DGIST Scholar: Energy-efficient Task Offloading and Sensing Decision for Edge Computing-enabled Vehicular Networks

Department of Electrical Engineering and Computer Science Theses Ph.D.

Cited time in webofscience

Cited time in scopus

Energy-efficient Task Offloading and Sensing Decision for Edge Computing-enabled Vehicular Networks

Title: Energy-efficient Task Offloading and Sensing Decision for Edge Computing-enabled Vehicular Networks

Description: 차량 네트워크; 엣지 컴퓨팅; 최적화 문제; 차량 센서 네트워크; 센싱 결정 기법; 강화 학습; 정보의 최신성; Vehicular networks; convex optimization; edge computing; resource allocation; sensing decision

Table Of Contents: I. INTRODUCTION 1
1.1 Outline and Contributions 3
1.1.1 Chapter 2 3
1.1.2 Chapter 3 3
1.1.3 Chapter 4 4
II. Task Offloading for Edge Computing-enabled Vehicular Networks 6
2.1 Contribution of This Chapter 6
2.2 Notations 7
2.3 System Model 7
2.3.1 Network Model 7
2.3.2 Cooperative Offloading Model 8
2.4 Total Energy Minimization 15
2.4.1 Problem Formulation 15
2.4.2 Optimal Solution in Multi-Vehicle Case 17
2.4.3 Optimal Solution in Single-Vehicle Case 19
2.5 Extension to Online Scenario 24
2.5.1 Network Model and Cooperative Offloading Models in Online Scenario 25
2.5.2 Online Optimization 25
2.6 Numerical Results 27
2.6.1 Single-Vehicle Case 28
2.6.2 Multi-Vehicle Case 29
2.7 Conclusion 35
III. Sensing Decision for Edge Computing-enabled Vehicular Sensor Networks 36
3.1 Contribution of This Chapter 36
3.2 System Model 36
3.2.1 Network Model 36
3.2.2 Vehicular Sensing Model 37
3.2.3 Performance metric 40
3.3 reinforcement learning (RL)-Based Sensing Decision Algorithm 41
3.3.1 Markov decision process (MDP) Formulation 41
3.3.2 Proposed Network Architecture 43
3.4 Simulation Results 43
3.5 Conclusion 45
IV. Conclusions 46
References 47