Design and implementation of a smart learning environment for teaching computer networking

The use of computers and computer technology has expanded rapidly since the mid-1990s, in part due to the ‘Internet revolution’ (Tronco, 2010) and also due to an almost ubiquitous availability of technological innovation related to computing (e.g., see Premsankar & Di Francesco, 2020). Many teachers at the tertiary level, however, remain entrenched in a traditional teaching approach: ‘commonly structured with the assumption that knowledge resides in the teacher and textbook, that knowledge is static and that teaching is the process of transferring that knowledge from the source (the teacher) to the students’ (Lightfoot, 2005 in Albakry et al., 2016, p. 9). This approach, as applied in integrating technology into tertiary-level instruction, appears to rely on the ‘knowledge reproduction model’, an objectivist paradigm drawn from Bernstein (2000). Lightfoot (2000) has argued that this approach results in static teaching (and passive learning), which involves lectures, handouts, drills and practice, all within structured classroom activity during regulated hours.

An awareness of the ineffectiveness of such passive learning approaches has led to the promotion of ‘knowledge-building’ approaches, such as constructivist approaches advocating learner discovery (Zajda, 2018) and approaches structured around learner-centred environments where students self-determine their own progress—the so-called active learning approaches (e.g., see Lightfoot, 2005). Active learning has been reported in the literature as an effective teaching approach in science, engineering and mathematics fields, where students actively participate in the learning process and construct their knowledge by participating, discovering, processing, exchanging, applying information and receiving feedback (Freeman et al., 2014). Along with the development of active learning approaches have come better understandings of its limitations and a move towards guided discovery and well-tested improvements in instructional design (see e.g., Kirschner et al., 2006; Nordmann et al., 2020).

Therefore, an active, rather than passive, learning approach may be optimal for students to obtain technical and complex concepts needed for expertise in computer networking. Ideally, teaching approaches applied in computer networking courses should enable active participation of students in the learning process (Goldstein et al., 2006; Hailin, 2017; Mikac & Horvatić, 2019; Sllame & Jafaray, 2013; Vinay & Rassak, 2015; Zhang et al., 2012). Such an approach would enable students to acquire the requisite knowledge, skills and experiences, both theoretical and practical, through learning activities. Most importantly, such activities, if designed appropriately, should enable students to play an active role in obtaining the knowledge, skills and experiences needed to configure and manage real-world computer network scenarios (see e.g., Bui et al., 2020; Woolcott et al., 2022).

The implementation of active learning approaches in teaching computer networks is not, however, straightforward, especially in the context of online education. Lectures on computer networking are often structured following the Open System Interconnection (OSI) layering model (see Zimmermann, 1980), either bottom–up or top–down, which allows interoperability between components produced by different computer manufacturers. The structuring technique of layering remains in common use, with the physical media generally in the first layer and other media layers or entities, each belonging to one system (Khan, 2020).

A common teaching approach in F2F and blended learning environments is to embrace an instructional design where practical and lab sessions are scheduled between the theory presented in lectures, so that the lecture content is complemented by learning activities that strengthen the network concepts within each layer. For example, students could have hands-on experience on different types of network cables and connectors in a lab session following lectures on the physical layer (i.e., layer 1 of the OSI model), while the next lab session might cover the data-link layer (i.e., concepts from layer 2), for example, by analysing an Ethernet frame, again following theory presented in the lecture format.

Although lectures, practical sessions and/or structured classroom activities remain the preferred teaching approach in computer networking, both the approach and the instructional design used remain problematic in promoting an active learning process. The main issue appears to be that learning activities are contextually disconnected within the instructional design, regardless of the teaching approach used. For example, students may not find a contextual link between the network cables they used in one lab activity and the Ethernet frame they analysed in another lab activity. The lack of contextual connection between the activities means that students often fail to transfer knowledge, skills and experiences obtained in each activity to scenarios, as required, to solve a complex network problem.

In transitioning to online education, a commonly used strategy is to replace F2F lectures with online and/ or pre-recorded video sessions. Instead of resolving the issue of contextualisation, however, this strategy adds another layer of difficulty for the students. For example, F2F lectures and lab sessions are generally conducted sequentially in a single session lasting from 2 hr to 3 hr, or alternatively, as a 1-hr lecture and a separate 2-hr lab session. This time duration has proven onerous for students who are undertaking subjects that are fully online, and many of them find it difficult to stay engaged in online settings, especially when their previous engagement was enhanced by an F2F setting. This phenomenon is called ‘Zoom fatigue’ (Nadler, 2020). In addition, the use of pre-recorded activities does not necessarily provide a deep understanding of the content as many students only follow the step-by-step guide just to finish tasks as quickly as possible with minimal effort (Goldstein et al., 2006).

Like many other courses in science and engineering fields, practical laboratory (lab) work is considered to be a critical component of courses in computer networks (Abdullah & Ehsan, 2012; Dobrilovic et al., 2012; Hailin, 2017; Javidi & Sheybani, 2008; Pan, 2010; Ruiz-Martinez et al., 2013; Sarkar, 2006a). A common teaching approach for practical lab work in such courses (F2F) is to use instructor-led lab sessions, for example, as workshops or tutorials, where students are asked to follow step-by-step instructions to configure network devices and protocols, then report and explain what they observe.

Although such a lab session can be viewed as an activity, for example, in simulating a real-world example, it is not active learning as students are taking only a passive role—students can follow instructions without understanding the reason behind each instruction or without the interaction required for active learning. As a result, students may not obtain a deep understanding of the content even after completing all learning activities (Zhang et al., 2012). In F2F settings, teachers may use in-class discussion to address this problem, but such discussion is difficult to implement in online settings, especially if the learning context is asynchronous. In addition, students may lose their motivation more quickly in engaging in activities online, especially when they experience technical problems with no immediate assistance and feedback (Al-araibi et al., 2019; Almaiah et al., 2020; Sani, 2020).

The learning environment may become more complex with the loss of the advantages of F2F interaction when transitioning to fully online education. For example, in F2F or blended learning environments, students generally learn about network cabling systems, physical routers and switches, while physically visiting a server room or data centre, or they may learn about security and regulations while working in the network facilities. These options, however, are not available online. Although visualisation and simulation have been shown to be effective approaches to teaching the theory of networking (Goldstein et al., 2006; Jovanović & Zakić, 2018; Lampi, 2013; Rashid et al., 2019; Ruiz-Martinez et al., 2013), it remains difficult to use these approaches to convey advanced networking concepts that would otherwise require hands-on experience (Pan, 2010; Sarkar, 2006a, 2006b). For example, working in a simulated scenario provides only a superficial experience compared to working with the real network and the real network devices (Shimba et al., 2017). Some industry publications suggest that simulation tools such as Packet Tracer (2020) should be used only as an additional learning support.

A major challenge, therefore, is to develop and implement teaching approaches and instructional design that allows for both effective teaching of theory and high-quality online lab activities that will motivate the students and enable them to obtain a deeper understanding of theoretical concepts. The teaching approaches and instructional design should, ideally, allows teachers to assess whether or not students have obtained the required knowledge, skills and experiences necessary for mastery (Prvan & Ožegović, 2020).

The following sections outline a teaching approach and instructional design, implemented as a fully online learning environment in 2020, that appears to answer this question, comparing effectiveness with that of the teaching approaches and instructional design used previously in a blended, but predominantly F2F, learning environment.

Identifying and applying effective instructional design, along with appropriate teaching approaches to enact that design, is critical to the success of online learning for computer networking students (Costley et al., 2017; Dabbagh & Bannan-Ritland, 2005). Teaching approaches that are appropriate for online learning environments should engage a set of teaching principles, such as being available and responsive to students, engaging and interacting with students, and providing prompt feedback (Adinda & Mohib, 2020). Such approaches can be built around instructional design that has been reported to work in online teaching, which should include such features as dividing the teaching content into smaller units to help students focus, strengthening students’ active learning ability outside of class, combining online learning and offline selflearning effectively and making plans for teaching when unexpected problems occur (Bao, 2020). This section explores how the approaches for teaching computer networks in online settings can be derived from such instructional design.

As mentioned earlier, conventional teaching approaches in computer networks often use an instructional design that follows either a top–down or bottom–up sequence based on the OSI model. The difficulty that ensues is related to a lack of contextual connection between the lectures if each lecture (and associated lab) focuses on a specific layer—students often do not see the computer network as a whole. As a result, students often fail to bring together the appropriate knowledge, skills and experiences related to each layer such that they can solve a complex problem that engages the entire network. It does not necessarily follow that breaking the lecture-plus-lab block into smaller mini-blocks is a logical progression in adapting the ‘knowledge reproduction model’ of instructional design to online delivery.

This article, therefore, explores an instructional design that abandons the current top–down or bottom– up layers approach in computer networking, instead designing the instruction to begin at the network layer of the OSI model. There are several advantages to this novel design. Firstly, it allows teaching approaches, in the initial online sessions, that encourage critical discussion regarding the concept of computer networks as a network of nodes, each associated with a unique identifier, that is, a network address. This design provides the foundation for the teacher to set up the learning context for other networking topics such as routing, switching and transport, thereby creating the contextual links missing in the current design and approaches. Once the contextual link between the topics is established, subject content can be divided into smaller components, which suit the shorter engagement favoured in online learning while, at the same time, maintaining content connectivity.

Secondly, this novel instructional design helps familiarise students with the twin concepts of ‘Internet protocol’ (IP) and ‘IP addressing’ early in the subject learning sequence, allowing them time to fully assimilate these foundational concepts, an assimilation that students often find difficult (Dulaney & Harwood, 2012). These twin concepts are critical to the success of teaching computer networks, particularly since IP addressing is known to be one of the hardest topics to master in an introductory computer networking subject (Dulaney & Harwood, 2012). This design also provides a way for the teacher to progress the students easily to a focus on network applications or network communication technologies, while not losing the contextual link, effectively enabling a teaching approach that models the OSI by either up or down movement across layers. This type of instructional design requires modification to the learning activities and assessments, and these are discussed in the next section. An instructional design that links learning activities and assessment (that examines whether teaching and learning was effective) is critical in promoting active learning in online learning environments (Freeman et al., 2014).

Learning activities in many online courses are designed to stick to the following sequence: 1) Students are requested to read textbook materials or watch video lectures of learning concepts and then 2) to answer a set of questions to check their understanding. Students may also be asked to participate in online chat discussions after reading or watching the materials. Although this approach works in some cases, it does not work well for teaching computer networking, largely because the approach does not facilitate the acquisition of the knowledge, skills and experiences needed for problem solving, either within layers of the OSI or across the network—although students may understand the concepts, they often fail to apply them successfully in a practical scenario (Elias & Ahmad, 2014).

The design of instruction for online learning activities, therefore, should facilitate student learning of problem solving, preferably in a collaborative learning environment. Ideally, teachers should take the role of mentors or supporters (the guide on the side) in this student-centred approach, rather than their primary instructor role (the sage on the stage) seen in the teacher-centred traditional approach (Dabbagh & Bannan-Ritland, 2005; King, 1993). Some success has been obtained with problem-based learning (as a type of active learning) in designing instruction that embraces the practical (lab) activities for computer networking (de Oliveira & dos Santos, 2018; Linge & Parsons, 2006), but there are a number of approaches that can be applied with student-centred learning in mind (Karanja & Grant, 2020).

The instructional design sequence presented in Figure 1 describes how a learning activity can be structured to engage students with problem-based learning. In this sequence, students are asked to study a given problem and answer a set of questions to determine if they understand the problem and can present a correct solution. Online discussion, scaffolding activities and online support session are provided as needed, and the sequence has inbuilt loops to allow the process to be repeated as needed. The online discussion is provided so that students can collaborate in scoping possible approaches to solving the problem, as well as identifying any missing data or processes required for problem solution. Scaffolding activities and online support are provided to help students navigate the problem-solving process. This sequential design dictates a specific teaching approach—the role of the teacher in this sequence is to provide guidance and feedback for solving the problem and to direct students to the resources that may assist in problem solution, such as missing data. Teachers, therefore, need to be available and responsive to students and to provide prompt feedback.

