Graham, J., and Decker, K.S.: Towards a distributed, environment-centered agent framework. Intelligent Agents VI, N.R. Jennings and Y. Lesperance (Eds.), LNAI1757, Springer Verlag (2000) 290-304.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....it to bene t from these services. However, adaptation supported by RAJA is beyond this technical level , i.e. physical layer of message transport. DECAF (Distributed Environment Centered Agent Framework) is a modular platform for the rapid design, development, and execution of intelligent agents [14]. It provides a range of architectural services such as communication, planning, scheduling, execution monitoring, and coordination. The framework supports di erent types of adaptation including organizational, planning, scheduling, and execution time adaptation [5] The current implementation of ....

J. Graham and K. Decker. Towards a Distributed, Environment-Centered Agent Framework. In Proc. of International Workshop on Agent Theories, Architectures, and Languages (ATAL99), 1999.

....resource adaptivity. In the context of pervasive and mobile computing numerous middleware for resource adaptive applications have been developed. Examples are the non agent based approaches such as Odyssey [12, 13] and Agilos [9] and the actor or agent based approaches like TLAM [15] and DECAF [7]. RAJA fills the gap of supporting resourceadaptivity left by existing specifications. It distinguishes itself by its greater flexibility, which characterizes agent based approaches and by its concept of controllers, whose advantages have been demonstrated in detail in section 2.2. 5 Conclusion ....

J. Graham and K. Decker. Towards a Distributed, Environment-Centered Agent Framework. In Proc. of International Workshop on Agent Theories, Architectures, and Languages (ATAL99), 1999.

....easily be applied to FIPA ACL as well. 2.2. Existing Agent Architectures Many agent architectures have been developed to support multi agent systems. Here we highlight a few and describe how they relate to our work. The DECAF (Distributed, Environment Centered Agent Framework) Agent Framework [4] is a Java based multi agent system. It provides a matchmaker agent that accepts KQML performatives to allow for agent communcation. SRI International has developed an agent architecture called Open Agent Architecture (OAA) 5] Agent interaction in OAA is done through an agent facilitator using a ....

....scheduling [10] is an extension of the design to time algorithms. The model here is more general in that it can take into account any scheduling criteria, such as time, cost, and quality, and it can use uncertainty as part of the decision making process. The DECAF architecture described above [4] is incorporating scheduling algorithms based on the designto criteria model. ASTRO [11] is a model for real time agents in which each goal is divided into subgoals that each have a deadline. The scheduling algorithm used by the ASTRO schedules all subgoals with their minimal duration. It then ....

John Graham and Keith Decker. Towards a Distributed, Environment-Centered Agent Framework. In Proceedings of the 1999 Intl. Workshop on Agent Theories, Architectures, and Languages [ATAL-99], Orlando, July 1999.

....final scheduling phase. A resource modeling component is also used during this analysis to ensure that resource constraints are also 1 From private conversations, it appears that the technique used to generate this schedule is similar to that seen in the DRU scheduler from the DECAF framework [4], a system developed concurrently with the research presented here. Task2 q min Set Parameters enables2 Track q max Track Low Track Medium Track High enables3 Send Results lock2 release2 RF 0.0 1000.0 1000.0 Figure 2: An example TMS task structure for tracking. respected. A conflict ....

John R. Graham and Keith S. Decker. Towards a distributed, environment-centered agent framework. In Proceedings of the Sixth International Workshop on Agent Theories, Architectures, and Languages (ATAL), Florida, jul 1999.

....4 2.2 Agent Architectures Many agent architectures have been developed to support multi agent systems. Here we highlight a few to indicate the wide range of approaches that have been taken in developing such systems. The DECAF (Distributed, Environment Centered Agent Framework) Agent Framework [4], developed at the University of Delaware, is a Java based multi agent system. It provides a matchmaker agent that accepts KQML performatives to allow for agent communcation. DECAF also provides a toolkit for building agents to remove the low level details from the development process. FIPA ....

