The realm of the Internet of Things (IoT), while continually transforming as a novel paradigm in the nexus of technology and education, still contends with numerous obstacles that hinder its incorporation into higher education institutions’ (HEIs) e-learning platforms. Despite substantial strides in IoT utilization from industrialized nations—the United States, the United Kingdom, Japan, and China serving as prime exemplars—the scope of its implementation in developing countries, notably Saudi Arabia, Malaysia, Pakistan, and Bangladesh, lags behind. A significant gap exists in research centered on the trajectory of IoT integration within e-learning systems of economically disadvantaged nations. Specifically, this study centers on Saudi Arabia to illuminate the main factors catalyzing or encumbering IoT uptake within its HEIs’ e-learning sector. As a preliminary step, this research has embarked on an exhaustive dissection of prior studies to unearth critical variables implicated in the IoT adoption process. Subsequently, we employed an inferential methodology, amassing data from 384 respondents in Saudi Arabian HEIs. Our examination divulges that usability, accessibility, technical support, and individual proficiencies considerably contribute to the rate of IoT incorporation. Furthermore, our data infer that financial obstacles, self-efficacy, interactive capability, online surveillance, automated attendance tracking, training programs, network and data safeguarding measures, and relevant tools significantly influence IoT adoption. Contrarily, factors such as accessibility, internet quality, infrastructure preparedness, usability, privacy concerns, and faculty support appeared to have a negligible impact on the adoption rates within HEIs. This research culminates in offering concrete recommendations to bolster IoT integration within Saudi Arabian HEIs, presenting valuable insights for government entities, policy architects, and HEIs to address the hurdles associated with IoT implementation in the higher education sector.

Smart parking is a crucial component of smart cities that aims to enhance the efficiency and sustainability of urban environments. It employs technology such as sensors and IoT devices to optimize the use of parking resources and improve drivers’ experiences. By reducing traffic congestion, decreasing air pollution, and enhancing accessibility, smart parking systems can contribute to the overall well-being of urban areas. IoT-enabled smart parking refers to the application of IoT technology to optimize and improve parking efficiency in smart cities. However, security and privacy challenges in IoT-enabled smart parking pose risks and concerns related to the collection and use of data by parking systems, such as unauthorized access or misuse of data, potential data breaches, and the need to ensure responsible data collection and usage to maintain user trust and confidence. To address these challenges, we propose a novel hybrid approach to trust management using machine learning algorithms to enhance the security and privacy of the system. Our approach consists of SVM and ANNs, taking into account credibility, availability, and honesty as key parameters. Furthermore, we use ensemble machine learning to select the best-predicted model from different trained models, leading to efficient performance and a trustworthy environment. Our results show that the proposed hybrid SVM classifier with a trust parameters approach achieved an accuracy of 96.43% in predicting and eliminating malicious or compromised nodes.

Load balancing is a very important concept in cloud computing. In this work, studies are conducted on workload traces at Los Alamos National Lab (LANL). The jobs in this trace are asymmetric in nature as most of them have small time limit. Two hybrid algorithms, a Genetic Algorithm combined with First Come First Serve (GA_FCFS) and Genetic Algorithm combined with Round Robin (GA_RR), are proposed here. The results obtained are compared with the existing First Come First Serve (FCFS), Round Robin (RR) and Genetic Algorithm (GA). Makespan and Resource Utilization are used for the comparison of results. In terms of Makespan, it is observed that GA_RR outperforms the other methods for all the batch sizes. Although the performance of GA_FCFS is much better than that of the other three well-established algorithms FCFS, RR and GA, it is still worse than that of the GA_RR algorithm for all the cases. GA_RR performs best in terms of Resource Utilization also and GA_FCFS is a close competitor. Overall, GA_RR outperforms all the other algorithms.