Page 173 - Proceedings of the 2017 ITU Kaleidoscope

CONTRACT THEORY BASED CACHING AND PRICING STRATEGY FOR CONTENT
CENTRIC NETWORKS

Chen Li, Jintian Li, Zhou Su, Qichao Xu

Shanghai University, P.R. China, Corresponding author: zhousu@ieee.org

ABSTRACT mal contract to improve the performance of content caching
in CCNs. Firstly, a novel two-layer heterogeneous network
Content centric networks (CCNs) have emerged to deliver a
large amount of contents in the networks. However, it has be- model is proposed in CCNs. In this heterogeneous network
come a new challenge to efficiently cache the contents in the model, users can access the contents from both Small-Cell
CCNs. Therefore, in this paper, we design a contract theory Base Stations (SBSs) and CCNs. Then, we present the opti-
based content caching scheme to improve the performance of mal contract to users and content provider to obtain the max-
CCNs. Firstly, a two-layer heterogeneous network model is imum utility. In addition, we compare the proposal with oth-
introduced to study the interaction between users and con- er previous schemes in simulation. Experiment results show
tent providers. Secondly, based on the contract theory, the that our proposal can be efficient to deliver contents in the
optimal caching and pricing strategy can be obtained under CCNs.
two constraints in CCNs. Finally, simulation experiments are The rest of the paper is organized as follows. Related work
carried out to prove that our proposal can efficiently improve is reviewed in Section 2. The system model is introduced in
the cache performance of CCNs. Section 3. The proposed solution is proposed in Section 4.
Simulation results are shown in Section 5. The conclusion of
Keywords — Content centric networks, edge cache, con-
this paper and the future work are given in Section 6.
tract theory

1. INTRODUCTION 2. RELATED WORK

An ever-increasing number of contents are delivered and Le et al. [10] compare several previous caching policies to e-
shared in content centric networks (CCNs) with the rapid valuate the performance of the network, whereby the caching
development of communication technologies. These large- policies with content replacement strategies are also estimat-
sized contents put a high pressure on the network opera- ed. Hajimirsadeghi et al. [11] develop an analytical frame-
tor to provide users with a satisfied quality of experience work for popular content distribution, in which the caching
(QoE). Different from the conventional IP-based scheme, the and pricing strategies can be determined through a game to
content-based scheme in CCNs [1–3] can reduce the content reach the Nash equilibrium. Su et al. [12] propose a nov-
transmission delay and consumption. Due to the limited el caching scheme to reduce the delivery latency in CCNs,
cache capacity in CCNs and network congestion, caching whereby the content store and pending interest table can be
the contents in the edge node becomes a novel method to utilized efficiently. Guo et al. [13] propose a novel archi-
improve the network performance. tecture to cache and transmit contents for cooperative nodes
With caching content at the edge of networks, the cached in CCNs, whereby the content popularity degree is consid-
content can be provided to the users from a nearby node ered. Zhang et al. [14] propose an implicit cooperative cache
where the delay for users to obtain the content can be re- scheme to cache the contents in CCNs, whereby the content
duced. Besides, as the content need not to be fetched from cache delay can be reduced and the cache resource utilization
the remote server, the network traffic can be reduced at the can be improved.
same time. Ramaswamy et al. [15] propose an utility-based data place-
Although many efforts have been made to study the content ment scheme in edge cache networks, whereby the virous
cache performance [4–6], most of them focus on the cache costs and caching benefits of the individual edge cache node
resource allocation [7–9], where the cache performance of are included. Wang et al. [16] propose a novel analytic mod-
the CCN can not be fully optimized. Besides, the curren- el based on the Markov-chain to estimate the caching perfor-
t optimal strategy and pricing strategy seldom take hetero- mance among different edge nodes, in which the bandwidth
geneous network into account. Therefore, it is a new and consumption among user equipments can be minimized.
open problem to design an efficient caching scheme for large Wang et al. [17] review the latest research works on mobile
amounts of contents in CCNs. edge networks including the application and use cases, which
To address these issues, we design a contract theory based can accommodate the user demands due to its distributed net-
caching and pricing strategy to provide users with the opti- work architecture. Wang et al. [18] propose a novel Markov-

– 157 –