Journal of Universal Computer Science (Apr 2019)

Cost Estimation of Blended Learning Course Delivery Through Public Cloud

  • Vladimir Ciric,
  • Ivan Milentijevic,
  • Oliver Vojinovic,
  • Nemanja Manic

DOI
https://doi.org/10.3217/jucs-025-04-0310
Journal volume & issue
Vol. 25, no. 4
pp. 310 – 333

Abstract

Read online Read online Read online

There is a lot of educational institutions that own blended learning platforms, and even more that do not. The era of cloud computing offers a possibility for those who do not own such platforms to utilize public cloud resources for courses delivery. The goal of this paper is design and cost estimation of flexible blended learning platform suitable for cloud implementation. In this paper we present technical requirements and create a mathematical model which allows one to estimate the cost of course delivery through public cloud. As a basis for the cost estimation, a model of blended learning course and a platform that supports the model will be proposed. The platform will be based on open source software, and it will be designed to be suitable and simple for cloud implementation. The proposed blended learning model will be applied and the platform will be implemented on the public cloud on the example of computer networking course. In order to estimate the cost of public cloud hosting, a stress test will be performed varying both virtual hardware and the number of students. Three different virtual hardware configurations will be evaluated with up to 400 students. The analysis of the stress test results will be given, along with the recommendations regarding the virtual hardware in the respect to the audience. The cost function in the respect to the number of enroled students will be given, too. It will be shown that the proposed course delivery model is low-cost, while it is based on the state of the art learning tools. Moreover, the students' achievements will be discussed in the example of two different cohorts of students.

Keywords