Transactional Programming for Distributed Agent Systems

A new multiagent programming paradigm based on the transactional logic model is developed. This paradigm enables us to construct a distributed agent transactional program (DATRAP). Such a construction is carried out in two stages: first expressing a program into a production rule system, and then converting the rule applications into a set of transactions on a database of active objects represented using high-level data structures. The formal specification and refinement calculus are key features in the development of a DATRAP. We also indicate how to specify granularity of parallelism and also achieve several types of parallelism. One can associate with a DATRAP two different types of execution semantics called set-based and instance-based semantics. We also show how to prove correctness of DATRAP, achieve maximal concurrence and reduce the complexity of a distributed program