An Experimental Approach for Optimising Mobile Agent Migrations

TL;DR

This paper investigates optimization techniques for mobile agent platform design to reduce network traffic and latency during agent migrations, supported by timing experiments and proposed new methods.

Contribution

It introduces optimization ideas for mobile agent migration, backed by experimental results, to improve transfer efficiency and reduce delays.

Findings

Optimized migration reduces network traffic.

Timing experiments reveal key delay factors.

Proposed techniques lower migration latency.

Abstract

The field of mobile agent (MA) technology has been intensively researched during the past few years, resulting in the phenomenal proliferation of available MA platforms, all sharing several common design characteristics. Research projects have mainly focused on identifying applications where the employment of MAs is preferable compared to centralised or alternative distributed computing models. Very little work has been made on examining how MA platforms design can be optimised so as the network traffic and latency associated with MA transfers are minimised. The work presented in this paper addresses these issues by investigating the effect of several optimisation ideas applied on our MA platform prototype. Furthermore, we discuss the results of a set of timing experiments that offers a better understanding of the agent migration process and recommend new techniques for reducing MA transfers delay.