....decision making. This algorithm provides a mechanism for making decisions 5 within specified timing constraints. For instance, when a specified deadline is near, the algorithm chooses a quicker decision making method than if the deadline was longer. The DECAF architecture described above [4] has incorporated similar algorithms for agent scheduling. At the University of Washington in St. Louis, the DOVE [12] project (Distributed Object Visualization Environment) uses a real time agent framework based on TAO (The ACE ORB) 13] to provide a system for network management and performance ....

John Graham and Keith Decker. Towards a Distributed, Environment-Centered Agent Framework. In Proceedings of the 1999 Intl. Workshop on Agent Theories, Architectures, and Languages [ATAL-99], Orlando, July 1999.

....used to develop and evaluate multi agent system up to 20 agents. Even if those simulations work, they are involving a limited number of agents. Our next goal is to build a system with more than 100 agents using a distributed version of MASS. We are also working on integrating MASS and the DECAF [8] architecture from the University of Delaware to allow people to use MASS with their own agent architecture. 14 Based on those experiments, we found that MASS has a lot of necessary features for multi agent systems: using the scripting language built in has allowed us to have only single point ....

John R. Graham and Keith S. Decker. Towards a distributed, environment-centered agent framework. In Proceedings of the Sixth International Workshop on Agent Theories, Architectures, and Languages (ATAL), Florida, jul 1999.

....networks to be specified easily. We believe that this is an important contribution of this paper. There has been a lot of work on applying HTN planning techniques in practical applications. Only in this volume, we can find the use of a HTN planner in monitoring objectives queue of the DECAF system [2] or in developing a RETSINA Multi AgentSystem [11] Our implementation of the htn construct in PROLOG shows that hierarchical task networks can be solved effectively in logic programming. It also shows that the HTN planning techniques can be useful in a high level execution language such as ....

J.R. Graham and K.S. Decker. Towards a Distributed, Environment-Centered Agent Framework. Pre-Proceedings ATAL'99, 162--175, 1999.

No context found.

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In N.R. Jennings andY. Lesperance, editors, Intelligent Agents VI, LNAI-1757, pages 290--304. Springer Verlag, 2000.

No context found.

J.Graham and K.Decker. Towards a distributed, environment-centered agent framework. In Intelligent Agents IV, Agent Theories, Architectures, and Languages. Springer-Verlag, 2000.

....money) to spend to achieve an answer of some characterization (quality, certainty, etc. The next section will describe our realization of this general model using DECAF. 3 DECAF DECAF (Distributed, Environment Centered Agent Framework) is a Java based toolkit for creating multiagent systems [17]. In particular, several tools have been developed specifically for prototyping information gathering systems. Also, the internal architecture of each DECAF agent has been designed much like an operating system as a set of services for the intelligent (resource efficient, adaptively scheduled, ....

....the internal operating system of a software agent, to which application programmers have strictly limited access. The overall internal architecture of DECAF is shown in Figure 1. These modules run concurrently, each in their own thread. Details of the DECAF implementation can be found elsewhere [17]. Agent Initialization DECAF Task and Control Structures Plan File Incoming KQML messages Domain Facts and Beliefs KQML Messages Outgoing Action Modules Hashtable Action Queue Pending Results Queue Action Dispatcher Planner Executor Message Queue Incoming Queue Objectives Queue ....

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In N.R. Jennings and Y. Lesperance, editors, Intelligent Agents VI, LNAI-1757, pages 290--304. Springer Verlag, 2000.

....alternatives. Thus coordination mechanisms of many different styles can be thought of as ways to provide information to a local scheduler that allow it to construct better schedules. This paper concerns an implementation of these ideas using DECAF (Distributed Environment Centered Agent Framework) [6] as an architecture and set of agent construction tools for building real multi agent systems. DECAF is being used to build applications in electronic commerce, bioinformatics, and plant alarm management. 2. TASK STRUCTURE ALTERATION To achieve its desires, an agent has to select appropriate ....