Figure 1.

Along with learning activity design, however, assessment design is a burning issue in any transition from F2F or blended delivery to online delivery. Assessment is traditionally focused on invigilated pencil-and-paper exams that may not offer either a fair or authentic assessment (de Olivera & dos Santos, 2018; Marcus et al., 2021)—they sometimes fall short in assessing “critical competencies needed for solving real-world tasks that professionals typically face in their field” (Hassan, 2020, p. 1). The assessment in the previous learning activity sequence can take a number of alternative forms, such as a quiz related to the correct solution, a report (e.g., a reflection on the process) or a video presentation of the problem solution. A suggested solution to the problem can be provided to help students before moving on to the next learning activity.

A major criticism of assessment design in computer networking is that it does little to prevent the potential for academic misconduct, such as through cheating and plagiarism (the latter may be more common in computing than in other courses, see e.g., Fraser, 2014). For example, the use of a non-invigilated (or non-proctored) assessment, such as a ‘take-home exam’, may not allow the surveillance needed to make sure that a submitted assignment was a genuine attempt, that is, that it was done by the submitting student. Contract cheating (outsourcing an assessment to a third party), or some other form of non-genuine response, is possible for many assessment types, with an Australian study suggesting that students perceive multiple-choice exams, assessments with short turnaround times, and assessments that are heavily weighted to be the most likely to prompt some form of cheating (Bretag et al., 2019).

Bloxham and Boyd (2007) recommend nine assessment strategies including, changing assessment tasks regularly, spreading assessment (including drafts) across a sequence of nested tasks, drawing on recent events, and including a student’s personal experiences, such as in the form of a personal reflection, to minimise plagiarism. Bretag et al. (2019) suggested four assessment tasks that were the least likely to promote contract cheating: 1) in-class tasks, 2) personalised and unique tasks, 3) vivas (viva voce or spoken assessment), and 4) reflections, for example, on work placements or practical activities. Sotiriadou et al. (2019) suggested scaffolded assessment tasks that include real-world scenarios and interactive orals to help prevent academic misconduct. The findings from these studies suggest that assessment in online computer networking courses should be designed as a sequence of nested small and light-weight tasks that involve reflection on an approach to a real-world networking problem. Assessments within this sequence should include personalised and unique tasks that can be verified through interactive oral examination or recorded video sessions.

Simulation and virtualisation have been proposed in the literature as cost-efficient approaches to addressing the lack of access to physical networking facilities (i.e., in the real world) (Jovanović & Zakić, 2018; Montagud & Boronat, 2014; Prvan & Ožegović, 2020; Shimba et al., 2017; Sllame & Jafaray, 2013; Yalcin et al., 2015). Simulation and virtualisation are known to promote active learning (Wan et al., 2011), and early studies of Dale (see e.g., Dale, 1969) suggested that students could remember up to 90% of what they learned in simulating the real experience, even after 2 weeks (Figure 2). However, there is no guidance in the current literature on how to select simulation and virtualisation tools for teaching computer networking online. It is known, however, that it is important to select the right simulation or virtualisation tool for the right job, to model only the required aspects, features and behaviours of the real network infrastructure in an online learning environment driven by software (Gegenfurtner et al., 2014; Kabir et al., 2014).

Figure 2.

There are two main approaches to teaching using network simulation. The first aims at simulating behaviours of network devices and protocols from the network configuration perspective. For example, several vendor-specific network simulators, such as Cisco Packet Tracer (Packet-Tracer, 2020), Juniper Network Simulator (Juniper, 2020) and Huawei eNSP (eNSP, 2020), are used intensively in teaching vendor certification courses such as Cisco, Juniper and Huawei Certified Network Associates (CCNA, JCNA and HCNA). The main objective of these simulations is to provide students with an environment to learn and practice configuring vendor-specific network devices and protocols. Students can experiment with different network configurations and ask ‘what if?’ questions.

