Vol.:(0123456789)
SN Computer Science (2021) 2:483
https://doi.org/10.1007/s42979-021-00907-y
SN Computer Science
ORIGINAL RESEARCH
Fair Resource Allocation Policies in Reverse Auction‑Based Cloud
Market
Dinesh Kumar
1
· Gaurav Baranwal
2
· Deo Prakash Vidyarthi
3
Received: 16 May 2021 / Accepted: 26 September 2021
© The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd 2021
Abstract
The increasing number of Internet of Things (IoT) applications and their dependence on cloud computing for computational
services has resulted in the cloud market’s growth. This growth has attracted many business organizations to offer cloud ser-
vices, leading to significant competition amongst the cloud service providers. Reverse auction has been widely used to model
such competition when there are many cloud service providers. A study on fair resource allocation mechanisms is performed
in this work, and a family of such mechanisms is proposed. This work considers a reverse auction-based cloud market where
users submit their combinatorial bids. This work emphasizes the importance of fairness in cloud resource allocation and its
implementation in a cloud market. The proposed priority-based fair resource allocation mechanisms remove the bidder drop
problem in a cloud market where a few major cloud providers dominate and control the whole market. Performance of the
proposed fair resource allocation mechanisms is studied based on various metrics such as the number of winning auction
rounds, providers’ revenue, users’ procurement cost, etc., in a simulated environment. It is observed that the naïve and not
well-established providers in the market also win when fair mechanisms based on priority methods are implemented. They
also get a chance to win the auction and can offer the resources successfully to the customer.
Keywords Fairness · Resource allocation · Combinatorial auction · Cloud computing · Internet of Things (IoT)
Introduction
Two emerging technologies, Cloud computing and Internet
of Things (IoT), have evolved to complement each other.
Due to their complementary nature, each may take advan-
tage of the other. For example, IoT can be benefited from
the virtually unlimited processing and storage capacities
of the Cloud, whereas Cloud can be benefited from IoT by
spreading its scope to provide a variety of services in vari-
ous real-life scenarios. It further creates opportunities for an
IoT system by doing data integration, aggregation, sharing
and analysis.
Cloud computing is a business model of computing para-
digm in which computing resources and services are pro-
vided to its users over the internet on a pay-per-use basis
[1]. It has brought a revolutionary change in the IT indus-
try by delivering virtualized resources as a service [2]. US
‘National Institute of Standards and Technology,’ i.e. NIST
defines Cloud Computing as: “Cloud computing is a model
for enabling convenient, on-demand network access to a
shared pool of configurable computing resources (e.g., net-
works, servers, storage, applications and services) that can
be rapidly provisioned and released with minimal manage-
ment effort or service provider interaction” [3]. Resource
pooling, on-demand self-service, elasticity, pay-per-use,
etc., are some of the fundamental characteristics of cloud
computing [2].
In a Cloud environment, Cloud offers the resources to its
clients who pay for the resource usage. Thus, it is always
desired that the deal should be beneficial to both parties for
* Dinesh Kumar
dinesh.kumar@mnnit.ac.in
Gaurav Baranwal
gaurav.baranwal@bhu.ac.in
Deo Prakash Vidyarthi
dpv@mail.jnu.ac.in
1
Department of Computer Science and Engineering, Motilal
Nehru National Institute of Technology, Allahabad,
Prayagraj, UP 211004, India
2
Department of Computer Science, Institute of Science,
Banaras Hindu University, Varanasi, UP 221005, India
3
School of Computer and Systems Sciences, Jawaharlal Nehru
University, New Delhi 110067, India