Fault tolerant knowledge level inter-agent communication in open Multi-Agent Systems

Agent Communication Languages (ACLs) have been developed to provide a way for agents to communicate with each other supporting cooperation in Multi-Agent Systems. The importance of ACLs is especially highlighted in open systems (such as the Internet) where heterogeneous agents must be able to communicate and exchange knowledge with one another despite differences in hardware platforms, operating systems, architectures and programming languages. In the past few years many ACLs have been proposed for Multi-Agent Systems, such as KQML and FIPA ACL. The goal of these languages is to support high-level, human like, communication among agents exploiting Knowledge Level features rather than symbol level ones. Despite these efforts, an important issue in the research on ACLs is still open and concerns how these languages should deal with failures of agents in open Multi-Agent Systems. The main consequence arising from this lack is that it is not possible to develop Knowledge Level agents able to cooperate in an open MultiAgent System prone to agent failures. The Fault Tolerant Agent Communication Language (FT-ACL) presented in this Thesis addresses this issue. FT-ACL provides high-level fault-tolerant communication primitives and support for an anonymous interaction protocol which satisfies a set of well defined Knowledge Level programming requirements. We present a formal specification of FT-ACL and of the underlying agent architecture. To illustrate the language features we show how FT-ACL can be effectively used to write high-level executable specifications of fault tolerant protocols. Moreover, we show how FT-ACL can be used to support Web agent interaction as well as Web Service invocation on the Semantic Web. 1. Department of Computer Science, University of Bologna, Mura A. Zamboni 7, 40127 Bologna, Italy.