Mobile agents for distributed simulation

This paper presents the basic ideas behind the use of mobile agent technology for distributed simulation. We focus on a specific mobile agent implementation, D'Agents, that is gaining widespread use. Mobile agent software technology allows programs, along with their execution states, to migrate within a network. To understand the role of agents in distributed simulation, note that simulations often operate on static datasets and data sources. Many simulations would produce more-accurate results if they could access dynamically-changing data from other sources. From the perspective of one simulation, other simulations are data resources, producing information possibly relevant to the past, present, or future of the system being modeled. Mobile agents allow dynamic linking between distributed simulations and efficient monitoring of and access to remote data resources. Mobile agent technology permits dynamic, decentralized creation and execution of execution modules that can execute on remote machines and spawn clones for parallel computation. This allows foreign programs to execute locally on a server but entails solving many problems in security, resource allocation, directory services, and debugging. This paper presents an overview of existing and proposed solutions to these problems. To demonstrate the use of agents for distributed simulation, we present a case study involving operational ocean simulation and red-time data assimilation. To evaluate the efficiency of this approach, performance measurements will also be given.