data can conceivably include information such as where you are, where you've been (and when), and how fast and otherwise aggressively you're driving. Then again, however, I've admittedly been bewildered at the popularity of today's GPS- and OBD-II-based telematics devices, such as Progressive's Snapshot and alternatives from other insurance providers. Perhaps, sadly I might add, I've underestimated consumers' willingness to dispense with privacy for discounts on their policies and/or other perceived benefits.

Privacy aside, if you buy into my premise that a wireless network interconnecting vehicles with each other as well as with the broader Internet has value, the obvious next questions involve how such a network should be implemented; what protocol should be employed, and what spectrum swath should it occupy? Until recently, the answers to those questions seemed settled; back in the late 1990s, the U.S. FCC (Federal Communications Committee) allocated 75 MHz of spectrum in the 5.9 GHz ISM (industrial, scientific and medical) band for ITS (intelligent transportation systems) uses.

The approach, specifically referred to as DSRC (dedicated short range communications) in the U.S. (Europe, for example, is working on a conceptually similar approach which allocates 30 MHz of spectrum in the 5.9 GHz band), overlaps spectrum already in use for military, satellite, and amateur radio services, along with well-known 5.8 GHz Wi-Fi. Messy? Yes. The approach is conceptually akin to ad hoc Wi-Fi, with no centralized router-like infrastructure required and connections between vehicles and other nodes (vehicles, roadside transmitters and receivers, etc) set up (and subsequently torn down) "on the fly" based on proximity and functional need.

However, as regular EDN readers already know well, cellular data networks have increasingly come to the fore in recent years. In addition to their primary use with smartphones, cellular-enabled tablets and other mobile devices, they also service so-called "fixed broadband" setups ... as well as other mobile platforms, conceivably including vehicles (here, to be clear, I'm referring to the vehicles themselves, not to the smartphone-toting occupants of those vehicles). Tailored variants of both 4G LTE and upcoming 5G cellular have been developed that support specific combinations of bandwidth, latency, range, premises penetration, cost, power consumption, and other characteristics, in targeting various applications. For ITS purposes, unsurprisingly, they're referred to as C-V2X (cellular vehicle-to-everything) setups.


Source: Qualcomm

Who's going to win this technology tug-of-war? The answer to that question is quite unclear at the moment, and "neither" is a potential outcome ("both," while also possible, would IMHO be unfortunate, from the standpoint of cost, complexity, and other concerns). DSRC arguably has the early lead, with initial deployments already on the road and heavy investments already made by some automakers, along with their chip, subsystem, software, and services partners. Those investments were a reaction to Obama administration-era mandates regarding DSRC deployment in order to ensure a timely critical mass of supportive equipment; more recently, however, the subsequent Trump administration has backed off the industry pressure. And C-V2X backer Qualcomm (unsurprisingly), along with key partner (and former DSRC advocate) Ford, plans to conduct initial testing of C-V2X "down the street" from me in Denver, CO this summer.

At least one market analyst has recently claimed, in fact, that the ever-increasing intelligence within the autonomous vehicle will make the need for any sort of ITS unnecessary (you can see his entire talk here).



While I personally wouldn't go quite that far, I concur with Lux Research's Mark Bünger that as individual network nodes (i.e. vehicles in this particular case) become smarter, the amount of data they need to interchange will likely lessen (i.e. bandwidth), as will the required rapidity of both proactively providing that data and responding to requests from other nodes (i.e. latency). And I also can't discount the conceptual advantage of building out a network using minor variances of already widespread technologies (i.e. cellular data service) versus more proprietary alternative approaches.

Readers, where are you placing your bets? Sound off in the comments.


Brian Dipert is Editor-in-Chief of the Embedded Vision Alliance, and a Senior Analyst at BDTI and Editor-in-Chief of InsideDSP, the company's online newsletter.


Related articles: