Received: 8 April 2018 Revised: 24 December 2018 Accepted: 24 December 2018
DOI: 10.1002/cpe.5162
RESEARCH ARTICLE
Distributed Fog-to-Cloud computing system: A minority
game approach
Eman AbdElhalim Marwa Obayya Sherif Kishk
ECE Mansoura University, Mansoura, Egypt
Correspondence
Eman AbdElhalim, ECE Mansoura University,
Mansoura, Egypt.
Email: eman_haleim@mans.edu.eg
Summary
With the ever-increasing popularity of resource-intensive mobile applications, today,
Fog-to-Cloud (F2C) computing system becomes a prominent technology for the next gener-
ation wireless networks. Despite the benefits of fog computing regarding localized storage
and processing, it suffers from restricted power allowance and computational capability of the
edge nodes. User nodes also may suffer from extensive delay, especially in offloading periods.
Therefore, it is essential to develop a distributed mechanism for users' computation in offload-
ing periods. According to this mechanism, not only the computational servers are exploited
at their best capacity but also the users' latency constraints fulfilled. Consequently, this paper
develops automated distributed fog computing for computational offloading using the theory
of minority game. The proposed scheme achieves user satisfaction latency deadline as well as
Quality-of-Experience. Moreover, it guarantees an adaptive equilibrium level of F2C computing
system, which is suitable for heterogeneous wireless networks.
KEYWORDS
distributed mechanism, Fog-to-Cloud, Internet of Things, minority game, offloading,
Quality-of-Experience
1 INTRODUCTION
With increasing smart devices usage, and proliferation of data-hungry applications which become inevitable in health care, social networking, and
so on, consider a pervasive wireless network of densely distributed resource-limited things, all able to transfer in the real-time large data volume
of a heterogeneous environment. Therefore, due to the current computing-bandwidth-energy limitations of the wireless resources domain, a
system of this complexity becomes unfeasible. However, two new paradigms, namely, fog computing (FC) and the Internet of Things (IoT),
1
open
the doors for this complex system to be achievable.
The primary definition of FC was introduced by Cisco.
2
As in the work of Liu et al,
3
FC is a novel computing paradigms that aim to bring
the cloud computing (CC) services and facilities close to data sources, which can reduce the latency and cost of delivering data to a remote
cloud.
3
cloud computing is a general term for delivering hosted services over the internet.
4
Furthermore, it enables companies to consume cloud
resources, such as a virtual machine (VM), storage, or an application, as a utility rather than having to build and maintain computing infrastructures
in-house.
5
Notwithstanding the benefits of cloud computing, it introduces large delay especially in delayed sensitive applications.
6
For this
purpose, the concept of fog cloud or fog computing (FC) recently appeared.
The FC paradigm has a number of advantages over the cloud approach: reduces the traffic sent to core cloud, which lead to reduce congestion
and latency; provides a platform (by using edge devices resources) for filtering and analysis of the data generated by sensors or user nodes (UNs),
and reduces the transmitted traffic to the core cloud by allowing the placement of filtering operators close to the source of data as shown in
Figure 1. The primary definition of fog nodes (FNs) can be resource-rich devices such as cloudlet.
7
This cloudlet, fog node (FN), consists of a
small-scale data center or cluster of machines designed to quickly render cloud computing services to portable devices, such as smartphones,
wearable devices, and tablets within a close geographical area.
User nodes (UNs) utilize cloudlet to offload (part or all of their heavy-computation and latency-sensitive) tasks for remote execution.
8
Although
offloading can reduce energy consumption at the UNs, it may also incur a large execution delay which includes transmission time between the
Concurrency Computat Pract Exper. 2019;e5162. wileyonlinelibrary.com/journal/cpe © 2019 John Wiley & Sons, Ltd. 1 of 18
https://doi.org/10.1002/cpe.5162