Which is More Important for Spectrum Policy: FCC or DOT?

Which has more influence on spectrum policy: the Federal Communications Commission (FCC) or the various federal, state, and local departments of transportation? Conventional wisdom would place the FCC as the hands down winner. Surely, that continues to be the case. But if current technological trends continue, the departments of transportation, especially the United States Department of Transportation (USDOT), could end up a strong second. We're all familiar with the growing importance of municipal Wi-Fi. But why not highway Wi-Fi? Why should Wi-Fi stop at city boundaries?

That's where the USDOT's planned Intelligent Transportation System (ITS) could come into play.

In 1999, the FCC allocated 75 MHz of dedicated spectrum for ITS, and in December 2003 it adopted service rules for the band, called Dedicated Short-Range Communications Services (DSRC). See FCC Report & Order 03-324A1, posted to the Federal Register in August 2004.

DSRC works at very short ranges--under 100 yards, or about the same distance as WiFi--and can be used for communications between vehicles and roadside terminals or between vehicles that are close together.

The ITS band, located at 5.850 GHz to 5.925 GHz, is adjacent to the 5 GHz unlicensed band and has many of the same service characteristics of that band. Indeed, the basic standard for the band, 802.11p (also known as the Wireless Access for the Vehicular Environment [WAVE] standard), is derived from the 802.11 family of standards in the unlicensed bands.

You might ask why does USDOT need a dedicated short-range communications services band of spectrum for ITS when unlicensed spectrum could have done the same job? Such a question could easily have been dismissed in 1999, when the WiFi revolution had not yet begun. But, at least on policy grounds, it is harder to dismiss today.

Consider this fact: roads cover a lot of the United States but the cumulative territory they cover is less than 1% of U.S. land surface area. Why shouldn't other people be able to use that spectrum locally when it wouldn't otherwise be used? Similarly, only public safety vehicles are likely to demand continuous high bandwidth use of the ITS network. When the roadside network is not being used for transportation related information, why shouldn't it be used to transmit other types of information?

The key argument used by advocates for dedicated ITS spectrum was that when lives are at stake, you don't fool around with unlicensed spectrum. It is the same type of argument the public safety community used to win 50 MHz at 4.9 GHz for a similar short-range Wi-Fi-like service. Service rules for the 4.9GHz band were adopted in 2004, shortly after the ITS band rules.

Unfortunately, complaining that the 75 MHz is inefficiently allocated only to transportation is like complaining over spilled milk. I don't believe there is anything practical that can be done about it.

But this still leaves two other very big policy issues on the table. First, if the government spends $3 to $10 billion dollars developing this ITS network beginning around 2009, why doesn't it add Wi-Fi functionality on the adjacent 5 GHz unlicensed band for a very modest additional expenditure?

Second, the entire world is moving to short-range wireless networks because that is the only way to achieve super high speed broadband speeds with high quality of service. For example, that's why 3G mobile telephone networks, where towers are often spaced a mile or more apart, use more than twice as much spectrum but can only achieve speeds less than 5% the speed of Wi-Fi networks. It's why, for example, the 18,000 light posts in New York City have suddenly become such valuable telecommunications real estate. So why should we be banning exactly the type of roadside broadband network architecture for mobile Internet service that clearly uses the spectrum most efficiently and is the most technologically advanced?

Intelligent Transportation System

The Intelligent Transportation System (ITS) is actually a loose conglomeration of an amazing variety of different technologies that often have very little to do with each other and range from very low tech (such as highway EZ Pass lanes and electronic overhead weather-traffic alert signs) to very high tech (such as the science fiction scenario of cars and roads so intelligent that cars don't need human drivers). Much of the excitement about ITS consists of technologies at an intermediate level of sophistication. This includes applications that alert human drivers of dangerous conditions (e.g., drifting across lanes without a turn signal on) and activating machines without human intervention in highly circumscribed conditions (e.g., activating an air bag microseconds before two cars will crash into each other). Two helpful overviews of ITS can be found at http://www.its.dot.gov and http://www.itsa.org.

ITS advocates believe ITS can reduce the expenses associated with transportation accidents involving injuries or fatalities ($260 billion/year) and traffic congestion ($100 billion/year). The goal of ITS is zero highway fatalities and delays. The ITS literature is filled with factoids (such as 88% of rear-end accidents are caused by drivers who fail to concentrate, or get too close to the car in front, and in 2003 alone, 8,569 people died and more than 1.4 million suffered injuries as a result of intersection-related crashes).

