Semantical aspects of an architecture for distributed embedded systems

Abstract

We investigate the formulation of a formal semantics for the industrial software architecture Splice. In this paper, we present a set of basic Splice interaction primitives that is both powerful and easy to implement. We define a semantics for this language based on a conceptual global dataspace. It is shown that the semantics is equivalent to an implementation-biased semantics where each process has its own local dataspace and communication is established by means of asynchronous message passing. Hence, our language allows both convenient reasoning using a global dataspace and efficient implementation by means of distributed dataspaces. The equivalence result is checked mechanically by means of the inter* active theorem prover PVS.

Semantic models of a timed distributed dataspace architecture

“…In previous work on a denotational semantics for Splice [3], the semantics of local storages was inconvenient for compositional verification; it uses process identifiers and a partial order of read and write events with complex global conditions. In more recent work on the verification of Splice-systems [16], we used a complex denotational semantics with environment actions.…”

Formal verification of replication on a distributed data space architecture

“…Several formalizations of (fragments of) Splice already exist. We mention work on the process algebra SPA [9], the µCRL tool set [10,21] and a formalization in the higher-order logic of the theorem prover PVS [3]. Related work on the operational semantics of Linda and JavaSpaces has been presented in [6].…”

“…The logic of PVS is a typed higher-order logic in which we express the semantics of Splice. Earlier work on a denotational semantics for Splice [3] showed the equivalence of a global data space view and an implementation with local data spaces for a carefully selected set of Splice-primitives. This result, however, does not hold for the full Splice architecture, which is essentially based on distributed storages.…”

“…The semantics for local storages of [3] seems not very convenient for verification; it is based on a partial order of read and write events, with complex global conditions. It also uses process identifiers, which we would like to avoid if possible.…”

Formal verification of replication on a distributed data space architecture

“…Our semantic models of Splice are based on the models described in (3,21). They are less detailed than (10] in order to facilitate verification.…”