Research on multiplexing of remote experiments and training systems using container virtualization technology
While the number of online classes is increasing due to the spread of COVID-19 infection, it is difficult to implement practical classes using special equipment such as electrical circuit practice and experiment remotely.
In a remote learning system using actual equipment, the number of students is large and the hardware scale of individual systems (servers) is large, which increases the time cost and hardware resources required to build and multiplex the experimental environment. This research aims to reduce both hardware and management costs by multiplexing experimental training systems on containers. Therefore, this research will minimize server hardware by using container virtualization technology for server construction. Container virtualization is a technology that creates an area called a container on the host OS and runs one container as one virtual machine, allowing the virtual machine to run with fewer resources than other virtualization techniques.
In this research, a remote experiment system is constructed to learn digital circuit design with FPGA (Field Programmable Gate Array) board. Learners access the server container from a client PC via a browser to code, compile, and write a circuit configuration files to the practice board. The results of the experiment can be confirmed by immediately displaying the video captured by the FPGA board on the browser. The system is designed to enable this type of training. Build as many training containers as the number of training boards connected to the server using Raspberry Pi 3 to which the training boards and cameras are connected via USB. Docker will be used to build the training containers. It will implement and deploy a web server, web application, and FPGA development environment.
After introducing the system, we will ask students to use this remote training system and evaluate usability and comprehension through questionnaires. In addition, the system will be compared with other virtualization methods and evaluated for resource reduction.