Page 2... continued

THE DEPENDABILITY FACTOR

The dependability of wireless networks concerns Alisha Malloy. With Varshney as her adviser, Malloy is currently working on her dissertation on wireless communication systems and specifically on the dependability and quality of service (QoS) of those networks.

Currently, wireless and mobile services are not as dependable as a wired network. Failures affect not only simple voice and data transmission but also limit complex applications such as multimedia applications and video conferencing, which require higher bandwidth. "In the past, the convenience of mobility was enough to satisfy users of wireless networks, despite the decrease in the quality and scope of service offerings compared to wireline networks," Malloy writes in her dissertation. She predicts, however, that users ultimately will demand the same dependability and QoS guarantees they have with wireline-based telecommunications.

Malloy has immersed herself in telecommunications since first serving as a vault officer in the U.S. Navy, where she issued secure cryptographic software and hardware for the entire Atlantic Fleet. She earned a masters degree in engineering management at Old Dominion University in Norfolk, where one of her teachers ­ Dr. Billie Reed, who worked with NASA ­ taught her to consider whether the tools she designed were user friendly. "I learned that there was a whole other side to these black boxes we make," Malloy comments.

After leaving the Navy, she took a position as a network analyst for Sprint, designing data networks for customers and later as a business analyst performing special pricing and maintaining customer relations. In her return to the classroom as a doctoral student in the CIS Department, Malloy returned to her technical strengths.

In articles co-authored with Varshney and CIS professor Andrew Snow, Malloy has examined the reliability and dependability of mobile networks. In one such article published in Computer, the authors identify components that are prone to fail in a wireless infrastructure. A base station, for instance, serves hundreds of mobile users in a given area, or cell, by allocating resources that allow users to make new calls or continue their calls if they move to the cell. A base station controller provides switching support for several neighboring base stations and serves thousands of users. A mobile switching center is a larger switch capable of serving more than 100,000 users. Home location and visiting location registers keep track of users who are permanently registered or who are just visiting the area. Signaling System 7 performs the call setup between mobile switching centers and also to the public switched telephone network (PSTN). Finally, high-capacity trunks carry calls between mobile switching centers and the PSTN. A failure in some of these areas might affect hundreds of thousands of users.

Although the FCC requires wireline networks to report all failures affecting 30,000 or more customers for 30 minutes or longer, the agency has not required wireless and mobile networks to report such information. "We have no such data for wireless networks," Malloy says. "What we know is strictly based on word of mouth."

For her dissertation research, Malloy created a simulation of the daily functions of a wireless network. Previous computer models looking at wireless networks assumed the system was always up and operational. Malloy took a different approach, looking for where the system might break down. Her simulation quantifies wireless network performance during both normal operations and when glitches in the system cause service disruptions.

Malloy's research emphasizes the need for architectural changes in wireless infrastructures. "Wireless is not set up for backup or recovery," Malloy says, but with changes in the design, architects could remedy that problem by introducing such elements as redundancy and network overlays to the system.

Malloy will get a chance to apply her knowledge upon graduation in the summer of 2002. She has accepted a faculty position in the Culverhouse College of Commerce and Business Administration at the University of Alabama. She will help build the school's newly established PhD program as well as work in its Enterprise Integration Lab, a research vehicle that helps organizations improve their effectiveness by seamlessly integrating business activities with the use of information technology. One area of particular interest to Malloy is working to improve school scores in the "black belt," an area that stretches from Georgia through Alabama and Mississippi to Louisiana and is home to an underserved population of minorities. Wireless networks offer schools there the possibility of raising test scores through tools available online and connections to the information superhighway.

LOOKING AHEAD

Dependability and QoS are just two elements that may delay or contribute to the success of the wireless frontier. Convenience and price are others. "People need to see value in the technology before they will use it, and it has to be relatively cheap," asserts Varshney. Smart phones, for example, already exist in the United States, but their price tag at approximately $699 is too high, as are usage fees. In addition, their small keyboards are hard to use. The devices need to be bigger and the bandwidth cheaper, Varshney said. He believes a pricing restructuring will have to precede extensive use, such as occurred when Internet providers changed from charging by the hour to a monthly fee for unlimited access.

The Internet is the predecessor for wireless commerce in another aspect. "A lot of technologies were built without thinking of applications," Varshney explains. "For example, the Internet began as a way of communicating between universities. Then business saw a way for it to be used in e-commerce although it wasn't originally designed for that purpose."

Now the same is happening for wireless. An infrastructure is in place, and experts like Varshney and Malloy are searching for the applications to make full use of its possibilities. They are asking the questions of the future: which applications need location services, how do we tailor those applications for certain products, which systems are currently available for location tracking, what middleware is necessary to support the applications. They are on the frontier of a brave new world.

Rhonda Mullen