Current Result Document : 23 / 23
| 한글제목(Korean Title) |
A Federated Architecture for Proactive Infotainment Content Caching in Moving Edge |
| 영문제목(English Title) |
A Federated Architecture for Proactive Infotainment Content Caching in Moving Edge |
| 저자(Author) |
Avi Deb Raha
Md. Shirajum Munir
Choong Seon Hong
|
| 원문수록처(Citation) |
VOL 49 NO. 01 PP. 1420 ~ 1422 (2022. 06) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
In moving edge computing, infotainment content caching is a potential strategy to lessen the ever-increasing data supply in wireless networks. Because caches have a limited capacity, it is essential to proactively anticipate content similarity and cache the most similar content in a local cache server. In this study, we consider moving vehicles as the cache servers, where infotainment contents are being proactively cached based on the similarity index of the current content. We devise a federated learning architecture, in which each roadside unit (RSU) acts as a client, and a base station (BS) plays the role of a central aggregator server. We adopt a cosine similarity metric to determine the similarity among the infotainment contents for finding the cache candidates at each RSU. Then, we develop an algorithm for the central aggregator to recommend infotainment contents for the specific service provider vehicle in advance. Additionally, we deploy a recurrent neural network for predicting an hour-ahead infotainment contents for moving edges. Our experiment shows a significant performance gain for predicting the similarity score for hour-ahead infotainment content caching, where the mean absolute error is 0.03. |
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
