Parallel and Distributed Systems Lab (PDSL)

Home Up Contents

Research - Mobile Computing

 

News
People
Research
Publications
Facilities
Support
Links
Private

Mobile Computing 

Mobile computing is becoming increasingly ubiquitous as more and more people carry wireless devices with them as they travel from home to office and well beyond. Services taken for granted in wired networks are not always well supported in mobile computing especially as a mobile user roams further away from their normal connection point to the wired network. Additional issues related to limitations of mobile devices and limited battery life compound these problems yet the benefits clearly outweigh the disadvantages as the adoption rate of wireless devices shows no sign of abating. As users expectations of being able to access their data and services they need "anytime, anywhere" grow, it is important to be able to develop new tools and techniques to meet the challenges of mobile computing without imposing complexity on mobile users who are, typically, not technically sophisticated.

Specific sub-areas of interest

Work on mobile computing in the PDSL has followed two major thrusts: performance improvement techniques for "traditional" mobile computing (e.g. web browsing, remote access, etc.) and optimization techniques for bandwidth intensive transmissions (most notably video). Some of the first research on mobile computing in the PDSL was based around an extension of the work of Walkty who proposed the use of mobile agents to dynamically transform web data downloaded to mobile devices to meet user, device and network preferences and conditions. By reassigning appropriate work into the wired network, the communication overhead between the "fixed" and "mobile" agents over the wireless link was reduced. This work is currently being extended again by Mr. Danny Liu (and MSc student supervised by Dr. Graham) to look at the possibility of doing "base station" load-balancing in urban hot-spots where agents are handed off to decrease load when needed. This work also lead, indirectly, to an assessment of the feasibility of mobile agent based routing in ad-hoc mobile networks done by another MSc student supervised by Dr. Graham, Mr. Lei Liang. Mr. Ian Scatliff, another MSc student of Dr. Graham's also did work on optimizing wireless web browsing, but without the use of mobile agents. He added power awareness (patterned after the work of Kahol) to the VAP pre-fetching strategy to create a technique known as SVAP which provides significant power savings on mobile devices. Currently, Mr. Channa Basani, an MSc student being supervised by Dr. Graham, is working on a new multicasting algorithm for use in mobile ad-hoc networks that incorporates group movement prediction to increase the probability of multicast trees maintaining their connectivity. Work on audio/video optimization for mobile computing in the PDSL began with the work of Mr. Xinliang Zhou (co-supervised by Drs. Graham and Eskicioglu) who looked at supporting collaborative applications using dynamically adaptive multicast trees. This work was picked up by Mr. Mohammad Shorfuzzaman for his MSc research (co-supervised by Drs. Eskicioglu and Graham) who looked at applying active network technology to support the in-network adaptation of media streams and implemented an MPEG-1 video transcoder based on the IXP2100 network processor.

Our Publications on Mobile Computing

bulletRelated Papers
bulletRelated Presentations
bulletOther Related Publications

Related Work on Mobile Computing

bulletLinks on Mobile Computing

For More Information Contact:

Parallel and Distributed Systems Lab
E2-556 EITC University of Manitoba, Winnipeg, MB, Canada R3T 2N2
Tel: 204-474-8837
FAX: 204-474-7609
Internet: pgraham@cs.umanitoba.ca

 

Send mail to pgraham@cs.umanitoba.ca with questions or comments about this web site.
Last modified: 03/23/06