Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Advertisement
Scientific Reports volume 15, Article number: 7619 (2025)
1499
Metrics details
The educational landscape is witnessing a transformation with the integration of Educational Technology (Edutech). As educational institutions adopt digital platforms and tools, the generation of Educational Big Data (EBD) has significantly increased. Research indicates that educational institutions produce massive data, including student enrollment records, academic performance metrics, attendance records, learning activities, and interactions within digital learning environments. This influx of data needs efficient processing to derive actionable insights and enhance the learning experience. Real-time data processing has a critical part in educational environments to support various functions such as personalized learning, adaptive assessment, and administrative decision-making. However, there may be challenges in sending large amounts of educational data to cloud servers, i.e., latency, cost and network congestion. These challenges make it more difficult to provide educators and students with timely insights and services, which reduces the efficiency of educational activities. This paper proposes a Regional Computing (RC) paradigm designed specifically for big data management in education to address these issues. In this case, RC is established within educational regions and intended to decentralize data processing. To reduce dependency on cloud infrastructure, these regional servers are strategically located to collect, process, and store big data related to education regionally. Our investigation results show that RC significantly reduces latency to 203.11 ms for 2,000 devices, compared to 707.1 ms in Cloud Computing (CC). It is also more cost-efficient, with a total cost of just 1.14 USD versus 5.36 USD in the cloud. Furthermore, it avoids the 600% congestion surges seen in cloud setups and maintains consistent throughput under high workloads, establishing RC as the optimal solution for managing EBD.
Education 4.0, often called smart education, represents a transformative approach to teaching, learning, and administration. By utilizing advanced technologies (e.g., 5G, Internet of Things (IoT), and Artificial Intelligence (AI)), it aims to enhance the effectiveness and appeal of educational practices1. Terms like “smart learning,” “smart classroom,” “smart learning environment,” and “smart university” are often used interchangeably to describe this concept2. In this context, “smart” refers to making intelligent decisions, providing personalized learning experiences, maintaining transparency, and adapting to individual needs3. With the continuous evolution of smart technologies, educational institutions are encouraged to embrace these advancements and transition from traditional methods to smart systems. This change is essential for maintaining relevance and effectiveness in the fast-evolving educational landscape4,5.
Big Data in Education.
Every day, organizations are overburdened by the vast volumes of data, commonly referred as big data6. It includes datasets that are so large and complex that traditional data processing applications are insufficient for handling them effectively. In the context of education, EBD have the vast amount of data produced by the educational environments, including student enrollment records, academic performance metrics, attendance records, learning activities, and interactions within digital learning environments7,8, as shown in Fig. 1. This information provides valuable insights into student behaviours, learning trends, and academic achievements, empowering educators and administrators to utilize data-driven approaches for enhancing teaching methods, customizing learning experiences, and ultimately improving educational results9.
Gross Enrollment Ration in Primary Education in 2022.
Currently, millions of students worldwide are enrolled in educational systems, accessing learning materials, and engaging with teachers and peers through digital platforms. Figure 2 illustrates the percentage of the population enrolled in school, indicating a high rate of school admission ranging from 90% to 100%. However, Fig. 3 reveals a significant disparity in the completion of secondary education among those initially enrolled, indicating a need for improvement in educational outcomes10. Consequently, there is a growing imperative to transition our education system to smart education, leveraging emerging technologies to effectively engage a larger segment of the population, as highlighted in recent research11.
Completion Rate of Secondary Education in 2021.
The extent to which educational institutions have adopted smart technologies varies widely. While some organizations have fully accepted the concept of smart education, many others are still planning12. It is projected that by 2030, a significant proportion of educational institutions will have transitioned to smart institutions, leveraging advanced technologies to enhance teaching and learning experiences13. This transition will greatly increase smart devices connected to the internet, allowing seamless communication and data exchange in education. As a result, the volume of data produced by these interconnected devices is expected to grow dramatically, giving educators and administrators valuable insights into students’ performance. This surge in data will drive further advancements in smart education and create opportunities for more personalized and effective learning experiences worldwide13.
The global big data market is expected to grow significantly, with revenues projected to reach 103 billion (USD) by 2027. This represents a compound annual growth rate of 12.7% from 2022 to 202714. In addition, data generation is forecasted to increase dramatically, with global output anticipated to reach 291 zettabytes by 202715, as illustrated in Fig. 4. In the education sector, the big data market was valued at 13.58 billion (USD) in 2020 and is forecasted to grow significantly, reaching 57.14 billion (USD) by 203016.
The growth in EBD has various challenges, including delays, network congestion, higher cost and performance issues17. These hurdles can resist timely access to critical information, disrupt online learning activities, and overburden the mainstream network18. Delays in accessing learning materials or receiving feedback may restrict the learning process and decrease student engagement, while network congestion and performance issues can disrupt lectures, discussions, and collaborative activities, impacting effective communication and interaction19. Prolonged disruptions could lead to decreased productivity and motivation among students, ultimately affecting their academic performance and overall learning outcomes. Addressing these data-related challenges is necessary for ensuring the smooth and effective delivery of education in digital learning environments20.
As educational technologies continue to evolve, the utilization of data generated within educational systems becomes increasingly crucial for shaping effective learning environments. Despite advancements, further exploration is essential, particularly in managing EBD. Therefore, the primary objectives of this initiative include:
Statistics of device connectivity, size of big data and total investment in big data.
Efficient management of EBD at regional servers instead of transferring it to the cloud servers in real-time.
