Adaptive RSU Assignment and Transmission Scheduling of Delay-Critical Emergency Messages and AR Traffic in MEC-Enabled Vehicular Environments

다국어 초록 (Multilingual Abstract)

Emergency message dissemination and augmented reality (AR) delivery are emerging as core components of next-generation intelligent vehicular systems. However, the coexistence of these two traffic types introduces conflicting quality-of-service requirements. On one hand, emergency short messages demand ultra-low latency, high reliability, and strict temporal guarantees to ensure safety. On the other hand, AR content involves larger data payloads and can tolerate slightly higher delays, yet still requires spatial and contextual relevance to maintain user immersion. Balancing these heterogeneous demands poses a significant challenge for multi-access edge computing (MEC)-enabled vehicular networks, where both dynamic mobility and limited RSU coverage must be considered. To tackle this challenge, this study formulates a joint RSU assignment and transmission scheduling problem in a multi-server, multi-user MEC environment. The problem is shown to be NP-hard, motivating the development of a two-stage optimization framework. In the first stage, the Penalty-Minimizing RSU Selection (PMRS) algorithm determines the optimal RSU for each vehicle by jointly minimizing deadline and coverage penalties. In the second stage, a hybrid scheduling algorithm, termed Deadline-Aware Priority Scheduling (DAPS), is proposed. DAPS combines the advantages of earliest-deadline-first (EDF) scheduling with simulated annealing (SA)–based optimization to ensure timely delivery of emergency messages while maximizing throughput and resource utilization for delay-tolerant AR traffic. Extensive simulations under realistic vehicular mobility and traffic scenarios demonstrate that the proposed PMRS + DAPS framework consistently outperforms conventional heuristic and metaheuristic baselines in terms of delay reduction, deadline satisfaction, and overall system efficiency. The results highlight the framework’s potential for enabling reliable and delay-aware content dissemination in practical MEC-assisted vehicular environments.

목차 (Table of Contents)

• Acknowledgments i
• Abstract ii
• 1 Introduction 1
• 1.1 Edge Computing as an Enabler for Low-Latency AR and Safety Services 2
• 1.2 Coexistence of Heterogeneous Vehicular Content Types 2
• 1.3 Motivation 3
• 1.4 The Proposed Framework 3
• 1.5 Thesis Organization 4
• 2 Related Works 5
• 2.1 Integration of Augmented Reality and Emergency Messaging in Vehicular Networks 5
• 2.2 Edge-Enabled RSU Architectures for Delay and Reliability Optimization 6
• 2.3 RSU Assignment, Resource Allocation, and Scheduling Strategies 6
• 2.4 Research Gaps and Motivation for Joint RSU Assignment and Scheduling 7
• 3 Background 8
• 3.1 Optimization 8
• 3.2 Nonlinear and Combinatorial Optimization 9
• 3.3 Optimization Algorithms for Scheduling and Resource Allocation 9
• 3.3.1 Heuristic Algorithms 9
• 3.3.2 Metaheuristic Algorithms 10
• 3.3.3 Hybrid Optimization Approaches 11
• 4 System Model 12
• 4.1 Communication Model 14
• 5 Problem Definition and Formulation 17
• 5.1 Problem Definition 17
• 5.2 Problem Formulation 17
• 6 Algorithms 20
• 6.1 PMRS 20
• 6.2 DAPS 21
• 6.2.1 Phase 1 22
• 6.2.2 Phase 2 22
• 7 Performance Evaluation 26
• 7.1 Simulation Setup 26
• 7.2 Baselines 28
• 7.2.1 HSS 29
• 7.2.2 Genetic Algorithm 29
• 7.2.3 Simulated Annealing 29
• 7.3 Simulation Result Analysis 30
• 7.3.1 Emergency Short Messages 30
• 7.3.2 Emergency Supplementary Messages 35
• 7.3.3 Advertisements 40
• 7.3.4 Overall Results 46
• 8 Conclusion 52
• 8.1 Future Works 52
• 8.1.1 Semantic Encoding of Emergency and AR Contents 53
• 8.1.2 Semantic-Aware RSU Assignment and Scheduling 53
• 8.1.3 Semantic Performance Metrics in Evaluation 53
• 8.2 Concluding Remarks 53
• Appendix A Baseline Algorithms 60
• A.1 HSS Algorithm 60
• A.2 Genetic Algorithm 61
• A.3 Simulated Annealing Algorithm 63