Designing Agent-Oriented Systems by Analysing Agent Interactions

We propose a preliminary methodology for agent-oriented software engineering based on the idea of agent interaction analysis. This approach uses interactions between undetermined agents as the primary component of analysis and design. Agents as a basis for software engineering are useful because they provide a powerful and intuitive abstraction which can increase the comprehensiblity of a complex design. The paper describes a process by which the designer can derive the interactions that can occur in a system satisfying the given requirements and use them to design the structure of an agent-based system, including the identification of the agents themselves. We suggest that this approach has the flexibility necessary to provide agent-oriented designs for open and complex applications, and has value for future maintenance and extension of these systems.

Trust-By-Contract: Modelling, Analysing and Predicting Behaviour in Software Architectures

Architecture description languages (ADLs) are used to specify high-level, compositional views of a software application. ADL research focuses on software composed of prefabricated parts, so-called software components. ADLs usually come equipped with rigorous state-transition style semantics, facilitating verification and analysis of specifications. Consequently, ADLs are well suited to configuring distributed and event-based systems. However, additional expressive power is required for the description of enterprise software architectures – in particular, those built upon newer middleware, such as implementations of Java’s EJB specification, or Microsoft’s COM+/.NET. The enterprise requires distributed software solutions that are scalable, business-oriented and mission-critical. We can make progress toward attaining these qualities at various stages of the software development process. In particular, progress at the architectural level can be leveraged through use of an ADL that incorporates trust and dependability analysis. Also, current industry approaches to enterprise development do not address several important architectural design issues. The TrustME ADL is designed to meet these requirements, through combining approaches to software architecture specification with rigorous design-by-contract ideas. In this paper, we focus on several aspects of TrustME that facilitate specification and analysis of middleware-based architectures for trusted enterprise computing systems.

Technology diffusion: analysing the diffusion of agent technologies

Despite several examples of deployed agent systems, there remain barriers to the large-scale adoption of agent technologies. In order to understand these barriers, this paper considers aspects of marketing theory which deal with diffusion of innovations and their relevance to the agents domain and the current state of diffusion of agent technologies. In particular, the paper examines the role of standards in the adoption of new technologies, describes the agent standards landscape, and compares the development and diffusion of agent technologies with that of object-oriented programming. The paper also reports on a simulation model developed in order to consider different trajectories for the adoption of agent technologies, with trajectories based on various assumptions regarding industry structure and the existence of competing technology standards. We present details of the simulation model and its assumptions, along with the results of the simulation exercises.