The Community for Technology Leaders
Green Image
Issue No. 10 - Oct. (2013 vol. 62)
ISSN: 0018-9340
pp: 1988-2000
Jonathan Lee , National Central University, Taiwan
Shin-Jie Lee , National Central University, Taiwan
Hsi-Min Chen , National Central University, Taiwan
Kuo-Hsun Hsu , National Taichung University, Taiwan
Recently, as more and more mobile devices embedded with intelligent software become part of our daily lives, Open Services Gateway initiative (OSGi) has gained increasing attention in the development of services for such devices. However, services residing in OSGi platforms are devised specifically for the platforms' local bundles. Although several works have enhanced OSGi with various communication mechanisms that enable remote service invocations, two crucial issues pertaining to mobile computing remain to be addressed: One is limited resources of mobile devices, and the other is unstable links experienced with mobile devices. To address these two issues, we propose an itinerary-based mobile agent approach with four major features: To implement mobile agents as OSGi bundles to deliver distributed OSGi services with lower resource consumption; to extend WSDL and WS-BPEL to help developers design mobile agent itineraries; to provide an opportunistic service matchmaking mechanism allowing mobile agents to adapt to context changes at runtime; and to devise asynchronous communication mechanism to enable reliable mobile agent transmissions among mobile devices with dynamic IP addresses and intermittent connection to the Internet. Through experimental evaluations, we demonstrate that both network traffic and turnaround time of our approach are better than those of client-server ones.
Mobile agents, Mobile handsets, Context, Mobile communication, Servers, Java, Computer architecture, mobile agent, Mobile agent itinerary, OSGi

H. Chen, S. Lee, J. Lee and K. Hsu, "Itinerary-Based Mobile Agent as a Basis for Distributed OSGi Services," in IEEE Transactions on Computers, vol. 62, no. , pp. 1988-2000, 2013.
82 ms
(Ver 3.3 (11022016))