Educational institutions initially store their data on regional servers, which is later migrated to the CC during off-peak time for subsequent utilization.
Ensuring minimal response times for educational resources and related information, despite the large amount of data generated by devices worldwide.
Edge and Fog Computing (FC) paradigms operate near users, processing and storing data at the edge instead of transmitting it to the cloud21. This approach significantly reduces processing time, transfer duration, and costs22,23. However, these paradigms may face challenges in handling big data, particularly data from social media, EBD, healthcare big data etc. The long-distance location of cloud servers can lead to delays and high workloads, resulting in decreased performance and increased costs. To address this issue, RC is emerging as a viable solution24.
This concept can be illustrated more clearly with a real-life example:
In EBD scenarios, when students request learning materials, the RC system checks if the content is available locally. If found, it’s streamed directly to their devices, reducing strain on the public network and cloud servers. Content filtering is done locally to manage peak-hour workload, ensuring smoother access. During off-peak hours, data is optimized for transfer to cloud servers for storage or processing, maintaining efficient data flow.
About the above problems and the objectives, below are the key contributions of this study;
Proposed RC infrastructure to address key challenges such as delay, cost, and network congestion by enabling localized data processing, reducing reliance on cloud-based systems, and ensuring efficient management of EBD.
Educational data from diverse sources is stored on regional servers and migrated to cloud platforms during off-peak hours for analysis and curriculum improvement.
Regulations and guidelines within RC centralize educational data and analyses, ensuring compliance with norms and streamlining institutional operations.
The structure of the paper is outlined as follows: Section 2 explores existing literature related to data offloading. Section 3 elaborates on the proposed methodology for the operation of the RC layer in managing EBD. Section 4 provides an evaluation of the framework through experimental results. Section 5 offers an in-depth discussion on edge, regional, and CC within the context of education. Finally, Sect. 6 concludes the study along with the future directions.
Previous research has largely concentrated on big data utilization for analytics, with less emphasis on big data transportation. Similarly, limited attention has been given to the processing and storage of EBD. Existing literature indicates that educational institutions, particularly Massive open online course (MOOC)s, generate a substantial volume of data daily. This data is processed in real time by students globally. This section covers the background in two subsection i) EBD and ii) Offloading Techniques.
Big data has emerged as a transformative force in education, providing opportunities to enhance teaching methods, advance educational research, and improve administrative processes25. Recent technological advancements have spurred increased interest and research in the application of big data in education26,27. Educational institutions are now exploring innovative ways to collect, store, and analyze vast amounts of information to better understand student behaviour, improve system efficiency, and integrate big data concepts into their curricula28. Techniques such as learning analytics and educational data mining have become essential tools for effectively utilizing this data. The categorization of big data in education, illustrated in Fig. 1, provides a structured framework for understanding its various applications29. The growing reliance on data generated by online educational platforms creates additional opportunities for insights and innovation, offering significant advantages to educators, students, and researchers alike.
Personalized learning has become a key focus in educational technology due to its ability to tailor content and subjects to meet the unique needs of individual learners. This approach allows students to engage with educational materials at a pace and depth that suits their specific requirements30. A study referenced in31 examined course recommendation systems utilizing big data analytics. The findings indicated that the proposed method surpassed existing algorithms in providing more accurate course recommendations. The authors in32 highlighted several key challenges in education, including the need to identify student misconceptions, predict dropout rates, and improve overall educational quality. Their study demonstrates the potential of utilizing data and advanced technologies to tackle these issues. They suggest various supervised learning methods as effective solutions to enhance personalized learning and improve educational outcomes.
Managing students and maintaining discipline are significant challenges for educational institutions. A study referenced in33 addressed this issue by utilizing big data to analyze students’ daily routines, learning preferences, and behaviours. This approach provided valuable insights that helped streamline student management. The authors in34 proposed a big data-based educational management model that improved information accessibility and expanded big data applications in administration. Similarly, authors in1 used similar strategies to enhance student management and engagement.
Big data is transforming teaching practices, with flipped classrooms35 and homeschooling36 standing out as prominent examples. In line with these innovations, the authors in37 proposed a hybrid teaching model. Their study demonstrated that this approach fosters greater student engagement compared to conventional classroom methods.
Education is completely connected to big data, serving as both a substantial producer and consumer of it. Millions of students, educators, and administrators actively contribute to and benefit from this evolving ecosystem36. The COVID-19 pandemic significantly accelerated the demand for virtual classrooms, underscoring big data’s critical role in adapting to these shifting educational paradigms38. Concepts such as personalized learning and home-based schooling are becoming increasingly prevalent, utilizing the analytical power and capabilities of big data39. This interconnected relationship continues to drive innovation and redefine the future of education.
The increasing volume of EBD, driven by IoT, AI, and mobile technologies, presents significant challenges in data processing and storage. CC has been widely adopted for scalable data management, but issues like latency and network congestion resist its effectiveness for real-time applications. To address these issues, offloading techniques like edge and FC provide localized processing near data sources, reducing latency and improving responsiveness in educational settings.
CC has become a valuable technology for addressing challenges in EBD, providing scalable and efficient methods for managing educational data. Many institutions have implemented cloud-based platforms to improve data storage and processing, which enable features for example student monitoring and adaptive learning systems40. However, despite its advantages, cloud computing also faces significant challenges, including latency, network congestion, and delayed responses, which are particularly critical in time-sensitive educational applications. Additionally, issues related to system integration and organizational management have been identified as major barriers to the widespread adoption of cloud-based solutions in educational environments41,42,43.