J.Graham and K.Decker. Towards a distributed, environment-centered agent framework. ATAL, Springer-Verlag, 2000.

....for every agent. Our goal is to enable the agents to adapt themselves under different environments and to select the best instantiation of the coordination components based on proper coordination mechanism selection strategy. We use DECAF(Distributed, Environment Centered Agent Framework) system [11] as agent construction tool. Each DECAF agent consists of several components, which maintain the agents functions and keep congruent to the traditional BDI (Belief, Desire,Intention) model. The coordination mechanism construction is independent from the agent construction. That is to say, DECAF ....

J. Graham and K. Decker. Towards a distributed, environment-centered agent framework. Intelligent Agents IV, Agent Theories, Architectures, and Languages, Springer-Verlag, 2000.

.... DECAF Task and Control Structure. We will discuss Plan Files, KQML messages (both outgoing and incoming) and the Action Modules (Java code) Domain Facts and Beliefs are the facts and beliefs that the agents users have about the domains in which they are solving problems with DECAF. Reference [1, 2] describe the insides of the box. 1.3 Experience the Demonstration DECAF has a demonstration accessible through the DECAF homepage. The demonstration produces a five agent world: ffl ANS is the Agent Name Service, which is essential to have running for any DECAF agent application to find the ....

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In N.R. Jennings and Y. Lesperance, editors, Intelligent Agents VI, LNAI-1757, pages 290--304. Springer Verlag, 2000.

....on our work on RETSINA; the DECAF toolkit; our initial annotation system; and our new extensions for functional annotation, EST processing, and metabolic pathway reasoning. 2 DECAF DECAF (Distributed, Environment Centered Agent Framework) is a Java based toolkit for creating multi agent systems [13]. In particular, several tools have been developed specifically for prototyping information gathering systems. Also, the internal architecture of each DECAF agent has been designed much like an operating system as a set of services for the intelligent (resource efficient, adaptively scheduled, ....

....and the information extraction agent shell) we have built to implement multi agent information gathering systems. The overall internal architecture of DECAF is shown in Figure 1. These modules run concurrently, each in their own thread. Details of the DECAF implementation can be found elsewhere [13]. Agent Initialization DECAF Task and Control Structures Plan File Incoming KQML messages Domain Facts and Beliefs KQML Messages Outgoing Action Modules Hashtable Action Queue Pending Results Queue Action Dispatcher Planner Executor Message Queue Incoming Queue Objectives Queue ....

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In Intelligent Agents VI, LNAI-1757, pages 290--304. Springer Verlag, 2000.

....in the next section) provides an implementation of these middle agents, reusable agent classes, and other tools for building multi agent information gathering systems. 3 DECAF DECAF (Distributed, Environment Centered Agent Framework) is a Java based toolkit for creating multi agent systems [16]. In particular, several tools have been developed specifically for prototyping information gathering systems. Also, the internal architecture of each DECAF agent has been designed much like an operating system as a set of services for the intelligent (resource efficient, adaptively scheduled, ....

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In N.R. Jennings and Y. Lesperance, editors, Intelligent Agents VI, LNAI-1757, pages 290--

....ABSTRACT Before the powerful agent programming paradigm can be adopted in commercial or industrial settings, a complete environment, similar to that for other programming languages, must be developed. This includes editors, libraries, and an environment for the completion of agent tasks. The DECAF[8] Agent architecture is a general purpose agent development platform that was designed specifically to support concurrency, distributed operations, support for high level programming paradigms, and high throughput. The architecture has been designed with built in scalability which adapts itself to ....

J. R. Graham and K. S. Decker. Towards a distributed, environment-centered agent framework. In N. Jennings and Y. Lesperance, editors, Intelligent Agents VI --- Proceedings of ATAL-99, Lecture Notes in Artificial Intelligence. Springer-Verlag, Berlin, 2000.

