Refining interfaces of communicating systems

Abstract

There are now several theories for describing and reasoning about the behavior of communicating systems, where the behavior of a communicating system is described in terms of its capabilities to perform communication actions in cooperation with its environment. In such theories, preorders or equivalences are defined as criteria for when one system is an acceptable substitute or implementation of another. Existing theories of communicating systems define preorders or equivalence relations only between systems with identical sets of communication actions. In many practical design situations, however, it may be desirable to refine a system by changing its set of communication actions. We present a simple method for carrying out such refinements. The method is first formulated in a general setting, and then elaborated in more detail in the trace model and a simple version of the failure model. We illustrate the usefulness of our method by an application to 1.451, an ISDN access protocol.

Secure Interoperation of Secure Distributed Databases

“…Other than our own papers on the subject, some of the best examples are Brinksma, Jonsson, and Orava's work on connector refinement [2], work on component refinement [3], and Philipps and Rumpe's work on refinement of information flow architectures [12]. However, work on relating models of connector behavior at different levels of abstraction by Abowd, Allen, and Garlan [1], Luckham, et al [5], and Moriconi and Qian [6] is closely related.…”

SDTP: a verified architecture for secure distributed transaction processing

“…Other than our own papers on the subject, some of the best examples are Brinksma, Jonsson, and Orava's work on connector refinement [2], work on component refinement [3], and Philipps and Rumpe's work on refinement of information flow architectures [13]. However, work on relating models of connector behavior at different levels of abstraction by Abowd, Allen, and Garlan [1], Luckham, et al [5], and Moriconi and Qian [6] is closely related.…”

“…Dataflow-level SADL description of X/Open architecture tation 2. A priori, it was not known what percentage of the design history encoded in the X/Open DTP hierarchy could be reused in the SDTP hierarchy.…”

“…A priori, it was not known what percentage of the design history encoded in the X/Open DTP hierarchy could be reused in the SDTP hierarchy. We therefore developed implementation-level descriptions of the three abstract ar-2 The bindings of the variables in the transformation can, and usual do, change. But the bindings in the replay can be automatically determined from the bindings in the original derivation by tracking the correspondence between objects in the two hierarchies as the new hierarchy is constructed from the old.…”

Abstract Interactions and Interaction Refinement in Model-Driven Design

“…Several authors approach interaction refinement from a pure formal perspective (e.g., [5], [6]). We believe that, in many cases, these approaches make simplifications at the cost of the usefulness of the formal model for pragmatic engineering purposes (as argued in [23]).…”