To mitigate these limitations, Multi-access Edge Computing (MAEC) has been proposed as a solution that processes data closer to its source. By minimizing latency and bandwidth usage, Mobile Edge Computing (MEC) improves the responsiveness of data-intensive educational applications. Studies have demonstrated the effectiveness of MEC in scenarios requiring real-time analytics, such as adaptive assessments and immediate feedback systems44. Additionally, frameworks designed for cooperative edge networks have shown improvements in task efficiency and performance in MEC environments45. However, scalability and resource allocation remain significant challenges for MEC, particularly when managing the simultaneous processing of large datasets46,47,48.
Edge and FC have also gained traction as viable alternatives for handling real-time educational applications. Edge computing offers localized data processing at or near the source, making it ideal for time-sensitive tasks such as classroom performance analytics49. For example, educational systems leveraging edge computing have optimized resource allocation, reduced operational costs, and improved overall system performance through advanced models like game theory-based approaches50. Similarly, the adoption of edge computing in smart classroom environments has enhanced bandwidth usage and response times51. Nonetheless, edge computing faces limitations in scaling to meet the demands of large educational institutions.
FC, particularly for applications involving wearable devices and remote learning platforms, provides a middle layer for processing and storing educational data. By reducing latency and alleviating network bandwidth constraints, FC enables real-time analytics for personalized learning interventions and predictive assessments52. Innovations such as the Smart Learning Gateway demonstrate the potential of FC to integrate local data storage, real-time analytics, and embedded intelligence within IoT-based educational systems53. While addressing mobility and scalability issues, FC still struggles with challenges related to data synchronization across multiple institutions.
Additionally, the concept of cloudlets, small-scale data centers positioned near data sources, has been explored to improve data management. Cloudlets show promise for supporting localized EBD but encounter scalability and resource limitations, particularly during peak usage54. Hybrid models combining fog and CC have been proposed to address these challenges, utilizing the strengths of both approaches to ensure efficient data processing, classification, and storage in dynamic educational environments55.
While big data offers number of benefits for education, it also presents several concerns that must be addressed. One primary concern is the risk of data misuse, especially given the vast amount of learner data, including institutional and individual geographical details, which could be mishandled56. Moreover, there are worries regarding data bias and algorithmic bias, which could lead to unfair outcomes and discrimination57. Additionally, with the increasing connectivity of devices worldwide, there is a corresponding rise in data generation, leading to network congestion and delay issues21. To avoid needless data transmission to the cloud, it is crucial that all industries, especially those producing large amounts of data, thoroughly review their data and have filtering processes in place at the organizational level. Similarly, educational institutions should filter their data at the edge or regional level before transferring it to the cloud to mitigate these concerns effectively.
The proposed methodology aims to adapt RC for EBD, enhancing the efficiency of educational data processing and storage while minimizing delays and network congestion. With the existing system struggling to handle the massive influx of EBD, innovative solutions are imperative. The proposed system for EBD involves three layers: T he Internet of Educational Things (IOET) Layer, the RC Layer, and the CC Layer. The proposed layout of RC for EBD is illustrated in Fig. 5. Table 1 shows the terminologies used in the study.
Structure of Regional Computing for Educational Big Data.
The IOET Layer operates at the educational institutions, where smart devices of students and teachers generate data. This layer produces data which is collectively handle by the educational edge and then sent towards the regional servers for processing and storage. The RC Layer serves as a buffer between the education layer and the cloud, processing and storing EBD before transferring it to the cloud21, as shown in Fig. 6. Placing computing infrastructure closer to educational institutions minimizes delays and costs associated with data transmission. Educational data is processed and stored regionally, reducing the burden on the network and optimizing data transfer during off-peak hours. The CC Layer functions as the central repository for EBD, facilitating coordination and analysis across educational institutions. This layer stores and manages the vast amounts of educational data generated by various sources, enabling collaborative efforts and data-driven decision-making.
The Educational Layer encompasses the smart devices used within educational institutions, including students and teachers’ devices, along with the local servers (edge computing) situated within the educational environment. This layer facilitates data generation, processing, storage, and communication among students, teachers, and administrators.
During educational activities such as online lectures, collaborative projects, and assessments, this internal system is actively utilized. Students and teachers interact with digital learning platforms, access educational resources, and participate in virtual classrooms using their smart devices. The servers within educational institutions handle the processing of educational data, including student assignments, assessment results, and learning analytics.
IOET equipped with sensors generates massive data in real-time. Sensors such as cameras, microphones, and accelerometers capture student interactions and activities, while communication features facilitate collaboration and communication among users. The servers process this data using sophisticated algorithms and AI to provide personalized learning experiences and academic support.
Educational institutions need high-speed internet to communicate EBD to the regional or cloud servers. So the total delay faced by the end devices is as;
Where (Del_{tran}) represents transmission delay, (Del_{prop}) stands for propagation delay, (Del_{proc}) indicates data processing delay, and (Del_{que}) denotes queuing delay.
The transmission delay ((Del_{text {tran}})) is calculated as;
Here, W represents the workload (data), B denotes the channel bandwidth capacity, SNR stands for the signal-to-noise ratio, ME (Modulation Efficiency) indicates the efficiency of the modulation scheme, and ER represents the rate of transmission errors.
