Latency-Aware Edge Computing Architecture for Real-Time Immersive Learning in AR/VR-Enabled Smart Classrooms
Nilufar KholmatovaLecturer, Termez State Pedagogical Institute, Termez, Uzbekistan. xolotovanilufar04@gmail.com0009-0006-2653-5830
Dilfuza GafforovaLecturer, Termez State University, Termez, Uzbekistan. Uzbekistan.dilfuza@tersu.uz0009-0006-2791-3506
Khilola GofurovaLecturer, TIIAME National Research University, Tashkent, Uzbekistan. hilolagafurova85@gmail.com0000-0002-6135-5898
Zoxidakhon DjumayevaTeacher, Department of Primary Education Methodology, Denov Institute of Entrepreneurship and Pedagogy, Surkhandarya, Uzbekistan. z.jumayeva@dtpi.uz0009-0003-4056-8485
Mashkhura YuldashevaAssociate Professor, Bukhara State Pedagogical Institute, Bukhara, Uzbekistan. mashhurayuldasheva1973@gmail.com0000-0002-4048-2775
Zulkhumor ShukurovaKarshi State University, Karshi, Uzbekistan. zulxsh0815@gmail.com0009-0006-5732-4016
Gulbahor QurbonovaLecturer, Jizzakh State Pedagogical University, Jizzakh, Uzbekistan. gulbahor8808@gmail.com0000-0003-0941-9062
The growth of Augmented Reality (AR) and Virtual Reality (VR) technologies in education has created amazing possibilities to facilitate immersive teaching. Nevertheless, producing timely and immersive high-quality content in classrooms is often still restricted because of the limitations in the latency of standard cloud-computing models. The paper presents the Latency-Aware Edge Computing Architecture (LAECA), which enables end-to-end latency of less than 20 ms in an AR/VR-based smart classroom. The VE4T system previously laid the foundation by using edge computing for educational VR based on a Genetic-Simulated Annealing Algorithm (GSAA) task scheduler and reached the average response latency close to 84 ms. In this paper, LAECA makes three improvements over it. (i) multi-threshold adaptive task scheduler: the scheduler selectively routes render and tracking jobs by using real-time QoS information. (ii)unified AR/VR session manager with IoT-enabled classroom context awareness; and (iii) QoE-latency optimization by minimizing motion to photon delay and maximizing classroom immersive education quality. The simulation experiments done in NS3 with as many as 150 simultaneous students indicate that LAECA results in 75% decrease in average response time (21 ms in comparison to 84 ms in VE4T), intraframe prediction accuracy of 98.7% at 8k resolution, and an average QoE score of 8.9 out of 10. Such findings also prove that latency-aware edge computing can offer the essential infrastructure for immersive classroom experiences.