Dynamic composition of functionality using code mobility for autonomous objects

Abstract

Internet-based distributed systems are required to possess the capabilities such as scalability, robustness, flexibility, dynamicity, etc. In order to support these capabilities, a new network computing paradigm exists: Autonomous Object Paradigm (AOP). An autonomous object is a mobile object capable of navigating autonomously through the network and performing tasks at the nodes it visits. If an autonomous object can dynamically compose functionality in other nodes into its own behavior by code mobility at run time, AOP can be closer to sufficiently satisfy the above capabilities. In this thesis, for dynamically composing functionality of autonomous objects by code mobility at run time, we propose Dynamic Composition of Functionality (DCF) based on code mobility. DCF based on code mobility can enhance the capabilities of AOP with respect to dynamic extensibility and changeability of functionality, modularity, fault-tolerance, and reducing network traffic. In order to realize such capabilities of AOP, we propose the language constructs for DCF based on code mobility and a prototype of autonomous object platform, called DC-AOP. We also present case studies using DC-AOP to show the usefulness of our proposal.