The second approach aims at simulating behaviours of the network from the network traffic engineering perspective. This approach often involves the simulation of the full network protocol stack, that is, a simulation of how data are passed across networks (following the OSI model). The main objective of this approach is to study performance and behaviour of a network under different traffic conditions. For example, students can ask questions like ‘What is the variation in delay of IP packets crossing the network under a certain traffic conditions?’ or ‘What is the variation of the Window size of a TCP session if the network background traffic is double of the current level?’ Network simulators following this approach include NS3 (Nsnam, 2020), OPNET (Opnet, 2020) and NetSim (Tetcos, 2020). As network simulators do not attempt to simulate every aspect, feature and behaviour of the network, they cannot completely replace real network devices in teaching more advanced networking topics such as network traffic engineering, network management and network security.

Virtualisation is seen as an alternative solution to simulation. The availability of high-performance computing devices and cloud computing allows abstraction (virtualisation) of any computing hardware within software. As a result, it is possible to use software to create and simulate a fully functional network with virtual routers, switches, terminals and their connectivity, with little to no difference in terms of functionality between the virtual network and its physical counterpart. This allows a virtualised network to seamlessly connect to the external physical network, making the virtualised network accessible from anywhere in the world. The instructional design features that are made available from virtualisation software offer the potential for valuable teaching approaches in computer networking, particularly related to understanding and application of complex concepts. Currently, there are several network simulation tools, also referred to as network emulators, that follow virtualisation approaches, for example, GNS3 (GNS3, 2020), VIRL (Virl, 2020), EVE-NG (Eveng, 2020) and Huawei eNSP (eNSP, 2020).

The selection of a network simulation or visualisation tool for teaching computer networks in online learning environments should enhance the following aspects of the instructional design.

Learning outcomes: The simulation or visualisation tool selected for the subject or course should enable delivering the learning content required for achieving the specified learning outcomes.

The features and suitability of the most commonly used network simulation or visualisation tools for teaching online computer networking courses are summarised in Table 1. Apart from network virtualisation, another form of virtualisation, virtual reality (VR, sometimes referred to as augmented, mixed or extended reality, see e.g., Brown et al., 2020), is seen as a powerful technology for conducting activities that can serve as a replacement for some sets of practical experience in online settings (Radianti et al., 2020). For example, VR technologies have been successfully used for equipment, operational task and safety training, due to their ability to create powerful realistic and immersive visualisation (Wang et al., 2018). VR technologies can be used to teach several practical skills in computer networks such as familiarisation with network physical facilities and hardware device operation, network physical security training, and prototyping physical network design. The availability of prebuilt 3D models of network devices and free 3D design software such as Blender (Blender, 2019) and Unity3D (Unity, 2019) makes VR technology readily accessible for teaching computer network courses.

This section presents the case context for this study, aligned with the theoretical background presented earlier, including an outline of the change in instructional design and teaching approaches used when the computer networks subject was delivered in the online learning environment.

At the study university, computer networks is an advanced-level subject taught in the first year of a master of information technology degree. The subject is designed to give students an in-depth understanding and practical experience in planning, configuring and managing computer networks that would be typically used in small-to-medium enterprises (SMEs). This in-depth understanding and practical experience includes, therefore, the knowledge, skills and experiences that will enable students to conduct different network management tasks including monitoring network performance, security and configuration, troubleshooting network issues, and providing advice and guidance to management on planning and upgrading networks.

The subject was chosen to illustrate the transition from a largely F2F (or blended) design to the one that was fully online, largely because the blended design followed a conventional bottom–up approach of the OSI model, with and extended lecture plus lab component typical of an advanced-level computer networking subject. The decision to change the instructional design was made in part to accommodate life on ‘planet COVID’, but this change also presented as an opportunity to alleviate a high student attrition and failure rate through allowing a teaching approach that provided for improved motivation and understanding in the subject.

