Recent research has developed service-oriented solutions for reconfigurable hardware, with a top-down functional approach supported by virtual instrumentation systems like "NI LV" - National Instruments LabVIEW. The broad spectrum of Cloud-based management and deployment of such solutions - including resource-intensive "Cloud compilation" in case of using FPGA - have a great potential for ODL (Open and Distance Learning), as being distributed and easily interfaced with our "target user" that can be a trainee not very professional with HDL (Hardware Description Languages). The paper goes beyond consolidated solutions using NI LV like RIO (Reconfigurable Input-Output), aiming to approach also the processing, not at the level of modern SoC (Systems on Chip), but on cellular-"tissue"/"fabric" structures like CA (Cellular Automata) or IFA (Intelligent Fabric of Automata). The authors are presenting a demonstrator based on Xilinx Spartan family of FPGA, with prototypes of elementary processing machines built to be easily put together in scalable configurations. There are approached the memory segmentation solutions, oriented for the employment of communication data-bases, e.g. tables for flows identification, classification and routing in modern streaming. The specific FPGA configurations are oriented on the use of the micro-automata for digit processing, with predictive calculations like those using PCRE - Perl Compatible Regular Expressions ("regex") - in order to recognize the "digital fingerprint" of multimedia flows. There are discussed use-cases oriented to packet processing in groups of virtual machines that execute specific tasks of routing, fragmentation, encapsulation, coding etc. with an event-driven control that can be extended to "policies" that are able to be dynamically modified.

distance learning, virtual instrumentation, FPGA, Intelligent Fabric of Automata, streaming