| Project/Area Number |
19K21539
|
|---|
| Project/Area Number (Other) |
18H06471 (2018)
|
|---|
| Research Category |
Grant-in-Aid for Research Activity Start-up
|
| Allocation Type | Multi-year Fund (2019)
Single-year Grants (2018) |
|---|
| Review Section |
1001:Information science, computer engineering, and related fields
|
|---|
| Research Institution | National Institute of Informatics |
|---|
Principal Investigator |
Chiang Yi-Han 国立情報学研究所, アーキテクチャ科学研究系, 特任助教 (10824196)
|
|---|
| Project Period (FY) |
2018-08-24 – 2020-03-31
|
| Project Status |
Completed (Fiscal Year 2019)
|
| Budget Amount *help |
¥2,990,000 (Direct Cost: ¥2,300,000、Indirect Cost: ¥690,000)
Fiscal Year 2019: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2018: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
|
|---|
| Keywords | モバイルエッジコンピューティング / オフロード / スケジューリング / コタスク / 混合整数非線形計画 / Mobile Edge Computing / Task Offloading / Delay Sensitivities / Actor-Critic Method / Reinforcement Learning / mobile edge computing / cotask offloading / cotask scheduling / cotask / mixed integer program / task offloading / task scheduling / offloading & scheduling |
|---|
| Outline of Research at the Start |
This research focuses on the problem of cotask-aware offloading and scheduling for MEC systems. Due to the problem intractability, the applicant designs efficient online algorithms, prove the achieved competitiveness, and then evaluate the performance through testbed experiments and simulations.
|
| Outline of Final Research Achievements |
Mobile edge computing (MEC) systems provide mobile devices (MDs) with low-latency cloud services by deploying edge servers (ESs) in the vicinity. In fact, various mobile applications may generate cotasks, each of which is completed only if all its constituent tasks are finished. In this research, we investigate the problem of joint cotask-aware offloading and scheduling in MEC systems (Cool-Edge), and we formulate it as a mixed integer non-linear program (MINLP) to minimize average cotask completion time (ACCT). To cope with the Cool-Edge problem, we propose two low-complexity algorithms to offload cotasks based on an LP rounding technique and schedule them according to an earliest-cotask-arrival-first rule, respectively, and we further prove the approximation factor jointly achieved by the two algorithms. Finally, we conduct testbed experiments and simulations to demonstrate the effectiveness of our proposed solution, and we also show how ACCT varies with the network environment.
|
| Academic Significance and Societal Importance of the Research Achievements |
コタスクの存在は、計算資源を効果的に活用するために、サブタスクのオフロードとスケジューリングの共同設計を必要とします。この研究では、混合整数非線形計画として定式化し、その問題のNP困難も証明しました。そして、設計したコタスクのオフロードとスケジューリングのアルゴリズムに基づいて、到達した近似率はシステムの計算能力と無線遅延によって決まります。テストベッドとシミュレーションの結果は、設計されたアルゴリズムがコタスクをバランスよくオフロードして、ほぼ同じタイミングでスケジュールできるので、ローカル計算とエッジ計算のトレードオフが改善されることを示し、ネットワークの計算資源を効率的に使用できました。
|