The propagation delay ((Del_{text {prop}})) shows the time taken to traverse the medium from its origin to destination. This delay relies on the distance ((Dis)) between the device and the server and the transmission speed ((tr_{s})). (Del_{text {prop}}) is calculated as:
The data processing delay ((Del_{text {proc}})) refers to the time needed for the system to process the data. This delay is directly proportional to the volume of the data ((W)) and the processing speed of the system ((P_{r})). The processing delay, denoted as (Del_{text {prop}}), is calculated as:
The queuing delay ((Del_{text {que}})) shows the time the task spends in a queue awaiting processing. This delay is directly proportional length of the packet ((L)), the arrival rate of packets ((a)), and inversely proportional to the packet processing rate ((R)). Queuing delay (Del_{text {que}}) is calculated as:
Similarly, the cost of data offloading rises with both the distance travelled and the size of the data. This relationship is represented by the following equation:
Here, (C_{t}) denotes the total cost incurred by data transmission to the regional server, which comprises both the transmission cost ((C_{text {tran}})) and the propagation cost ((C_{text {prop}})).
Data flow from Educational Institutions to Regional Computing and Cloud Computing.
RC situated within a designated educational region serve as centralized hubs for processing and storing EBD. These servers collect data from various sources within their region, including smart devices used by students, teachers, administrators and institutions. The data transmitted to regional servers encompasses diverse data such as student enrollment records, academic performance metrics, attendance records, learning activities, and interactions within digital learning environments
Regional servers process EBD to derive meaningful insights and information. This involves analyzing learning patterns, identifying classroom challenges, and highlighting areas where students excel. These servers play a crucial role within educational institutions by transforming aggregated educational data into actionable knowledge, thereby supporting strategic planning and data-driven decision-making.
During peak hours, EBD is temporarily stored on regional servers to minimize network congestion and response time. During off-peak time this data is sent to the CC for long-term storage and additional analysis. By using this strategy, educational institutions may guarantee real-time responsiveness to changing student needs and reduce data transmission delays.
Moreover, the delay ((Del_t)) experienced during data transmission, as depicted in Equation 1, is influenced by factors such as distance (Del) and workload (L). Processing and storing EBD locally during peak hours mitigate the total delay, enabling timely responses to educational queries and minimizing network strain. This localized approach to data management optimizes the educational environment’s efficiency and enhances the overall educational experience for students and educators alike.
Educational Regional Computing
The EBD management at regional level is shown in Algorithm 1. This is designed for handling EBD within the educational region. It receives data from sources such as learning management systems, student information systems, digital learning tools, and administrative systems. This algorithm decide whether to migrate the data to the RC or to the CC.
During peak-hours, the algorithm offload to the regional servers to ensure real-time responsiveness. Conversely, during off-peak hours, data is migrated to cloud servers for long-term storage and in-depth analysis. If regional resources are insufficient, the algorithm prioritizes transferring critical data to the cloud, maintaining efficiency within the educational environment. The algorithm facilitates data transfer to the cloud for training across different educational regions, ensuring smooth operations in various geographical locations.
The educational cloud layer operates as a vital component in the ecosystem of EBD management, complementing the functionalities of the regional layer. While the regional layer actively engages in real-time educational operations and local processing, the cloud layer serves as a passive yet powerful resource for advanced computing capabilities and extensive data analysis.
The regional layer utilizes the cloud layer for scalable resources and offloads the EBD during off-peak time to enhance processing. The cloud layer serves as a central hub, storing EBD from global educational institutions for processing and collaboration. The processing time ((T_{text {cloud}})) for managing all educational workload at cloud servers is given by Equation 7, where (W) represents the total educational workload and (R) represents the computing rate of the CC.
The total educational workload ((W)) is calculated by summing the data produced by each educational source ((W_i)) from 1 to (n), as shown in Equation 8:
Upon examining Equation 1, it becomes clear that the round-trip to the cloud can lead to increased delays, primarily due to the large volume of EBD ((W)) being processed. Furthermore, factors such as propagation delay and queuing delay are likely to increase in relation to the growing workload and the distance between the educational sources and the cloud servers.
Educational Cloud Computing
To manage EBD effectively, the Algorithm 2 was developed. It performs analyses including anomaly identification and student performance analysis, preprocesses data, and uses analytics for insights. It works with local servers and edge computing, optimizes data storage and retrieval, and feeds administrators’ and educators’ systems with real-time changes. It also establishes procedures for data retention and compliance, tracks system performance, and makes data sharing easier for more extensive educational applications.
Energy consumption increases with distance, affecting operational costs for data transfer in education. The energy consumption calculation is as follows:
Energy utilization between consecutive stages, denoted as (E_{text {tran}}(i, i+1)), increases as the stages progress.
Where
In this context, (D_{i, i+1}) represents the distance between consecutive devices, (P_{i}) denotes the power, and (T_{i, i+1}) indicates the time taken by these consecutive devices. The equation shows that power consumption increases with distance.
Similarly, (E_{text {other}}) refers to the energy consumption associated with other activities, which include processing ((E_{text {pro}})), storage ((E_{text {stor}})), and cooling of the data centers ((E_{text {col}})).
As per the above, the total energy utilization is calculated as:
In this context, (E_{text {total}}) represents the total energy usage, while (E_{text {tran}}) denotes the energy utilized for data transfer, which includes networking devices (e.g., medium, switches, routers etc).
It is also known that,
The energy consumption ((E)) is proportional to the operational cost ((Cost_{text {oper}})), meaning that as energy consumption increases, operational costs also rise.
