Data processing in fusion experiments with remote participation

Abstract

Remote participation in fusion experiments is becoming more and more widespread. From the point of view of the data processing that each experiment generates, remote participation increases the present needs of existing systems. On the one hand, these systems must be able to acquire the data from each discharge (with hundreds of channels and high-sample frequencies) and, on the other hand, they must be able to compress, analyze, and process data in as short period of time as possible. This would mean a decrease in network bandwidth use when sending results to remote investigators, a considerable saving of space in the local database, and a reduction of computation time in the server. Furthermore, in remote participation experiments, the system would allow each user to configure their own measures and analyses (downloading their own high-level programs in the system—writing in 4GL languages, for example in LABVIEW—and receiving the processed data as soon as possible). What is proposed as a solution for data processing in experiments with remote participation (that will be applied in the stellarator type device (TJ-II) of the Research Center for Energy, Environment, and Technology (CIEMAT), of Madrid Spain, is a distributed system that guarantees noise immunity and electromagnetic shielding (for example, based on several PCI (peripheral component interconnect) extensions for instrumentation (PXI) chassis with several digitizer boards in each one). It is not possible to process all data from the system’s own controller. Therefore, the following alternative will be presented: The design of the development of a processing board, that works in a peripheral PXI slot (nonsystem slot), running a real-time operating system (WNT+RTX, RT-LINUX,…), that communicates by PXI bus with the controller. In this way, the needs of data processing in remote participation experiments can be fulfilled.