Calit2

Adaptive Systems: The "ABC" of Wireless Connectivity

 


Project: Adaptive Systems
PI: Ramesh Rao, Prof, Jacobs School of Engineering; UCSD Division Director, Calit2
Funding: Ericsson, UC Discovery Grant Division: UCSD
Corporate Partner: Ericsson Wireless
Start Date: August 2002
PDF:PDF
DOC:DOC


"ABC" of Wireless Connectivity


Wireless users may be surrounded by half a dozen or more wireless "bubbles"-competing cell-phone networks, short-range Bluetooth, local area 802.11b (Wi-Fi) high-speed data service, or, in many cities, so-called third-generation (3G) coverage such as CDMA2000 networks. But as wireless systems and standards proliferate, how can these technologies work for consumers-not just confuse them?

With $2.7 million in funding from the UC Discovery Grant program and Ericsson, a team of up to 20 researchers from Calit2 and the UCSD Jacobs School of Engineering is undertaking a four-year project to enhance wireless connectivity. "Adaptive systems" would permit seamless interoperability among various network infrastructures to maintain constant-and the most powerful-connection to the Internet for mobile users using both licensed (e.g., cellular networks) and unlicensed (e.g., 802.11b) spectrum. Users would be switched invisibly and seamlessly between wireless networks based on whichever is the optimal connection at the time, hence the project's nickname, "Always Best Connected" (ABC).

"Think about the options facing a wireless user. The consumer needs to send packets of information from point A to point B and can choose among a variety of networks at different times of day. But to take advantage of alternatives, you need a way to prioritize. It’s like the airlines, which have figured out how to get planes off the ground, when to land them, and how to price the tickets."
- Ramesh Rao, UCSD Division Director, Calit2

The ABC project is part research, part deployment. On the UCSD campus, researchers have already installed hardware and software that allow user devices to roam among four network access technologies: 802.11b, GPRS, CDMA2000 1xRTT, and CDMA2000 1xEVDO, as well as local-area Ethernet. These networks vary in speed, and some are always-on. Capacity alone can be a major problem if you are dealing with multimedia and switching from high-bandwidth Wi-Fi to cellular access.

One major research focus is the job of encoding, delivering, and dealing with errors across systems. These tasks are usually designed around the particular characteristics of one system. But if you're going to run your application across five different systems, you have to adapt much more dynamically than before to the underlying characteristics. You have to adapt not only to channel conditions, but also to the user's conditions-how much battery power the device has, what type of display, and whether it's a color screen or not. One goal is to figure out the smart thing to do at the level of the mobile device. But ABC researchers are also investigating the smartest way to architect the system. There is also huge potential to develop new services that will be broadly available whatever your access technology. Co-PIs on this project are Pamela Cosman, Rene Cruz, Sujit Dey, and Geoff Voelker.

Contact: Ramesh Rao
(858) 822-4572
rrao@ucsd.edu