2022 Fiscal Year Annual Research Report
Mixed-Clairvoyance Task Offloading and Scheduling in Multi-access Edge Computing Systems: From Combinatorial Optimization to Machine Learning
| Project/Area Number |
20K19794
|
|---|
| Research Institution | Osaka Metropolitan University |
|---|
Principal Investigator |
江 易翰 大阪公立大学, 大学院工学研究科, 助教 (10824196)
|
|---|
| Project Period (FY) |
2020-04-01 – 2023-03-31
|
| Keywords | Edge computing / Internet of Things / Age of information / Serverless computing |
|---|
| Outline of Annual Research Achievements |
The overall research achievements can be summarized as follows.
(1) The problem of cotask processing in multi-access edge computing (MEC) systems can be formulated as an NP-hard combinatorial problem. For this, we proposed a non-clairvoyant deep dual learning method to update the primal and dual variables (governed by two deep neural networks) iteratively.
(2) The problem of information sampling and transmission scheduling in MEC systems with serverless computing can be formulated as another NP-hard combinatorial problem. For this, we designed both offline (clairvoyant) and online (non-clairvoyant) age-efficient algorithms for the information sampling and transmission scheduling with and without the prior knowledge of the invocations of serverless functions, respectively.
|
