Executable Specifications for Distributed Embedded Systems

Reusable components, executable specification language, real-time executive, communication task, trial architecture

Rapid prototyping of embedded distributed systems specifications can be approached with the help of prototyping hardware components and an executable specification language. Statements of such a language should cover process management, timing, and communication commands interpretable by real-time executive and communication task services of every node prototype. That language can be made executable by a technique employing attribute grammars and either a macroprocessor or Prolog. The overall prototyping technique consists of the following steps: (i) definition of a concrete specification language, which fits the particular application and implementation domains, and description of its semantics by an attribute grammar; (ii) description of this attribute grammar either by text macros of a powerful enough macroprocessor or by a definite clause grammar in frame of Prolog, aiming at a straightforward translator prototype implementation; (iii) design of a trial architecture and identification of its reusable components; (iv) specification of a target communication system using the trial system architecture and the devised specification language; (v) macroprocessor or Prolog-driven expansion of that specification into an executable code fitting the trial system architecture; and (vi) final compilation and execution or interpretation of the expanded code complemented by reusable system components of the trial architecture. While the steps according to sub (i), (ii) and (iii), discussed in more detail in sections 2, 3 and 4 respectively, can be realized once for a particular application domain, the rest of activities, i.e. sub (iv), (v) and (vi), briefly mentioned in section 5, have to be contributed to every new application.