A Systematic Approach for Identifying Operational Relationships in Embedded Computer Control Systems DeJiu Chen, Martin Törngren Department of Machine Design, Royal Institute of Technology SE-100 44 Stockholm, Sweden chen@md.kth.se, martin@md.kth.se Abstract Embedded computer control systems are a kind of safety critical real-time system, formed by a synergetic integration of various functions, technologies, and engineering disciplines. To modularize the design and to define the embodiments of functional modules in terms of software components, a more precise modeling of systems and support for assessments of modularity and related tradeoffs in designs is necessary. However, little support exists in this area. Existing system models and modularity metrics such as coupling and cohesion are often either too general or too specific by targeting only a specific implementation technology. This paper proposes a meta-level model for the systems and a fine-grained classification of relationship patterns established by communication, synchronization, and implementation. The work emphasizes a system perspective and aims to support a more precise assessment of coupling, considered as one of the most important criterion for component creation and integration as well as structuring. 1. Introduction Embedded computer control systems (ECS) are digital computer-based systems for advanced control (e.g. motion control), diagnostics, and monitoring in machinery [1]. Typical application areas include vehicles, avionics, and robotics. Because of the dynamics under control, such systems differ from other general-purpose computer systems in the aspects of real-time and safety criticality. Given the increasing complexity, time-to-market requirements, and expectations for high qualities and low cost, a paradigm shift from ad hoc to systematic engineering is necessary. One central issue is concerned with how to define, model, manage, and represent various system aspects and parameters so that decisions can be made using models. Another issue is how to parameterize nonfunctional properties for more precise predictions and tradeoffs in design. Currently, while multiple modeling and analysis techniques are exploited in the development of ECS [3], little support exists in the area of holistic system description, modularization, and assessments of nonfunctional qualities (e.g., modifiability). Existing solutions are often too general or too specific by targeting only a specific implementation technology or a particular aspect. In engineering practices, system descriptions are often ad hoc using block diagrams or in software, providing oversimplified or partial information with respect to requirement mapping, component definition, as well as dependency and consistency of solutions [18]. Often, systems fail to meet the intended quality goals because of an overlooked functional assumption of components (e.g., value range and failure modes), or an incomplete or imprecise assessment of the system- wide effects of a particular component. Component-based software engineering (CBSE) has proven effective in generic software domains. Instead of building systems from scratch, this approach aims to integrate existing solutions, normally using Object-Oriented technologies, hence reducing complexity and increasing efficiency of system development. On the other hand, current CBSE is still insufficient for ECS. One reason for this is that existing approaches mainly target generic software systems and focus on the implementation aspects such as packaging of binary components and middleware for interoperability (e.g., CORBA and EJB). See e.g., [17]. In ECS, software components constitute the embodiments of functional modules and issues concerning timeliness (e.g., end-to-end timing) and reliability needs to be explicitly covered. The paper includes six main sections. Section 2 discusses the aim and approach. Section 3 introduces the system meta-model. Various operational relation- ships are described in section 4. Due to limited length of the paper, we focus on the reasoning behind the classification. Finally, the related work and conclusion are given in Section 5 and 6.