Abstract
The mobile agent technology is emerging as a useful new way of building large distributed systems. The advantages of mobile agents over the traditional client-server model are mainly non-functional. We believe that one of the reasons preventing the wide-spread use of mobile agents is that non-functional properties are not easily grasped by system designers. Selecting the right interactions to implement complex services is therefore a tricky task. In this paper, we tackle this problem by con- sidering eficiency and security criteria. We propose a language-based framework for the specification and implementation of complex services built from interactions with primitive services. Mobile agents, Rpc, re-mote evaluation, or any combination of these forms of interaction can be expressed in this framework. We show how to analyze (i.e. assess and compare) complex service implementations with respect to eficiency and security properties. This analysis provides guidelines to service design- ers, enabling them to systematically select and combine difierent types of protocols for the efiective realization of interactions with primitive services.
Cite
CITATION STYLE
Fradet, P., Issarny, V., & Rouvrais, S. (2000). Analyzing non-functional properties of mobile agents. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1783, pp. 319–333). Springer Verlag. https://doi.org/10.1007/3-540-46428-x_22
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.