Distributed priority synthesis using knowledge

For distributed computing, orchestrations along predefined communication paths are used to obtain agreement between system components on the next chosen transition. Although the communication overhead can be high, it can be efficiently reduced by the introduction of knowledge, which provides each local component imperfect view on the global state during run-time. In this paper, given a safety criterion, we formulate the problem how to automatically orchestrate components in a system using stateless precedences between actions under the assist of statically computed knowledge. If the system is diagnosed as unsafe, the use of knowledge can be integrated in the synthesis process to enlarge the set of legal fixing candidates. These new solution candidates may disrespect predefined communication paths but their defined priorities are still guaranteed to be deployable.