To evaluate the effectiveness of the RC framework in managing EBD, we utilized EdgeCloudSim, a widely recognized simulation tool that allows for the simulation of cloud networks and assists in assessing the performance of networks and data centers. Our preliminary research focused on using this to analyze the effects of EBD on network behaviour, taking into account both RC and CC paradigms.
The experimentation setup of this study is categorized into two different scenarios;
In the first case, we evaluated the delay, cost, congestion and throughput parameters associated with offloading, processing, and storing EBD on clouds.
In the second, we investigated the delay, cost, congestion and throughput of offloading, processing, and storing EBD on RC.
In the first phase of the experimentation, a Data Center (DC) was set up in North America (Region 1), while educational institutions were represented in various regions around the world. To simulate the scenario, a workload of 10 KB per device was generated, initiating the offloading from the educational institutions to the cloud server. In the second phase of the experimentation, regional servers were deployed in their respective regions to process the workload specific to each area.
Each DC used five virtual machines (VMs) with 200 GB of RAM, 100,000 GB of storage, 6 GB of bandwidth, 4 CPUs, and a processing speed of 10,000 MIPS. Peak hours were from 01:00 AM to 09:00 PM, with 1,000 educational institutions submitting requests simultaneously. Off-peak hours lasted from 01:00 AM to 09:00 AM, with 100 institutions submitting requests.
The delay (in milliseconds) comparison of cloud and regional computing.
The comparative analysis of total delay between regional and CC , as shown in Fig. 7, demonstrates the superior performance of RC for EBD. Simulations were conducted for 100 to 2,000 devices, each transmitting 10 KB of data each. The results show that the total delay for the RC is consistently lower compared to CC. For instance, at maximum 2000 devices, the regional delay is 203.11 ms, while the cloud delay is significantly higher at 707.1 ms. By significantly reducing delays, RC ensures smoother operations and faster responses for EBD.
The processing cost comparison of cloud and regional computing.
The comparison of processing costs between regional and CC underscores the cost-effectiveness of regional systems in managing educational devices, as depicted in Fig. 8. Across simulations involving 100 to 2,000 devices, the processing cost in RC remains consistently lower than in CC. At the upper limit of 2,000 devices, the regional processing cost is only 0.57 USD, whereas the cloud processing cost escalates to 5.07 USD, nearly nine times higher. This significant difference underscores the economic advantage of RC, making it a more viable solution for processing tasks in educational environments.
Communication cost comparison of cloud and regional computing.
The communication cost comparison between regional and CC demonstrates the efficiency of regional for handling EBD, as shown in Fig. 9. Across simulations ranging from 100 to 2,000 devices, RC consistently incurs significantly lower communication costs. At the maximum of 2,000 devices, the communication cost in RC is only 0.29 USD, whereas the cloud communication cost reaches 0.57 USD, almost double. This considerable cost difference highlights the advantage of RC, particularly in scenarios requiring frequent communication between devices, making it a more economical option for educational systems.
Total cost comparison of cloud and regional computing.
The total cost comparison between regional and CC reveals a substantial cost advantage for RC in managing educational devices, as shown in Fig. 10. Across all simulations involving 100 to 2,000 devices, the total cost in RC is consistently lower than in CC. At the maximum of 2,000 devices, the total cost in the regional setup is just 1.14 USD, while the cloud total cost increases significantly to 5.36 USD, more than four times higher. This significant difference in total costs underscores the overall economic benefits of RC, making it a more cost-effective solution for educational institutions managing large-scale networks of devices.
The congestion comparison of cloud and edge computing.
As shown in Fig. 11, congestion in CC increases dramatically as the number of devices increases, reaching 75% at 800 devices and surging to nearly 600% at 2,000 devices. This steep rise in congestion leads to severe network bottlenecks and diminished performance due to the heavy data transmission load, which overwhelms the cloud infrastructure. In contrast, RC exhibits consistent stability, effectively supporting up to 1,000 devices without significant congestion. This reliability is attributed to its localized data handling and limited device scope, enabling regional systems to efficiently manage workloads and avoid the severe congestion challenges observed in CC.
The throughput comparison of edge and cloud computing.
Figure 12 highlights the differences in throughput between CC and RC as the number of devices increases. In CC, throughput initially rises sharply, reaching a peak of 5000 MB per second when managing approximately 1000 devices. However, the increasing number of devices overutilized the system, leading to a significant drop in throughput to 3333 MB per second at 2000 devices. Conversely, RC demonstrates stable throughput regardless of the number of devices. This stability is attributed to its ability to effectively distribute the workload, ensuring consistent performance and avoiding the degradation observed in CC.
The experimental findings offer significant evidence of the advantages of RC compared to CC for handling EBD. In terms of crucial performance indicators like delay, cost, congestion, and throughput, RC has shown consistent superiority over CC.
Delay analysis reveals that RC significantly enhances response times, achieving a delay of 203.11 ms for 2,000 devices compared to the cloud’s 707.1 ms. This reduction ensures smoother operations and faster responses for educational applications. Cost analysis shows the economic advantage of RC, with processing costs as low as 0.57 USD and communication costs at 0.29 USD for 2,000 devices, compared to cloud costs of 5.07 USD and 0.57 USD, respectively. The total cost difference is particularly striking, with RC incurring only 1.14 USD against the cloud’s 5.36 USD, making it a more sustainable and cost-effective option. Congestion management further highlights the robustness of RC. Unlike CC, which experiences congestion surges up to 600% at higher device counts, RC remains stable and manages up to 1000 devices efficiently. Similarly, throughput analysis demonstrates RC’s reliability, maintaining consistent performance across increasing workloads, while CC suffers from throughput degradation due to network overload.