Vehicle infrastructure Integration (VII) Initiative

The part of the ITS system that I believe is most interesting from a highway Wi-Fi perspective involves the architecture of the DSRC network and specifically what has in the past several years come to be known as the Vehicle infrastructure Integration (VII) initiative.

In its initial stages the VII network will not be ubiquitous. Indeed, the advocates of ITS have gone out of their way to assure the telecommunications industry that it will be a crippled network with highly limited coverage and service applications. But even then, it is anticipated to include about a half million hot spots centered at traffic intersections and other accident prone stretches of road such as sharp turns and bridges especially vulnerable to icy conditions. Some of the applications suggested for the VII network include intersection collision warning, curve speed warning, rollover warning, emergency brake warning (from nearby vehicles), and vehicle navigation map updates. Components of the network would include road sensors (including speed, rain, ice, snow, fog, temperature, and lane change sensors), in vehicle sensors (within a few years vehicles are expected to have up to 200 built-in sensors to monitor all facets of a vehicle's condition and maintenance needs), telecommunications to communicate with the sensors (.e.g. for roadside slippery conditions alerts and collision avoidance alerts), telecommunications to conduct transactions (e.g., payment of tolls and parking meters), and telecommunications to access the Internet (for weather, traffic, tourist, and other transportation related information). Even at vehicle speeds up to 120 mph, the network is expected to be able to communicate at broadband speeds up 27mbps, more than 30 times the speed of the fastest 3G networks.

Of great importance, the major automobile companies that are expected to deploy the 802.11p technology in their vehicles beginning around 2009 are expected to also include 802.11a/b/g Wi-Fi capability. One reason is that even if it takes 802.11p technology some years to have much practical use, the 802.11a/b/g technology will have immediate practical use, if only to download from home to car various media including music, audio books, and weather reports. Note that two updated Wi-Fi standards expected to be released later this year, 802.11r and 802.11s, will incorporate mesh and high speed handoff technologies necessary for non-proprietary highway Wi-Fi.

So what great harm would be done if when the 802.11p network went up a Wi-Fi hot spot, at nominal cost, was added to each node? The answer is that such a proposal would cause the mobile telephone industry to rise up in arms. If the transportation and public safety industries hadn't assured the mobile telephone industry that they would build a crippled, limited purpose ITS network, they never would have been able to get the spectrum in the first place. And if they renege on that promise, the political opposition to ITS could rapidly become overwhelming. What would you prefer: a crippled ITS network or no network at all? The practical folks at the departments of transportation are happy to take the crippled network.

Smart Politics vs. Good Public Policy

However, just because highway Wi-Fi is considered a political pipedream doesn't mean it isn't good public policy. More hopefully, what is considered politically hopeless one day may turn out not to be. After all, a lot of people felt the same way about municipal Wi-Fi not too long ago.

Moreover, highway Wi-Fi might be far less expensive than anyone could have dreamed of when the ITS proceeding first began at the FCC in the late 1990s. For example, the military has developed optical fiber with RJ45 nodes (the telephone plug like outlets used to plug in Ethernet wire and wireless routers) spliced into it at the factory (think of Christmas lighting with light bulbs factory attached to the wire). That means that in Iraq the military can quickly lay a super high-speed broadband network, including secure and inexpensive Ethernet or wireless nodes. At a rate of a mile or more per hour, a special moving vehicle can lay this fiber 6 inches below a dirt surface like flexible irrigation pipe can be laid underneath a lawn or golf course. Of course, many other implementation scenarios can be imagined. But the point is that it is not currently in the interests of anyone at national, state, or local transportation departments to think this way.

In 2005, the state of Arizona, with a grant from the Department of Homeland Security, deployed highway Wi-Fi along a 32 mile stretch of the CANAMEX highway near the Mexican border in Arizona. The network, developed by New Zealand company RoamAD (see www.RoamAD.com), allows public safety vehicles to access the Wi-Fi network at speeds up to 80mph and at broadband speeds faster than available even on the most advanced 3G networks.

The era of single purpose networks such as the one the USDOT is proposing should be brought to an end. They may be appropriate for the analog age but not for the digital age we are entering. In the digital age, it is an economic travesty to build single purpose networks because telecommunications is becoming a fixed cost business. Once a high-speed network is built, it costs relatively little to add extra users and applications. Perhaps transportation bits should have the highest priority. But when capacity is left unused, other types of bits should also be allowed to share the network. Imagine how you would react if one day you were told that over your home high speed network you could only send transportation-related bits. Using public resources for a highway Internet solely dedicated to transportation bits is a similar travesty.