New teaching approaches were based on an instructional design that included modified learning activities and assessment protocols built around problem-based scenarios to be delivered online. These scenarios are designed to engage students through activity in the situated investigative approaches that are drawn from the subject syllabus and which consider the relative nature of context in learning via solving real-world or open-ended problems (Barab & Plucker, 2002; Chris et al., 2018; Savery, 2015; Woolcott et al., 2017a, 2017b). These approaches arise therefore not from traditional didactic pedagogies but from theoretical approaches to student-centred learning that developed from research on problem-based learning and the related anchored learning and inquiry learning (Barab & Plucker, 2002; Barab & Squire, 2004; Savin-Baden & Howell-Major, 2004; Willis, 2006).

While these scenarios use a learner-centred approach to teaching computer networking (see e.g., Linge & Parsons, 2006; Savery, 2006), the scenarios are designed as experiential under guidance and may best be described as scaffolded in the sense that instruction is not minimal (see discussion in Efendi & Yulastri, 2019; Hmelo-Silver et al., 2007). As Savery (2006, p. 12) points out, ‘Critical to the success of the approach is the selection of ill-structured problems (often interdisciplinary) and a tutor who guides the learning process and conducts a thorough debriefing at the conclusion of the learning experience.’

The design used in the current study was centred specifically on an overall network management scenario with five network management problems following the OSI fault, configuration, accounting, performance and security (FCAPS) network management model (Clemm, 2006). The problems were presented to students in a specific order to allow scaffolding—students begin with the configuration management problem followed by performance management, accounting management, security and fault management problems. This order was dictated by a need for students to gain expertise in configuring the network before they could monitor its performance and perform network accounting tasks. Once expertise in these three areas was obtained, students could tackle the more complex security and fault management problems, building on what they learnt in configuration management, performance and accounting management. The overall design of learning flow presented in Figure 3 illustrates the order or the learning content and the relationship between the theoretical and the practical parts of the subject.

Figure 3.

The theoretical part, delivered as online lectures recorded for repeat viewing, began with revision of the IP, followed by IP address subnetting, IP address planning and IP routing. Since computer networking is an advanced-level subject, this fundamental knowledge was assumed, but the revision included acknowledged that time may have elapsed since students completed the pre-requisite, lower-level subjects. In addition, as mentioned earlier, it was considered beneficial to start the computer network course from the network layer with the view that this design would allow an easier establishment of a contextual link for the subsequent layers. In the practical part, which was designed to support the theory presented in the initial two theoretical parts (see Figure 3), students were presented with the problem of setting up a network simulation environment at home and configuring a networking scenario representing a typical SME network. This practical task allowed students to revise their knowledge on IP addressing, IP routing and some network configuration skills.

The flow design then led to coverage of the most important protocol in network management—the Simple Network Management Protocol (SNMP) and architectures for SNMP deployment. The associated practical parts required students to solve the problem of selecting a suitable network management architecture for the given network scenario and perform all necessary configuration tasks to successfully deploy SNMP for network management. The second half of the subject covers each network management function starting from the configuration management, followed by performance, accounting, security and fault management. At this stage, students received a personalised network management problem, which required them to perform actual network management tasks on their simulation networks on a daily or weekly basis. This strategy gave an authentic problem to each student while helping to minimise the chance for academic misconduct, such as contract cheating.

To support students, an online discussion board was set up to answer questions from students within 24 hr. In addition, two (paired) 1-hr online sessions were provided every week via Zoom or Blackboard Collaborate (an online learning system used for subject delivery). It is worth noting that two 1-hr sessions were used in place of the one 2-hr session used in the blended delivery to improve interactivity and responsiveness in teaching this subject online.

