Students' Perceptions and Self-Reported Experience of Using ChatGPT for Learning IP Subnetting in Higher Education: A Student-Centric Study

Abstract

This paper explores the use of artificial intelligence in computer networks, focusing on ChatGPT and its application to subnetting. The study examines how ChatGPT supports the calculation-based aspects of subnetting. Subnetting is an important topic in network management and security. A quantitative, questionnaire-based approach was used with 30 students enrolled in the Computer Networks course at the Department of Information Technology, Faculty of Humanities and Applied Sciences, University of Benghazi, Tocra, Libya. The questionnaire examined students’ difficulties with subnetting and their perceptions of using ChatGPT across key subnetting phases, including determining the assigned IP address space, identifying borrowed host bits, determining the number of possible subnets and hosts per subnet, calculating the new subnet mask, and determining the broadcast address and the usable range of the new subnets. The findings showed that subnetting was challenging for many students, with 53.3% (n =16) describing it as complicated. Time consumption was reported as a difficulty by 80.0% (n=24), while 66.7% (n=20) identified calculations as the most difficult aspect. In addition, 93.3% (n=28) of the participants reported that ChatGPT saved time when working on subnetting. Similar proportions explained ChatGPT 'uses in all of subnetting' levels. However, 46.7% (n=14) identified over-reliance on ChatGPT as a major challenge; also, 46.7% (n=14) rated its role in automating repetitive calculations as valuable. Overall, 90.0% (n=27) supported the use of ChatGPT as a supplementary learning tool and suggested some improvements to make it more useful for learning subnetting. These results indicate that ChatGPT can provide useful support for students working on calculation-based subnetting tasks.