Based on our experiments and the existing literature, we conducted a comprehensive comparison of RC and CC for EBD across various parameters, including delay, server mobility, area coverage, computational power, data sources, and capacity. The detailed comparison can be found in Table 2. This analysis demonstrates the potential of RC as a scalable and efficient solution for educational institutions. It addresses critical performance and cost challenges while highlighting areas for further optimization.
After evaluating the proposed methodology, it is evident that the RC paradigm effectively addresses the identified challenges. Nevertheless, several specific issues need to be addressed to positively impact both delay and cost.
The main challenges include the ownership cost, which creates a barrier since educational institutions need to invest massively in deploying and operating RC13.
Furthermore, processing data at the terminal level of regional servers raises significant privacy and security issues that need to be properly managed58.
A significant challenge arises from the differing cyber regulations across various regions, requiring the implementation of management strategies tailored to each region59.
RC emerges as a highly effective solution for managing EBD, offering significant improvements in delay, cost efficiency, congestion management, and throughput stability compared to CC. While the paradigm shows immense promise, addressing challenges such as ownership costs, security concerns, and regulatory variations will be crucial to its broader adoption and sustained impact on educational systems.
In this study, we introduced RC as a promising solution for storing and processing EBD at a regional level. This approach aims to reduce reliance on CC servers during peak hours, helping to alleviate the strain on public networks, improve real-time responses, and enhance the use of EBD for educational purposes. Our findings show that implementing RC significantly decreases the workload on the primary network and improves the performance of educational data communication. Looking ahead, future efforts could focus on addressing the challenges associated with implementing RC, such as the costs of ownership, security concerns, and regional cyber regulations. Therefore, further research is needed to explore the scalability and efficiency of RC in managing educational data across various jurisdictions and regions. Additionally, examining how emerging technologies like edge computing and blockchain can be integrated with RC, may provide valuable insights into improving educational data management and communication systems.
The datasets generated during the experimentation of this study are available from the corresponding author on reasonable request.
Badshah, A. et al. Towards smart education through internet of things: A survey. ACM Comput. Surv. 56, 1–33 (2023).
Article MATH Google Scholar
Das, N. Digital education as an integral part of a smart and intelligent city: A short review. Dig. Learn. Based Edu. Trans. Phys. Barriers 81–96 (2023).
Cosentino, G. & Giannakos, M. Multisensory interaction and analytics to enhance smart learning environments: A systematic literature review. IEEE Trans. Learn. Technol. (2023).
AlMalki, H. A. & Durugbo, C. M. Systematic review of institutional innovation literature: towards a multi-level management model. Manag. Rev. Quart. 73, 731–785 (2023).
MATH Google Scholar
Qureshi, M. S., Daud, A., Hayat, M. K., Song, M. & Park, Y. Heading towards sub-discipline rankings for higher education institutions. IEEE Trans. Comput. Soc. Syst. 10, 3536–3555 (2023).
Article Google Scholar
Raj, A. et al. Revealing ai-based ed-tech tools using big data. In 2023 3rd International Conference on Innovative Practices in Technology and Management (ICIPTM), 1–6 (IEEE, 2023).
Bai, X. et al. Educational big data: Predictions, applications and challenges. Big Data Res. 26, 100270 (2021).
Article MATH Google Scholar
Daud, A. et al. Predicting student performance using advanced learning analytics. https://doi.org/10.1145/3041021.3054164 (International World Wide Web Conferences Steering Committee, Republic and Canton of Geneva, CHE (2017).
Baig, M. I., Shuib, L. & Yadegaridehkordi, E. Big data in education: A state of the art, limitations, and future research directions. Int. J. Educ. Technol. High. Educ. 17, 1–23 (2020).
Article MATH Google Scholar
Ritchie, H., Samborska, V., Ahuja, N., Ortiz-Ospina, E. & Roser, M. Global education. Our World in Data. https://ourworldindata.org/global-education (2023).
Aljohani, N. R. et al. An integrated framework for course adapted student learning analytics dashboard. Comput. Hum. Behav. 92, 679–690 (2019).
Article MATH Google Scholar
Al-Shoqran, M. & Shorman, S. A review on smart universities and artificial intelligence. Fourth Ind. Rev. Implement. Artif. Intell. Grow. Bus. Success 281–294 (2021).
Analysis, S. E. M. I. Smart education market by end-user, component and geography – forecast and analysis 2023-2027. https://www.technavio.com/report/smart-education-market-industry-analysis (2023).
Research, A. & Consulting. Big data market size to reach usd 473.6 billion by 2030. https://www.acumenresearchandconsulting.com/press-releases/big-data-market (2023).
Statista. Volume of data/information created, captured, copied, and consumed worldwide from 2010 to 2020, with forecasts from 2021 to 2025. https://www.statista.com/statistics/871513/worldwide-data-created/ (2023).
in Education, B. D. A. Volume of educational data. https://www.alliedmarketresearch.com/big-data-analytics-in-education-market/ (2023).
Sorour, A. & Atkins, A. S. Big data challenge for monitoring quality in higher education institutions using business intelligence dashboards. J. Electronic Sci. Technol. 100233 (2024).
Suganeshwari, G. & Divya, D. Deep learning in big data: Challenges and perspectives. Big Data Comput. 132–144 (2024).
