Cache Replacement Schemes Based on Adaptive Time Window for Video on Demand Services in Femtocell Networks

Autor: Zohreh Hajiakhondi-Meybodi, Jamshid Abouei, Amir Hossein Fahim Raouf
Rok vydání: 2019
Předmět:
Zdroj: IEEE Transactions on Mobile Computing. 18:1476-1487
ISSN: 2161-9875
1536-1233
Popis: The cache replacement policy is a crucial phase in caching-based systems that deal with the process of selecting applicable cache contents. In this paper, we propose two novel cache replacement algorithms based on the dataset obtained from a typical wireless femto network. In the first algorithm, called Weighted Least Frequently used with an Adaptive Time Window (WLF-ATW), we aim to make a balance between the network's traffic and the recognition of popular contents. The WLF-ATW algorithm takes the frequency and the recency information of files into account to ascertain the popularity of contents. We suggest another new cache replacement policy namely Fairness Scheduling-based with an Adaptive Time Window (FS-ATW) that is based on fairness scheduling in order to minimize the user's access delay. The novelty of our proposed FS-ATW lies in ranking clients according to their last situations that lead to a further user's experience. The effectiveness of these new algorithms is evaluated from the cache hit ratio, transferred byte volume, user's access delay, user's experience, and load balance. A comprehensive numerical evaluation shows that the performance of the proposed WLF-ATW and FS-ATW algorithms is significantly better than some existing cache replacement strategies.
Databáze: OpenAIRE