期刊
IEEE TRANSACTIONS ON COMMUNICATIONS
卷 70, 期 12, 页码 7858-7871出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCOMM.2022.3214885
关键词
Receivers; Servers; Optimization; Encoding; Indexes; Videos; Delays; Proactive caching; dynamic features; index-coding
资金
- Natural Sciences and Engineering Research Council of Canada (NSERC)
- Qatar National Research Fund, QNRF [AICC03-0530-200033]
This article introduces a dynamic caching scheme that updates cache content at the edge and receivers by index-coding the update messages with delivery messages. The scheme reduces downloaded traffic and helps decrease other QoS metrics.
Proactive caching shows great potential to minimize peak download rates by caching popular data, in advance, at the edge. Fast-changing file features, such as fast-changing file popularities and fast-changing file contents (data freshness), represent a challenge for proactive caching if cache content update is much slower, which decreases the efficiency and usability of caching. We present a dynamic caching scheme that updates local user caches and optimizes the use of caching resources. The developed scheme index-code the updates with the delivery messages. The developed scheme is presented for a network with one cache-enabled server, that has a pool of files, communicating with K cache-enabled receivers with requests limited to the server's file pool. The developed scheme assumes partial knowledge of features variation. Asynchronous file delivery is assumed as a result of non-flexible receivers' request timing. We show that the file delivery messages can be used to proactively and constantly update the receivers' finite caches by index-coding the update messages with delivery messages at no additional rate-cost. We also show that this mechanism reduces the downloaded traffic and can be used to reduce other QoS metrics.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据