Agent-oriented computing for distributed systems and networks

Recently, we can observe an increasing interest in agent-based computing as a method for design and implementation of distributed systems and networks. In particular, software agents find their way to dynamic and complex environments where they provide the, much needed, robustness and flexibility. As a cause and effect of this process, the field of agent technology is rapidly maturing. The gathered body of knowledge allows in-depth reflection on the nature of designing and implementing agentbased solutions. For instance, experiences gathered by the Telecom Italia guide the development of the JADE agent platform. As a result, significant advances can be seen in (1) methods for designing agent-based systems, (2) agent-oriented formalisms for modeling distributed systems and networks, (3) agent tools/ platforms/ environments. The goal of this special issue is to discuss some of them. Papers found in this collection originate from two sources. First, we have invited best papers from the Agent-Based Computing: from Model to Implementation (ABC:MI’2011) workshop and selected papers from the Intelligent Distributed Computing (IDC’11) conference, to be extended to the journal format. Second, an open call for papers had been published. As a result a total of 27 papers have been received. After a stringent, multi-round, refereeing procedure, the best 7 papers have been accepted. These papers were conceptually divided into two groups. In the first, we find three papers dealing with agent system design and implementation. In the second, we have placed four application use case studies, illustrating how software agents can be actually used. Let us now briefly summarize content of these papers. The first paper, entitled “Expanded Borel Cayley Graphs (ExBCGs): A Novel Communication Topology for Multi-Agent Systems,” was authored by Dongsoo et al. It addresses the low end of the software stack and concerns network-related issues involved in agent communication. Specifically, in their paper, authors' introduce Expanded Borel Cayley Graphs (Ex-BCGs) as a communication topology for multi-agent systems. The Ex-BCGs preserve the communication efficiency of Borel Cayley Graphs (BCGs), but do not require new generating parameters for new sizes. Furthermore, the Cut-Through Rewiring (CTR) algorithm, used in the Ex-BCGs as a fault-tolerant methodology against node-failures, becomes an efficient way of recovering connections for neighbors of pruned nodes. Specifically, the experimental results show that the resized networks, obtained from the Ex-BCGs by the CTR algorithm, are more fault-tolerant with robust connectivity, efficient topological properties, and faster convergence speed than these based on BCGs. One of the well-recognized approaches to the design of agent system is by realizing some of its goals through agent mobility.