Opportunistic Computation Offloading with Learning-based Prediction for UAV Clustering Networks
- 주제(키워드) Offloading , Drones , Cluster , Wireless , ANN
- 발행기관 아주대학교
- 지도교수 Young-Bae Ko
- 발행년도 2019
- 학위수여년월 2019. 2
- 학위명 석사
- 학과 및 전공 일반대학원 컴퓨터공학과
- 실제URI http://www.dcollection.net/handler/ajou/000000028539
- 본문언어 영어
- 저작권 아주대학교 논문은 저작권에 의해 보호받습니다.
초록/요약
Drones or Unmanned Aerial Vehicles (UAV) have recently become popular, especially in civilian and military applications. Examples of drone utilization are reconnaissance mission, packet delivery system, and surveillance system. As the technology grows to be more advanced, drone’s tasks become more complex and they need more computation power. Hence, clusters or swarms of drones are preferred since the swarms can offer more flexibility, reliability, and coverage. One weak point of drones is their limited computing and energy resources; therefore, it is hard for drones to process all the application on time. One possible solution to mitigate the problem is by doing offloading scheme between drone clusters. In this study, we proposed an opportunistic computation offloading system, which permits swarm drones with high computation applications to opportunistically borrow neighbor drone cluster’s available computation resources. Furthermore, we designed a response time estimation module by adopting artificial neural network technique in order to decide whether it is faster to complete the jobs by offloading the tasks to the nearby drone clusters or not. The offloading process will be only done if the estimated offloading scheme’s response time is smaller than the response time of executing the tasks locally. Our simulation results show that the proposed scheme can decrease the application response time by conducting opportunistic offloading process.
more목차
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 BACKGROUND AND RELATED WORK 9
2.1 COMPUTATION OFFLOADING 9
2.2 MOBILITY MODEL FOR UAV AD HOC NETWORKS 13
CHAPTER 3 OPPORTUNISTIC COMPUTATION OFFLOADING SCHEME 15
3.1 OPPORTUNISTIC COMPUTATIONAL OFFLOADING SCHEME 17
3.2 COMPUTATION OFFLOADING DECISION MODULE 20
3. ANN-BASED RESPONSE TIME PREDICTION MODULE 25
3.4 TASK OFFLOADING SERVICE 28
CHAPTER 4 PERFORMANCE EVALUATION 29
4.1 SIMULATION ENVIRONMENT 29
4.2 PERFORMANCE EVALUATION 31
4.2.1 OFFLOADING PERFORMANCE 32
4.2.2 PREDICTION MODULE 37
4.2.3 ENERGY CONSUMPTION 41
CHAPTER 5 CONCLUSION 43
REFERENCES 45