....each economic entity agent is concerned with and in control of its own welfare, and is thus an autonomous agent. Third, each agent is viewed as intelligent because it engages in purposeful activity expected to achieve goals or objectives. We use these three observations and the DECAF agent toolkit [2] to model individual consumer and individual business behaviors explicitly, giving us a collection of economic agents which we expect to behave as received economic theory instructs us. Our agents have limited rationality (unlike traditional economic agents) can engage in certain specific ....

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In Proceedings of the 1999 Intl. Workshop on Agent Theories, Architectures, and Languages (ATAL-99), pages 162--175, 1999.

....entity agent is concerned with and in control of its own welfare, and is thus an autonomous agent. Third, each agent is viewed as intelligent because it engages in purposeful activity expected to achieve goals or objectives. Having made these three observations, we use the DECAF agent toolkit [5] to model potential individual consumer and individual business behaviors explicitly, giving us a collection of economic agents which we expect to behave as received economic theory instructs us. Our agents have limited rationality (unlike traditional economic agents) can engage in certain ....

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In Proc. ATAL-99, pages 162--175, 1999.

....ABSTRACT Before the powerful agent programming paradigm can be adopted in commercial or industrial settings, a complete environment, similar to that for other programming languages must be developed. This includes, editors, libraries, and an environment for the completion of agent tasks. The DECAF[9] Agent architecture is a general purpose agent development platform that was designed specifically to support concurrency, distributed operations, support for high level programming paradigms and high throughput. The architecture has been designed with built in scalability which adapts itself to ....

J. R. Graham and K. S. Decker. Towards a distributed, environment-centered agent framework. In N. Jennings and Y. Lesperance, editors, Intelligent Agents VI --- Proc. of ATAL-99, Lecture Notes in Artificial Intelligence. Springer-Verlag, Berlin, 2000.

....of a simple information gathering problem: 2 queries, one with multiple plans, and several clauses, some of which could be fulfilled at possibly multiple, limited capacity sources. This work is not being carried out in simulation, but rather in a real time distributed agent toolkit called DECAF [15]. In order to reduce broadcast communication and to make the system scalable, we introduce a resource manager agent, which enforces access to a mutex or limited capacity resource. Initially, the resource manager agent centralizes the bid processing (for some resource) carrying out the decision ....

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In Proceedings of the 1999 Intl. Workshop on Agent Theories, Architectures, and Languages (ATAL-99), 1999.

....the underlying structures. Thus, the programmer does not need to understand JAVA network programming to send a message, or learn JAVA database functions to attach to the internal knowledge base of the framework. The details of the internal mechanism of the DECAF architecture can be found in [5]. 2 THE EXPERIMENTAL STUDY Figure 1 represents the high level structure of the DECAF architecture. Structures inside the heavy black line are internal to the architecture and the items outside the line are user written or provided from some other outside source (such as incoming KQML messages) ....

John R. Graham and Keith S. Decker. Towrds a distributed, environment centered agent framework. In Proceedings of the of the Sixth International Workshop on Agent Theories and Languages, 1999.

No context found.

J. Graham and K.S. Decker. Towards a distributed, environment-centered agent framework. In Proceedings of the 1999 Intl. Workshop on Agent Theories, Architectures, and Languages #ATAL-99#, 1999.

No context found.

Graham, J., and Decker, K.S.: Towards a distributed, environment-centered agent framework. Intelligent Agents VI, N.R. Jennings and Y. Lesperance (Eds.), LNAI1757, Springer Verlag (2000) 290-304.

No context found.

Graham, J. and Decker, K.: Towards a Distributed, EnvironmentCentered Agent Framework. In: N. Jennings and Y. LeSperance (eds.): Intelligent Agents VI, Proceedings of the Sixth International Workshop on Agent Theories, Architectures, and Languages (ATAL-99). Lecture Notes in Articial Intelligence. Springer-Verlag, Berlin.",(2000).

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC