Content uploaded by Chaker Abdelaziz Kerrache
Author content
All content in this area was uploaded by Chaker Abdelaziz Kerrache on May 20, 2024
Content may be subject to copyright.
1
A Review of Service Selection Strategies in Mobile
IoT Networks
Abdessalem Mohamed Hadjkouider, Chaker Abdelaziz Kerrache, Ahmed Korichi, Yesin Sahraoui, Carlos T.
Calafate, Sahraoui Dhelim, Asma Adnane
Abstract—With the proliferation of service-oriented
architectures, the ubiquity of mobile networks, and an
expanding spectrum of services available to users, the
service selection process has become paramount in de-
signing and implementing contemporary applications.
Service selection strategies in mobile networks involve
choosing the optimal services from many available op-
tions based on QoS (quality of service), bandwidth,
and mobility, ultimately augmenting network efficiency
and user experience. Nonetheless, particular challenges
can impact the efficacy of these strategies. This paper
presents a comprehensive survey of service selection
strategies in mobile networks, transitioning from tra-
ditional approaches to emerging techniques, while also
introducing a taxonomy classifying service selection
methods based on criteria including mobile devices
as service providers, architectural perspectives, com-
puting devices allocation, evaluation metrics, and ser-
vice selection methods. Additionally, We offer insights
into the latest QoS-aware and energy-centric service
selection methodologies, while addressing challenges
related to battery life and energy efficiency, Quality of
Service (QoS), security and privacy, as well as scalabil-
ity and resource management. Additionally, we discuss
prospective trajectories and future direction.
Index Terms—service selection; mobile networks;
QoS; IoT;
I. Introduction
Mobile networks have become an integral part of our
lives, connecting us to various services and applications
wherever we go. From making calls and sending text
messages to accessing the internet and streaming media,
mobile networks provide us with the means to stay con-
nected and access information on the go Fig.1. However,
with the increasing complexity and diversity of services
A. M. Hadjkouider and A. Korichi are with Kasdi
Merbah University of Ouargla, Ouargla, Algeria (emails:
hadjkouider.abdesselam@univ-ouargla.dz;ahmed.korichi@univ-
ouargla.dz).
C.A. Kerrache is with Laboratoire d’Informatique et de
Math´ematiques, Universit´e de Laghouat, Laghouat, Algeria (email:
ch.kerrache@lagh-univ.dz).
Y. Sahraoui is with National Higher School of Artificial Intelli-
gence, Algiers, Algeria (email: yacine.sahraoui@ensia.edu.dz).
C. T. Calafate is with the Computer Engineering Department
(DISCA), Universitat Polit`ecnica de Val`encia, Valencia, Spain. (e-
mail: calafate@disca.upv.es)
S. Dhelim is with School of Computer Science, University College
Dublin, Ireland. (e-mail: sahraoui.dhelim@ucd.ie)
A. Adnane is with the Computer Science department, Loughbor-
ough University, UK. (e-mail: A.Adnane@lboro.ac.uk)
Correspondence should be addressed to (ch.kerrache@lagh-
univ.dz)
available, it has become crucial to have practical service
selection mechanisms in place [1] [2] [3].
With the emergence and integration of the Internet of
Things (IoT), mobile networks underwent substantial
transformations, resulting in the growth of what are now
known as mobile Internet of Things networks. This con-
vergence has prepared the way for a wide range of ap-
plications, including smart homes and health monitoring,
industrial automation, and environmental sensing, which
benefit from mobile networks’ increased connection and
flexibility [4]. In the context of Mobile IoT, service se-
lection is the process of selecting the most appropriate
services from a pool of available options that best meet
the specific needs and preferences of users or applications
while taking into account various constraints and criteria,
such as quality of service (QoS), energy efficiency, security,
cost, and latency. This selection procedure is critical for
maximizing user experience and system performance in
mobile situations when resources are limited and condi-
tions can change rapidly. [3].
