|
||
 |
||
|
After the RMS Titanic hit an iceberg and sank in 1912, the U.S. government decided to set up a dedicated emergency channel so that it would always be possible to radio warnings to ships and avert future disasters. The government actually had its facts wrong: Despite heavy outgoing radio traffic from passengers, operators on the Titanic had received numerous warnings about icebergs. The problem was with processes on the Titanic, where messages were either never passed on to the captain or where the captain ignored the warnings. Still, the point is that the government began to regulate the use of the airwaves. For decades, the Federal Communications Commission (www.fcc.gov) handed out the right to use a piece of the spectrum free, to those who made the most persuasive argument. In recent years, the FCC made what has been hailed as a major improvement: It now treats spectrum as property, selling slices to the highest bidders. But it is time for a more fundamental change. To this point, spectrum has been treated as scarce. It was, in the time of the Titanic, because of the technology available then, but not any longer. Instead, there is reason to see spectrum as almost limitless—with all the implications that change holds for the population at large, for the companies that now control so much of the available spectrum, and for government, whose restrictions on what can be broadcast stem from the notion that spectrum is a precious public resource. The notion that the airwaves need to be meted out carefully is based on the fact that, in the early 1900s, radio equipment was easily thrown off when signals of the same frequency from more than one source overlapped. We call this phenomenon “interference.” In fact, the waves sent out by different transmitters don’t interfere with each other at all. They pass right through each other unchanged. Interference occurs in the receiver, when its antenna picks up multiple signals of the same frequency and has trouble telling them apart. In other words, interference is a function of the intelligence designed into the receiver, not a function of what happens in the airwaves—and receivers can be a lot more intelligent than they were 90 years ago. Recent advances are enabling radio signals to be coded digitally so they can easily be separated from each other. No longer is there a need to chop the airwaves into distinct regions of frequency and geography. Once we no longer have to reserve certain frequencies for specific services, we can avoid problems such as those encountered on Sept. 11, 2001, when cellphone users couldn’t get calls through because airwaves set aside for mobile phone use were jammed, while wide swaths of spectrum allocated to other uses were virtually silent. Moreover, wireless phones would be able to tap into any network and stay connected as they move around the globe. It gets better. The airwaves’ capacity can actually increase as we add devices—cellphones, televisions, and other gadgets based on digital technologies—to these networks. Because the number of those devices will continue to explode as new technologies come to market, the capacity of spectrum can increase to accommodate almost any amount of demand. The growth will come because it will be possible to design networks that perform the neat trick of pulling themselves up by their own bootstraps, achieving what I call “cooperation gain.” Wireless devices can be designed so they can form self-organizing repeater networks—allowing messages to follow a path composed of short hops, repeating from device to device at low energy rather than transmitting directly from a source to a destination. Because each device in the network is a transmitter as well as a receiver, capacity expands as the number of devices increases. Other techniques, such as one called space-time coding, also generate cooperation gain. How much cooperation gain can be achieved is an area of active research and debate, but it is dramatic. Now for the hard part. Most current gadgets that communicate via the airwaves won’t do the trick. That is because those devices are limited in the way they send and receive signals. What we need are digital “software-defined radios.” Unlike devices that rely on fixed hardware designs, these wireless units would be able to adapt and shape the local airwaves to meet the instantaneous demands of users. We also need new network designs, because the sort of pre-allocation of spectrum that currently is the rule prevents the possibility of cooperation gain. In the late 1960s, a group of us designed communications standards that let the Internet be so adaptable that computers that connect to it can easily switch to new standards and take on new capabilities; in the past, such a switch would have required changes to the network that hooks the devices together. Flexibility has been a crucial part of the power of the Internet and should be mimicked in the design of wireless networks. Perhaps the biggest challenge is the one that technology can’t solve: how spectrum is regulated. The FCC still restricts technologies such as software-defined radio, preventing nearly all kinds of networking within frequency bands and, thus, keeping us from boosting bands’ capacity. Different kinds of content are restricted to different frequency bands, making it impossible to share spectrum as demands shift. Most of the available spectrum is carved up and allocated to existing services—which may be the thorniest issue of all. There are, however, ways to get started toward a limitless wireless network. For instance, as small bands of spectrum become available—perhaps as the needs of a broadcaster or telecommunications operator shift—the government could begin pooling the spectrum for common use with very limited licensing requirements. If we can reclaim even some spectrum from its legacy owners, we can create room for innovators to build more open wireless networks. Getting to a completely open network will take years, or even decades, but that is all the more reason to begin the transition now.
|