Parra-Ullauri, J. M. et al. Federated analytics for 6g networks: Applications, challenges, and opportunities. IEEE Network (2024).
Zou, H., Yao, J., Zhang, Y. & Huang, X. The influence of teachers’ intrinsic motivation on students’ intrinsic motivation: The mediating role of teachers’ motivating style and teacher-student relationships. Psychol. Sch. 61, 272–286 (2024).
Article Google Scholar
Badshah, A. et al. Use of regional computing to minimize the social big data effects. Comput. Indus. Eng. 171, 108433 (2022).
Article MATH Google Scholar
Dautov, R. et al. Data processing in cyber-physical-social systems through edge computing. IEEE Access 6, 29822–29835. https://doi.org/10.1109/ACCESS.2018.2839915 (2018).
Article MATH Google Scholar
Goudarzi, M., Wu, H., Palaniswami, M. & Buyya, R. An application placement technique for concurrent iot applications in edge and fog computing environments. IEEE Trans. Mob. Comput. 20, 1298–1311. https://doi.org/10.1109/TMC.2020.2967041 (2021).
Article MATH Google Scholar
Badshah, A., Ghani, A., Daud, A., Chronopoulos, A. T. & Jalal, A. Revenue maximization approaches in IaaS clouds: Research challenges and opportunities. Trans. Emerg. Telecommun. Technol. e4492 (2020).
Fischer, C. et al. Mining big data in education: Affordances and challenges. Rev. Res. Educ. 44, 130–160 (2020).
Article MATH Google Scholar
Ray, S. & Saeed, M. Applications of educational data mining and learning analytics tools in handling big data in higher education. Appl. Big Data Anal. Trends, Issues Chall. 135–160 (2018).
Amjad, T., Daud, A. & Aljohani, N. R. Ranking authors in academic social networks: a survey. Library Hi Tech 36, 97–128 (2018).
Article MATH Google Scholar
Baig, M. I., Shuib, L. & Yadegaridehkordi, E. Big data in education: a state of the art, limitations, and future research directions. Int. J. Educ. Technol. High. Educ. 17, 1–23 (2020).
Article MATH Google Scholar
Badshah, A. et al. Big data applications: overview, challenges and future. Artif. Intell. Rev. 57, 290 (2024).
Article MATH Google Scholar
Munshi, A. A. & Alhindi, A. Big data platform for educational analytics. IEEE Access 9, 52883–52890. https://doi.org/10.1109/ACCESS.2021.3070737 (2021).
Article MATH Google Scholar
Yuwen, Z., Changqin, H., Qintai, H., Jia, Z. & Yong, T. Personalized learning full-path recommendation model based on lstm neural networks. Inf. Sci. 444, 135–152. https://doi.org/10.1016/j.ins.2018.02.053 (2018).
Article Google Scholar
Kanth, R., Laakso, M.-J., Nevalainen, P. & Heikkonen, J. Future educational technology with big data and learning analytics. In 2018 IEEE 27th International Symposium on Industrial Electronics (ISIE), 906–910, https://doi.org/10.1109/ISIE.2018.8433753 (2018).
Zhang, N., Geng, B., Hu, W. & Wen, R. The applications of big data analysis in student management education. In 2021 2nd International Conference on Big Data and Informatization Education (ICBDIE), 55–58, https://doi.org/10.1109/ICBDIE52740.2021.00021 (2021).
Liang, J. Research on the application of big data in the informatization of higher education management mode. In 2020 International Conference on Intelligent Transportation, Big Data & Smart City (ICITBS), 799–802, https://doi.org/10.1109/ICITBS49701.2020.00175 (2020).
Hao, W. Empirical study on the application of flipper classroom innovation teaching under the context of big data. In 2021 International Conference on Computer Technology and Media Convergence Design (CTMCD), 138–142, https://doi.org/10.1109/CTMCD53128.2021.00037 (2021).
Inayatulloh et al. Extended e-learning model to support home schooling with collaboration between teacher, parents and student. In 2022 IEEE International Conference of Computer Science and Information Technology (ICOSNIKOM), 1–6, https://doi.org/10.1109/ICOSNIKOM56551.2022.10034900 (2022).
Hu, G., Liu, W. & Xu, H. Research on hybrid teaching assessment driven by big data. In 2022 2nd International Conference on Big Data Engineering and Education (BDEE), 206–209, https://doi.org/10.1109/BDEE55929.2022.00042 (2022).
Daniel, S. J. Education and the covid-19 pandemic. Prospects 49, 91–96 (2020).
Article PubMed PubMed Central MATH Google Scholar
Shemshack, A. & Spector, J. M. A systematic literature review of personalized learning terms. Smart Learn. Environ. 7, 1–20 (2020).
Article MATH Google Scholar
Karatas, M., Eriskin, L., Deveci, M., Pamucar, D. & Garg, H. Big data for healthcare industry 4.0: Applications, challenges and future perspectives. Exp. Syst. Appl.200, 116912 (2022).
Zeadally, S., Siddiqui, F., Baig, Z. & Ibrahim, A. Smart healthcare: Challenges and potential solutions using internet of things (iot) and big data analytics. PSU Res. Rev. 4, 149–168 (2020).
Article Google Scholar
Sasubilli, S. M., Kumar, A. & Dutt, V. Improving health care by help of internet of things and bigdata analytics and cloud computing. In 2020 International Conference on Advances in Computing and Communication Engineering (ICACCE), 1–4 (IEEE, 2020).
