Photon Netw Commun (2016) 32:348–358
DOI 10.1007/s11107-016-0665-9
ORIGINAL PAPER
Priority-based content processing with Q-routing
in information-centric networking (ICN)
Sibendu Paul
1
· Bitan Banerjee
2
· Amitava Mukherjee
3
· Mrinal K. Naskar
1
Received: 28 April 2016 / Accepted: 12 September 2016 / Published online: 26 September 2016
© Springer Science+Business Media New York 2016
Abstract With exponential increase in the number of users
and available data, service providers are facing hard times to
satisfy and improve end user experience. Researchers have
come up with the idea of exploiting increasing number of
routers in a network, and it leads to the development of
information-centric networking (ICN). Efficient usage of the
in-network caches and content forwarding methodology are
the key issues in an ICN architecture. ICN reduces average
hop count and correspondingly average content download
delay because the intra-domain routers in ICN have storage
capacity and they can act as temporary content provider. In
this paper, we address the content management issue in a
cache with finite storage capability and propose an efficient
content management policy that changes a router to a self-
sustained cache. We propose a novel methodology to process
content packets in the buffer of a cache and correspondingly
reduce the propagation delay through a cache. We simulate
our proposed algorithm over real-life network environment
and evaluate the performance of different user experience
metrics, e.g. average latency, throughput, goodput, and link
load. Simulation results suggest that our proposed model
B
Bitan Banerjee
bitan@ualberta.ca
Sibendu Paul
sibendu.paul01@gmail.com
Amitava Mukherjee
amitava.mukherjee@in.ibm.com
Mrinal K. Naskar
mrinaletce@gmail.com
1
Department of Electronics and Telecommunications
Engineering, Jadavpur University, Kolkata, India
2
Department of Electrical and Computer Engineering,
University of Alberta, Edmonton, Canada
3
IBM India Private Ltd., Kolkata, India
outperforms the existing state-of-the-art on-path caching
strategies.
Keywords ICN · Routing · Delay · Link load · Hit rate ·
Markov chain
1 Introduction
In the recent years, a paradigm shift in the communica-
tion model has taken place with the shift in focus from
the client–server model to the dissemination and content
retrieval. Initially networking was a user-based service;
however, slowly amount of information has outgrown the
number of users and it motivated researchers to imagine a
parallel networking universe with information as the first-
class citizen. In present networking model, data packets
are delivered between two end users independent of net-
work topology, content location, and popularity. However,
researchers noticed that users are more concerned about the
information rather than the source of the content [1]. Treating
contents as the primary objects guided the transition from
peer-to-peer networking to ICN [2]. Several ICN models
and infrastructures had been proposed as better replace-
ments of current Internet methodologies [3–5]. To satisfy
ever-increasing users’ demand and a burst of data on a fixed
infrastructure, Chai et al. [1] proposed opportunistic caching
of contents (i.e. cache the contents in nodes having the max-
imum probability of hit) for better utilization of network
resources and for speeding up the content delivery.
Along with the steep increase of users, the popularity of
ICN system has increased exponentially. Therefore, every
ICN realization will inherently face a key challenge to sat-
isfy the quality-of-service (QoS) requirements using a fixed
routing and caching mechanism. Earlier, researchers used
123
@Phil_Robichaud
@deepthiw
@JoseServera