A significant challenge in service selection in mobile IoT
networks is determining the optimal services that fulfill
users’ needs, considering the diverse range of capabilities,
variations in latency, performance attributes, and the
extensive array of available services. These services span
from fundamental voice and messaging functions to data-
intensive applications like video streaming, Augmented
Reality (AR), online gaming, and IoT-based applications
such as smart homes, wearable devices, and industrial IoT
[5].
To address these challenges, Service selection in mobile
IoT networks has rapidly become an attractive topic
for researchers and industries by leveraging various
techniques, Decision-Making Algorithms [6], machine
learning-based algorithms [7], and game theory models
[8].
The introduction of fifth-generation 5G technology has
further revolutionized service selection in mobile networks
and played a crucial role in enabling wireless communica-
tion and connectivity for mobile devices, allowing people
to stay connected and access a wide range of services on
the go [9] [10].
On the other hand, The QoS requirements significantly
influence the service selection process. which involves
commonly used metrics in the service selection, such as
latency, bandwidth, reliability, and cost [11]. Different
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
2
algorithms and techniques have been applied to assist
users in selecting suitable Services, given that specific
features, limitations, and prices characterize every service.
[12]- [13].
Furthermore, cloud computing has gained significant
traction as a predominant computing paradigm, driven
by the proliferation of wireless network applications and
the apparent limitations in resources and storage capacity
for end users. This paradigm offers extensive computing
resources, continually evolving and enhancing the method-
ologies available to help consumers choose the most appro-
priate services [14] [15].
cloud
Edge computing
Base station
Internet of things
Fig. 1: service IoT Networks
this paper aims to offer a comprehensive overview of
service selection within mobile networks, focusing on the
employed strategies.
The rest of this paper is organized as follows. Section
II provides a brief background to understand the main
concepts of service selection in the mobile environment.
and Section III provides the Existing Service Selection
Works. Section IV presents the Proposed taxonomy for
service selection approaches in mobile networks. Service
selection solutions’ methods and techniques are presented
in Section V. Section VI discusses The main challenges
and future directions. Finally, we conclude in Section VII.
II. Background
We introduce in this section concepts and paradigms
essential for readers to comprehend the proposed topic.
Then, we proceed to review the main applications of
service selection. Table1 presents a comprehensive com-
pilation of acronyms utilized in this research article to
enhance the overall legibility and comprehension of the
study.
A. Useful concept definitions
•Service selection: Choosing the most suitable ser-
vices from a pool of candidates is called service se-
lection. This process relies on non-functional features,
commonly known as Quality of Service (QoS) metrics,
which include price, reliability, security, and reaction
TABLE 1: List of Acronyms
Acronym Definition
5G Fifth-Generation
AHP Analytic Hierarchy Process
ANP Analytic Network Process
API Application Programming Interface
BS Base Station
CC Cloud Computing
CV Consumer Vehicles
GA Genetic Algorithm
ILP Integer Linear Programming
IoT Internet of Things
IoV Internet of Vehicles
MANET Mobile Ad-hoc Network
MCDM Multi-Criteria Decision Making Methods
MEC Mobile Edge Computing
ML Machine Learning
PV Provider Vehicle
QoE Quality of Experience
QoS Quality of Service
REST REpresentational State Transfer
RFID Radio Frequency Identification
SAW Simple Additive Weighting
SLR Systematic Literature Review
TOPSIS Technique for Order of Preference by Sim-
ilarity to Ideal Solution
UAV Unmanned Aerial Vehicle
UAVNET Unmanned Aerial Vehicle ad hoc Network
VANET Vehicular Ad-hoc Network
WS Web Service
WSN Wireless Sensor Networks
time. The quality of service selection may fluctuate
over time due to its operation within a limited envi-
ronment. Therefore, the system for selecting services
must be flexible enough to accommodate consumer
preferences. Consequently, it is feasible to substitute
a service with an alternative that provides equivalent
features and comparable or improved Quality of Ser-
vice (QoS) during its lifespan [16].
•Service discovery: IoT service discovery and se-
lection entail the thorough examination of services
provided by IoT devices to pinpoint the most suit-
able options. This method takes into account the
limitations inherent in IoT systems as well as the
expectations of users, with the aim of improving the
quality of service (QoS) [17].
•Quality of Service (QoS): QoS: It is the ability
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
3
to achieve the desired level of performance regarding
a set of parameters, such as cost, bandwidth, latency,
delay, packet loss ratio, jitter, and throughput, among
others. The QoS places greater emphasis on the con-
siderations and criteria involved in the selection of
services. The QoS traits are recognized and employed
during service selection to enhance the discovery pro-
cess and choose the best appropriate services from a
pool of potential options [18].
•Internet of Things(IoT): The term IoT is semanti-
cally related to two words, “Internet” and “Things,”
where the Internet is known as the global system that
uses TCP/IP protocol suite to interconnect different
computer networks, while Things refer to any objects
that surround us and have the capability to sense
and collect data about its environment. Therefore,
IoT can be defined as a global system based on an
IP suite, in which objects equipped with sensors,
radio frequency identification (RFID) tags, or bar-
codes have a unique identity, operate in an innovative
environment, and are seamlessly integrated into the
information network by using intelligent interfaces
[19]. IoT relies on a wide range of materials, network
infrastructure, communication protocols, Internet ser-
vices, and computing technologies [20]
•Cloud Computing (CC): CC refers to a specific
approach for managing adjustable and scalable in-
frastructures. This involves utilizing remote servers to
define data centers, deployed centrally on a worldwide
scale [21]. According to Sunyaev et al. [22], the cloud
computing platform provides various on-demand ser-
vices categorized by usage tranches. These services
include processing power, storage, development envi-
ronments, bandwidth, and others.
•Edge Computing (EC): EC is a paradigm that
extends the capabilities of CC to edge nodes, such
as smart sensors, smartphones, smart vehicles, and
edge servers. This paradigm enables computation and
storage to be performed at the nearest level, typically
within the Base Station of cellular mobile networks
[23]. Edge devices connect to cloud data centers via
the core network only when advanced processing is
required [24]. Edge computing enhances the capabili-
ties of cloud computing by facilitating the execution of
computational tasks at the edge nodes. This approach
is particularly beneficial for real-time applications and
services that require low latency and high bandwidth.
By distributing computation and storage closer to
where data is generated and consumed, edge comput-
ing reduces network congestion and improves overall
system performance.
•Mobile Ad hoc NETwork (MANET): MANET
refers to a wireless network in which mobile devices
establish communication without relying on a fixed in-
frastructure or centralized resources [25]. The devices
can operate in either stationary or mobile configu-
rations, enabling the formation, reconfiguration, and
self-repair of networks in a dynamic or ad-hoc manner.
MANETs have the potential to provide users with ac-
cess to information and offer a more cost-effective al-
ternative compared to conventional networking tech-
nologies. They can be used in scenarios where fixed
network infrastructure is impractical, impaired, or
impossible [26]. Some advantages of MANETs over
fixed-topology networks include flexibility, scalability,
and cheaper management expenses [27].
•Vehicular Ad hoc NETwork (VANET): VANET
is a specific type of Mobile Ad-Hoc Network
(MANET) designed for communication between ve-
hicles in motion. It is established by employing the
fundamental concepts of MANETs within the context
of vehicular environments [28]. VANET is a decentral-
ized wireless network that allows various mobile cars
and connected devices to establish communication
and exchange valuable information with each other
[29].
•Unmanned Aerial Vehicle(UAV): a UAV refers
to an aircraft without a human pilot. Instead, it
is operated manually by an individual situated on
the ground or autonomously by utilizing a software
program [30]. UAV networks facilitate the intercon-
nection of many UAVs. According to Kerrache et al.
[31], these networks possess a wide range of potential
applications, including but not limited to military
surveillance, search and rescue missions, and various
other uses. In an ad-hoc UAV network, a subset of
UAVs is connected to the ground base stations, while
all UAVs collectively establish an ad-hoc network. In
this particular system, UAVs can establish communi-
cation channels with both other UAVs and the ground
base stations [1].
