Next-generation wireless research advances
An ambulance races a critically ill patient to a near-capacity emergency room. An emergency medical technician at the patient's side confers with a surgeon at the hospital. The surgeon and technician make decisions regarding the patient's care by sharing voice, X-rays, patient history and other medical data over a single wireless communications link.
That scenario could not happen today. Current mobile communications technology cannot handle data on that scale. Anyone who's had a cellular phone lose its signal can understand why. Until researchers devise methods to reliably and efficiently send large volumes of data over the limited frequencies available, large-scale E-commerce will remain largely tethered to a wire. A team of electrical and computer engineers is working to change that through intelligent coupling.
By investigating the boundaries between traditional layers of communication, the researchers believe they can increase the amount of data that travels usable frequencies between 800 megahertz and 30 gigahertz by a factor of 1,000 or more. Specifically, Professors Barry Van Veen, Parameswaran Ramanathan, James Bucklew and Rajeev Agrawal, Associate Professor Daniel van der Weide, and Assistant Professors Susan Hagnessand Akbar Sayeed are finding potential for great efficiency where the physical layer interfaces with the network layer. A wired network can send a packet of information without a care for the mode of transport, be it copper, fiber or cable. But for the wireless world to operate on the level of the wired, intelligent coupling of these layers is a must.
"With fiber or copper the quality is constant and this separation of layers works well," says Van Veen. "If a packet of information doesn't show up, there is feedback and the network resends. With wireless, the quality is changing all the time. If the quality is bad, a device will keep asking to resend and the whole system gets bogged down. What we are trying to do is exchange information between these separated layers. We're passing information to the network when the signal is poor that causes it to wait. When the signal is good, the network is informed that it can send at a high rate."
The idea of intelligent coupling lies not only between communication layers, but between traditional engineering disciplines. Hagness has knowledge of antenna design and how radio waves propagate through complicated media. van der Weide is an expert in broadband microwave circuitry and Bucklew specializes in fast simulation of rare events. Sayeed and Van Veen have expertise in signal processing, and Agrawal and Ramanathan are network specialists. The team also works with researchers at the University of Colorado and collaborates with the French telecommunications school Eurecom and industrial partners including Motorola and Texas Instruments.
Additional opportunities for intelligent coupling could arise if the Federal Communication Commission establishes a secondary licensing system for the spectrum. Under this system, a primary license holder could sell a portion of their frequency band. A license holder that has no use for its frequency at night, might sell rights to another service provider who wants to offer higher data rates at that time. Van Veen says this scenario calls for an intelligent telephone system that understands which frequencies are available at various data rates and costs.
"We don't know what the future looks like, but it seems clear that these kinds of developments support the need for intelligent coupling between wireless devices and networks," says Van Veen. "A person can't assume that their phone or wireless gadget will transmit at 854 megahertz all the time. There are already phones that work in both digital and analog formats, but we are talking about even greater flexibility. Each frequency band is going to have different qualities and data rates that must be coupled with the network."
Intelligent phones will need to be coupled to an intelligent network that can prioritize requests for information. A request for voice might get lower priority but at a lower cost. People might pay for guarantees of higher data rates or allow their signal to be bumped in exchange for a lower price. It's not unlike the arrangements electric utilities make with customers when demand for power outstrips supply on a hot day.
"What drives all of this is the fact that the usable radio spectrum is finite. Above 30 gigahertz you start to have problems with atmospheric affects. It's a finite band and when you talk about this explosion of services being proposed it basically means that we have to wring out of the spectrum every last bit of performance that we can."