Massively Multi-Agent Technology

With the proliferation of consumer devices, computation is becoming truly ubiq- tous. Multi-agent systems hold the promise to enable utilization of these compu- tional resources for ground-breaking new applications. However, as the number of agents involved in multi-agent computations rises, traditional approaches of bui- ing multi-agent systems fail to scale. Massively multi-agent systems address this challenge of scale. Massive, in this context, is meant to capture the complexity of such systems, which precludes decision making to focus on individual agents. Agents making decisions have to reduce the complexity of the space in which they must decide. The focus, therefore, is on the approaches to manage the compl- ity, implications of this abstraction, as well as on identifying ways of applying the paradigm to problems. Papers appearing in this volume have been selected from three inter- tional workshops held in conjunction with the International Conferences on - tonomous Agents and Multiagent Systems (AAMAS) held in 2006 and 2007: First International Workshop on Coordination and Control in Massively Mul- agent Systems (CCMMS 2007); and the jointly held Second International Workshop on Massively Multi-agent Systems (MMAS 2006) and the Third Int- national Workshop on Challenges in the Coordination of Large-Scale Multi-agent Systems (LSMAS 2006).

Klappentext

This book constitutes the thoroughly refereed post-workshop proceedings of the joint AAMAS 2006 International Workshops on Massively Multi-Agent Systems, MMAS 2006, and on Large scale Multi-Agent Systems, LSMAS 2006, held in Hakodate, Japan, in May 2006, and of the International Workshop on Coordination and Control in Massively Multi-Agent Systems, CCMMS 2007, held in Honolulu, HI, USA, in May 2007 as associated event of AAMAS 2007.

The 13 revised full papers presented were carefully selected from the presentations made at the 3 workshops fall in 4 broad categories, presenting a snapshot of current research. Included are implementation strategies addressing coordination in the space of spatial and temporal distributed systems; approaches to deal with complexity to make decisions such as task allocation and team formation efficiently, by creating implicit or explicit encapsulations; and finally, a diverse range of applications to which these approaches may be applied, from large-scale agent based simulations to managing different types of networks to image segmentation.

Inhalt
Massively Multi-agent Infrastructure.- A Platform for Massive Agent-Based Simulation and Its Evaluation.- Distributed Coordination of Massively Multi-Agent Systems.- Community-Based Load Balancing for Massively Multi-Agent Systems.- Task Allocation in Massively Multi-agent Systems.- A Study of Coordinated Dynamic Market-Based Task Assignment in Massively Multi-Agent Systems.- Dynamic Pricing Algorithms for Task Allocation in Multi-agent Swarms.- Teams and Organization.- Evolve Individual Agent Strategies to Global Social Law by Hierarchical Immediate Diffusion.- Team Formation Strategies in a Dynamic Large-Scale Environment.- Using Swarm-GAP for Distributed Task Allocation in Complex Scenarios.- Applications of Massively Multi-agent Systems.- Mousetrap 2.0.- Application of a Massively Multi-Agent System to Internet Routing Management.- An Agent-Based Approach for Range Image Segmentation.- Coordination in Disaster Management and Response: A Unified Approach.- The Scalability of an Environment for Large-Scale Sensor-Networks.