Loading…

Stability, Rate, and Delay Analysis of Single Bottleneck Caching Networks

Caching has been widely considered an efficient way of reducing and balancing the growing traffic in communications networks in recent years. The cache network of interest consists of one content server connected via a shared link to a number of caching nodes, also known as a single bottleneck cachi...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on communications 2016-01, Vol.64 (1), p.300-313
Main Authors: Rezaei, Fatemeh, Khalaj, Babak H.
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Caching has been widely considered an efficient way of reducing and balancing the growing traffic in communications networks in recent years. The cache network of interest consists of one content server connected via a shared link to a number of caching nodes, also known as a single bottleneck caching network. In this paper, for the first time, the stochastic requests traffic model in such networks is considered and a performance analysis is provided based on such a realistic assumption. In addition, we introduce new comprehensive performance metrics, which simultaneously take into account, the cache hit probability, load on the bottleneck link, and requests arrival rates. The main contribution of this paper is to present a system model based on queuing theory and provide an analysis of the stability, maximum stable throughput, load on the bottleneck link, and average response delay for various coded and uncoded caching schemes. Moreover, we propose a novel hybrid scheme that improves the shared link utilization factor, maximum stable throughput, and delay of single bottleneck caching networks compared to existing methods. Our results, validated against simulations and real trace-driven experiments, provide interesting insights into the performance of single bottleneck caching networks.
ISSN:0090-6778
1558-0857
DOI:10.1109/TCOMM.2015.2498177