main menu:
  » home
  » introduction
  » indigolog overview
  » oaa overview
  » interface model
  » developing agents
  » simple example
  » advanced example
  » other interface models
  » implementation details
  » tips
  » links

 
   

Multiagent Platforms

 
The popularity of agents is continuing to grow in the research community as well as in the industry. This has resulted in emergence of many new agent programming languages.

Since most researchers agree that social ability is one of the most important features of an agent, a lot of new developments in the agent community are in the area of multiagent systems.

A number of multiagent platforms based on different architectures, different agent communication languages, and targeted at various domains and applications were presented in the past several years. These systems provide various levels of support for the developer of multiagent software.

Some of them just provide the infrastructure for the multiagent system including communication libraries, agent communication languages, and some king of matchmaker or facilitator agents while not assuming anything about the way individual agents are written. The developer is free to choose the programming language for the agents, the level of their intelligence and other characteristics. OAA provides communication libraries and an agent communication language, but does not include agent programming language, so it could be included in this group.

On the other hand, there are multiagent platforms that provide not only the communication infrastructure, but also the agent programming language. These systems often include integrated environments for rapid application development. The programmer does not have much choice over the way individual agents are written and often can only choose predefined patterns for inter-agent communications. While this may be a good solution for some projects, these systems usually cannot be applied to a broad range of problems.

IndiGolog in MAS



Previous attempts were made to create multiagent systems using IndiGolog. The agents used low-level TCP/IP communication to exchange information. While TCP/IP allows agents written in many languages to communicate with each other, many would prefer most of the low-level code hidden in order to concentrate on the functionality of the agents.

The goal of this work was to provide an interface that would allow IndiGolog agents to be easily integrated into Open Agent Architecture system while retaining all of their functionality and the IndiGolog style of programming. IndiGolog OAA agents using this interface are more powerful than the regular OAA agents: they are both reactive and proactive while generally OAA agents can be either reactive or proactive. Therefore, these agents are able to overcome one of the major limitations of OAA. This is the solution to the problem that arises due to the fact that IndiGolog controller agents have their event loops while OAA has its own. Here we propose a solution that integrates these event loops, therefore allowing an IndiGolog agent to monitor and react to both OAA and IndiGolog events.

 

 
 
© 2001 All Rights Reserved. See Legal Notice.