The basis of this context-sensitive computational approach will be a behaviour-based specification language. It will abstract from the realisation details in terms of architecture and implementation of the machine which is supposed to implement the specified behaviour and give the “designer” of a system the possibility to mainly specify behavioural aspects of the machine to be generated. The translation process finally automatically generates the complete machine in a rigorous manner. This will not be “yet another specification language” as it is inspired by the links between observable biological symmetries and their corresponding behaviour. Equipping this language with operational semantics will allow the automatic generation of an interaction machine which exhibits comparable power to biology.
This mapping will establish new connections in the field of algebra, coalgebra, and as a consequence in category theory. It will rely on work which links structure – represented by algebra – with its corresponding behaviour – represented by coalgebra. In particular so-called behaviour-realisation adjunctions will be investigated which were first recognised by Goguen in 1972 (J.A. Goguen, 1972a). They form the basis for deriving an appropriate operational semantics. Similar work conducted in 1981 by G.D. Plotkin (Plotkin, 2004), which layers coalgebra on top of algebra will support this task. Also the more general framework of universal (co)algebra, an important branch of category theory, will be relevant for this research. While the translation process from algebras into logic has been understood fairly well (Dini & Schreckling, 2008; Andréka et al., 2009), universal algebra and algebraic logic will offer powerful tools for comparing and studying logics. This tool will help us to cover new ground in the fields of coalgebraic logic and universal coalgebra and identify the required elements for the specification language described above.
In order to remain manageable and scalable the specification language will also have to rely on language features which allow to build hierarchies, known from the paradigm of object-oriented programming. Thus, the design of the language will also rely on prior research (Goguen and Malcolm, 1997). With the novel insights gained from the definition of this specification language, BIOMICS will contribute to novel approaches to software security and the security design domain. It will abolish the ex post “patching” paradigm and integrate security mechanisms only if specified by the context, thus not overstraining the user with highly specific information which he will not be able to understand.