Akherfi, K., Gerndt, M. & Harroud, H. Mobile cloud computing for computation offloading: Issues and challenges. Appl. Comput. Inf. 14, 1–16 (2018).
MATH Google Scholar
Awad, A. I., Fouda, M. M., Khashaba, M. M., Mohamed, E. R. & Hosny, K. M. Utilization of mobile edge computing on the internet of medical things: A survey. ICT Express 9, 473–485 (2023).
Article Google Scholar
Raj, J. S. & Jennifer, S. Optimized mobile edge computing framework for iot based medical sensor network nodes. J. Ubiq. Comput. Commun. Technol. (UCCT) 3, 33–42 (2021).
MATH Google Scholar
Talebkhah, M. et al. Edge computing: architecture, applications and future perspectives. In 2020 IEEE 2nd International Conference on Artificial Intelligence in Engineering and Technology (IICAIET), 1–6 (IEEE, 2020).
Li, C., Xue, Y., Wang, J., Zhang, W. & Li, T. Edge-oriented computing paradigms: A survey on architecture design and system management. ACM Comput. Sur. (CSUR) 51, 1–34 (2018).
ADS MATH Google Scholar
Akherfi, K., Gerndt, M. & Harroud, H. Mobile cloud computing for computation offloading: Issues and challenges. Applied computing and informatics 14, 1–16 (2018).
Article MATH Google Scholar
Abirami, S. & Chitra, P. Energy-efficient edge based real-time healthcare support system. In Advances in computers, vol. 117, 339–368 (Elsevier, 2020).
Dong, P. et al. Edge computing based healthcare systems: Enabling decentralized health monitoring in internet of medical things. IEEE Network 34, 254–261 (2020).
Article MATH Google Scholar
Rajavel, R., Ravichandran, S. K., Harimoorthy, K., Nagappan, P. & Gobichettipalayam, K. R. Iot-based smart healthcare video surveillance system using edge computing. J. Ambient. Intell. Humaniz. Comput. 13, 3195–3207 (2022).
Article Google Scholar
Kumar, D., Maurya, A. K. & Baranwal, G. Iot services in healthcare industry with fog/edge and cloud computing. In IoT-based data analytics for the healthcare industry, 81–103 (Elsevier, 2021).
Rahmani, A. M. et al. Exploiting smart e-health gateways at the edge of healthcare internet-of-things: A fog computing approach. Futur. Gener. Comput. Syst. 78, 641–658 (2018).
Article MATH Google Scholar
Lo’ai, A. T., Mehmood, R., Benkhlifa, E. & Song, H. Mobile cloud computing model and big data analysis for healthcare applications. IEEE Access4, 6171–6180 (2016).
Alshammari, H., El-Ghany, S. A. & Shehab, A. Big iot healthcare data analytics framework based on fog and cloud computing. Journal of Information Processing Systems 16, 1238–1249 (2020).
MATH Google Scholar
Amo, D. et al. Seven principles to foster privacy and security in educational tools: Local educational data analytics. In Eighth International Conference on Technological Ecosystems for Enhancing Multiculturality, 730–737 (2020).
Braunack-Mayer, A. J., Street, J. M., Tooher, R., Feng, X. & Scharling-Gamba, K. Student and staff perspectives on the use of big data in the tertiary education sector: A scoping review and reflection on the ethical issues. Rev. Educ. Res. 90, 788–823 (2020).
Article Google Scholar
Bauer, E. Improving operational efficiency of applications via cloud computing. IEEE Cloud Computing 5, 12–19 (2018).
Article MATH Google Scholar
Alwakeel, A. M. An overview of fog computing and edge computing security and privacy issues. Sensors 21, 8226 (2021).
Article ADS PubMed PubMed Central MATH Google Scholar
Download references
These authors contributed equally: Bader Alshemaimri, Afzal Badshah, Ali Daud, Amal Bukhari, Raed Alsini and Omar Alghushairy.
Software Engineering Department, College of Computing and Information Sciences, King Saud University, Riyadh, Saudi Arabia
Bader Alshemaimri
Department of Software Engineering, University of Sargodha, Sargodha, Pakistan
Afzal Badshah
Faculty of Resilience, Rabdan Academy, Abu Dhabi, United Arab Emirates
Ali Daud
Department of Information Systems and Technology, College of Computer Science and Engineering, University of Jeddah, Jeddah, Saudi Arabia
Amal Bukhari & Omar Alghushairy
Department of Information Systems, Faculty of Computing and Information Technology, King Abdulaziz University, Jeddah, Saudi Arabia
Raed Alsini
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
All authors contributed equally to this study. Bader Alshemaimri conceptualized the framework and methodology. Afzal Badshah led the experimental analysis and manuscript drafting. Ali Daud supervised the research and provided critical revisions. Amal Bukhari managed data collection and evaluation. Raed Alsini validated the computing model and optimized performance. Omar Alghushairy handled system integration and testing. All authors reviewed and approved the final manuscript.
Correspondence to Afzal Badshah.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
Alshemaimri, B., Badshah, A., Daud, A. et al. Regional computing approach for educational big data. Sci Rep 15, 7619 (2025). https://doi.org/10.1038/s41598-025-92120-7
Download citation
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-025-92120-7
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
Advertisement
Scientific Reports (Sci Rep)
ISSN 2045-2322 (online)
© 2025 Springer Nature Limited
Sign up for the Nature Briefing: AI and Robotics newsletter — what matters in AI and robotics research, free to your inbox weekly.