Since conducting integrated practical lab activities is seen as one of the major challenges in teaching computer network courses in online environments, design elements were introduced to enable network virtualisation within VR learning environments (VRLEs). Practical activities were designed based on criteria outlined earlier and using the following open-source (free) software: Huawei eNSP, Wireshark, Oracle VirtualBox and Zabbix. Huawei eNSP is a network simulator that uses the Oracle VirtualBox virtualisation technology as the core. Since eNSP makes use of virtualisation, networks simulated in eNSP can be connected to virtual and physical networks outside the simulation environment, allowing the management of the simulated networks with a real network management solution such as Zabbix. This capability was considered critical to achieving subject learning outcomes.

It is worth noting that while several other network simulators (e.g., GNS3 and NetSim) offer a similar capability, Cisco Packet Tracer does not. Zabbix is a network management software platform, which has been widely used in the industry because it is light weight and enterprise class (Zabbix, 2019). Zabbix can easily run on a Linux virtual machine, which makes it highly accessible to students. Wireshark is a network and protocol analyser, which has been integrated as part of eNSP. Another advantage of this software combination is that each of them can run on a single Windows PC with an Intel i5 processor and 8GB RAM, a common configuration of a home PC, which makes them suitable for students.

The final instructional design for online delivery had three main practical activities, aligned with subject assessments to determine students’ success. Each practical activity was scaffolded via a series of activities linked to learning resources and quizzes to guide students to an activity goal. The detailed step-by-step, instructions as seen in similar activities (e.g., Cisco training materials (Deal, 2008)), were not immediately provided to students. Instead, a series of structured questions was used to guide students in finding solutions. The following list provides the goal and timeline of each of the three activities.

Activity 1: The goal of this activity is to help students in setting up a simulation environment (completed by students in week 3).

To help students complete each activity, a short (5–10 min) video providing the solution to the problem was made available to students in the last week of each activity.

In addition to the aforementioned activities, students were also provided a virtual tour of a server room using the VR technology. The VRLE was constructed using Unity (2019) and exported as a 360-video experience accessible via YouTube. Students were able to watch the video with or without a VR headset, although such headsets would give students an immersive experience not available otherwise. Figure 4 shows a screen shot from the tour.

Figure 4.

The instructional design included a sequence of four assessments with personalised and unique tasks, following the proposed assessment design for this online teaching approach outlined before (Bloxham & Boyd, 2007; Bretag et al., 2019). Each assignment was weighted equally (25% each). There was no examination in the online subject offering.

Assessment One comprised two parts, one theoretical and one practical. In the theoretical part, students were asked to document the physical and logical design of a given network. Each student was assigned with a unique IP address space. In the practical part, each student was required to set up a network simulation environment on their personal device and configure the given network in a simulation environment that was set up for them. Each student was then asked to create a 5-min video to demonstrate the working network in the simulation environment. This assessment was due at the end of week 4 (of the 12-week subject).

Assessment Two also comprised two parts, theory and practice. First, each student was asked to present an argument for their selection of a network management architecture to manage the network given to them in the first assignment. Next, each student was required to configure the network for SNMP deployment. Zabbix was used in the role of network management systems. Finally, as in Assessment One, each student had to create a short video to demonstrate the working network management architecture in the simulation environment. This assessment was due at the end of week 7.

Assessment Three was a practical task that did not contain a theoretical part. Students were required to perform a set of network management tasks on their simulation networks and document the process. The tasks include performance baselining, configuration backup and restoration, collecting and accounting information for a network interface, performing security checks and developing a troubleshooting plan for a network issue. Students were considered successful in Assessment Three if their process documentation satisfied given criteria. This assignment was at the end of week 11.

In Assessment Four, the final assessment, each student was asked to conduct research on recent network technologies and present a case to managers to upgrade the existing network. This assignment was at the end of week 12.

In the 2019 blended offering, students were required to complete two assignments and an online invigilated exam. Assignment One was the same in both offerings, and Assignment Two in the blended offering was the same as Assignment Four in the online offering. The 2019 examination was a 3-hr paper comprising mainly questions on theory, with no real testing that required understanding of problem-solving scenarios. In contrast, Assignments Two and Three in the online offering actively tested students’ problem-solving skills.