| Presentation | 2021-03-04 Latency-based Dynamic Edge Offloading Framework of Containerized Applications for Connected Cars Toru Furusawa, Hiroshi Abe, Akihiro Nakao, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Edge offloading, which executes applications for vehicles on edge servers, enables the reduction of power consumption and development costs of in-vehicle devices. In particular, edge offloading is effective for the kinds of applications that require a response within a few seconds but do not directly affect driving control, such as license plate image recognition of surrounding vehicles. However, when an increase in network delay or a decrease in the processing capacity of the edge server occurs due to vehicle movement or changes in the surrounding environment, the application processing time temporarily and significantly increases. On the other hand, with the spread of container virtualization technology in recent years, it is predicted that container-based applications will be able to run on in-vehicle devices and edge servers in addition to the cloud in the future. In this paper, we propose a framework for offloading container-based applications for vehicles to the optimal destination among in-vehicle devices, edge servers, and clouds. To stably continue offloading by dynamically changing an offloading destination based on a measured value of processing delay between an in-vehicle device, each edge server, and a cloud. The framework is implemented using Kubernetes. Experiments show that when network delays or edge server CPU processing time degradation occurs, the offloading destination switches within one second and offloading can continue stably while lowering the average processing time. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | connected car / edge computing / edge offloading / container / Kubernetes |
| Paper # | NS2020-131 |
| Date of Issue | 2021-02-25 (NS) |
| Conference Information | |
| Committee | IN / NS |
|---|---|
| Conference Date | 2021/3/4(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | General |
| Chair | Kenji Ishida(Hiroshima City Univ.) / Akihiro Nakao(Univ. of Tokyo) |
| Vice Chair | Kunio Hato(Internet Multifeed) / Tetsuya Oishi(NTT) |
| Secretary | Kunio Hato(Hiroshima City Univ.) / Tetsuya Oishi(KDDI Research) |
| Assistant | / Shinya Kawano(NTT) |
| Paper Information | |
| Registration To | Technical Committee on Information Networks / Technical Committee on Network Systems |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Latency-based Dynamic Edge Offloading Framework of Containerized Applications for Connected Cars |
| Sub Title (in English) | |
| Keyword(1) | connected car |
| Keyword(2) | edge computing |
| Keyword(3) | edge offloading |
| Keyword(4) | container |
| Keyword(5) | Kubernetes |
| 1st Author's Name | Toru Furusawa |
| 1st Author's Affiliation | Toyota Motor Corporation/The University of Tokyo(Toyota/Univ. of Tokyo) |
| 2nd Author's Name | Hiroshi Abe |
| 2nd Author's Affiliation | Toyota Motor Corporation(Toyota) |
| 3rd Author's Name | Akihiro Nakao |
| 3rd Author's Affiliation | The University of Tokyo(Univ. of Tokyo) |
| Date | 2021-03-04 |
| Paper # | NS2020-131 |
| Volume (vol) | vol.120 |
| Number (no) | NS-413 |
| Page | pp.pp.49-54(NS), |
| #Pages | 6 |
| Date of Issue | 2021-02-25 (NS) |