Tightly Synchronized Distributed Measurement and Event Triggers

In this paper we describe the principles, design and implementation of a method to reliably perform tightly synchronized data acquisition and event triggering in a distributed control system. In parallel to event triggers, our timing system facilitates maintaining tightly synchronized clocks to achieve high quality timestamps which can be tailored to the event triggers if wanted. We introduce concept of priority-based clocks to handle timing hardware failures and to provide flexibility and commonality of time management in a large scale distributed control system. Contribution of this work is mainly: 1) Integration of our hardware-assisted data timestamping method into a control and data acquisition framework 2) Priority-based clock mechanism which overcomes hardware clock failures and allows computing nodes continue monitoring and collecting distributed events with a coarser precision. This framework allows us to build a tightly synchronized distributed control and measurement system without mandating the use of any special hardware.