The formal verification of software (programs and protocols) and hardware (chip design) is growing in importance as systems become more complex and production processes more expensive, but testing is widely recognised as being insufficient to guard against malfunctioning. For industrial applicability it is essential that some unity emerges from the competing theories of concurrency, and that the verification process be supported by software tools. The CONCUR Action has explored the relationships between algebraic concurrency theories and has investigated their possible integration. Such algebraic methods deal with the constructs in a program and its specification according to mathematically sound rules in order to show that the program implements the specification. This process is supported by a range of software tools. s.
Concurrency theory (CONCUR) is important for the formal verification of concurrent and distributed programs. The different algebraic approaches to concurrency theory have been explored, and investigations made into the possibilities of integrating these approaches, or incorporating the features of 1 system in another. Appropriate software tools have been developed.
CONCUR has obtained a unified view on the different schools involved in process algebra, classified and initiated the development of a wide range of process algebra extensions and worked towards a common basis for tool design and implementation.
APPROACH AND METHODS
The relationships between and relevant roles for algebraic versus model-based theories of concurrency have been investigated. Methods pioneered in one group have been applied to problems originating in another, and the results compared. In addition to cooperating at the theoretical level, the partners are collaborating through the development, use, and comparison of software reasoning tools, as experience has shown that the development and use of software tools acts as a catalyst for theoretical advances.The tools are being applied in case studies relevant to the European software industry. The work of CONCUR will be continued in a new ESPRIT project 7166 (CONCUR II).
PROGRESS AND RESULTS
-obtained a unified view on the different schools involved in process algebra
-classified and initiated the development of a wide range of process algebra extensions
-worked towards a common basis for tool design and implementation.
The CONCUR Action is expected to lead to the development of accepted standards and thus a higher level of compatibility between the different formalisms and approaches in algebraic concurrency semantics, allowing interusability of results. CONCUR has alsoenabled the development of a common tool-set, which will provide a strong basis for further software development in the area of the verification of concurrent systems.
1009 DB Amsterdam
BN1 9QT Brighton
OX1 3QD Oxford