Author
Raouf, Amir Hossein Fahim, Abouei, J.
Publication Date
2018-09-25
Publication Place
-
IEEE
Subject
Caching, Cache replacement, Video on demand
Type
Document
Language
English
Digital
Yes
Manuscript
No
Library
Özyeğin University
Library Asset ID
978-1-5386-4916-9
Record ID
1c78514c-d25f-4ae4-9adb-777b0e8cdc3f
Date
2018-09-25
Sample Text
In this paper, we propose a novel two-phased cache replacement algorithm based on the dataset obtained from a typical wireless mesh network. In the first phase, we aim to make a trade-off between the networks resources and refresh the cache content. To deal with this issue, we apply a sliding window to propose a new parameter, called equivalent active time, to forecast the user's behaviour pattern, i.e., make a decision whether to update the cache contents or utilize the network's resources for other services. In the second phase, we introduce, for the first time, the parameter time to live (TTL) that shows the video popularity lifetime. Our proposed cache replacement strategy uses frequency counters and the TTL information in the victim selection process to prevent the cache pollution and make better use of the cache space. Numerical results show that our replacement algorithm outperforms some existing cache replacement strategies in term of a portion of satisfied requests.
DOI
10.1109/ICEE.2018.8472723