•Wireless Sensor Networks (WSNs): WSNs are
composed of numerous miniature devices that can
communicate with one another while operating under
constrained power resources. These wireless sensors
are strategically placed in a specific environment to
detect and monitor various environmental phenomena
[32]. Due to the limited power resources of sensor
nodes, the data collected from the designated envi-
ronment is transmitted directly to the base station.
B. Standards and technologies
Various protocols and standards in IoT mobile networks
enable devices to communicate seamlessly with the Inter-
net. Support critical public safety functions. These stan-
dards are critical to the expanding environment of mobile
networks, ensuring that devices communicate consistently,
securely, and efficiently in an increasingly connected world.
These standards include:
•NB-IoT Standardization:
NB-IoT, a fundamental technology for the Internet
of Things (IoT), was standardized in 3GPP Release
13 (LTE Advanced Pro). This standardization was an
important milestone, showcasing the 3GPP’s ability
to quickly react to emerging market needs. NB-IoT
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
4
is one of several technologies geared at addressing
the IoT industry, demonstrating 3GPP’s commitment
to providing a diverse portfolio of technologies for
operators to use based on their market requirements
(3GPP) [33] .
•Advances in 5G for IoT:
The evolution to 5G Advanced (5G Advanced) in-
volves important advancements with 3GPP versions
17 and 18, which focus on increasing radio access
network features such as MIMO enhancements and
dynamic spectrum sharing. These enhancements are
critical for enabling new IoT use cases because they
improve the network’s ability to manage high mo-
bility scenarios, increase robustness, and reduce user
equipment power consumption. Release 17 includes
substantial enhancements in beamforming, multiple-
input and multiple-output (MIMO) operations, user
equipment power reductions, and positioning. [34]
These improvements aim to raise network perfor-
mance and support ultra-reliable, low-latency commu-
nication (URLLC), which is critical for time-sensitive
IoT applications such as industrial automation and
remote-control applications.
•3GPP’s Role in IoT Standards:
3GPP has helped shape IoT standards by developing
a ”portfolio of technologies” that includes NB-IoT,
eMTC (also known as LTE Cat M1), and EC-GSM-
IoT. These technologies support a wide range of IoT
applications, including enhanced mobile broadband
(eMBB) and ultra-reliable and low-latency commu-
nications (URLLC), which are critical for the next
generation of IoT deployments. The standardization
efforts include enhancements spanning numerous re-
leases, with a focus on improving positioning, power
efficiency, and compatibility for new IoT applications.
With each release, 3GPP refines and expands IoT
capabilities, guaranteeing that mobile networks can
meet the diverse and changing needs of global IoT
deployments (3GPP).
Proximity Services (ProSe) over 3GPP networks is
vital because ProSe is a key standard enabling di-
rect connection between devices in LTE (Long Term
Evolution) and beyond. ProSe, as specified by 3GPP,
allows devices in close proximity to connect directly
with one another via the LTE network, skipping the
core network. This feature is notably useful for public
safety applications, as it allows first responders to
communicate in places where the network is over-
crowded or non-existent. [35] It also creates chances
for new commercial services and Internet of Things
applications that benefit from direct device-to-device
connectivity.
C. Applications relying on services selection
-The service selection process is even more challenging.
Selecting the best service that matches the end user’s
functionality and quality requirements is necessary. It
involves determining the most suitable service that
fulfills the user’s requirements and optimizing the
network resources to provide the desired level of quality
and performance within various contexts, such as
software development, cloud computing, service-oriented
architectures, Internet of Things (IoT), Healthcare
Systems, Network Management, E-commerce, Web
Services, etc. Fig.2 presents the Service Selection
Applications.
•Cloud Computing(CC):
Within cloud computing, it is common for companies
to encounter the need to make informed decisions
regarding selecting cloud services. These services in-
clude a range of offerings, including but not limited
to storage, computing power, and databases [36]. The
primary objective of such decision-making efforts is to
identify the most suitable cloud services that match
the organization’s specific goals, considering aspects
such as performance, cost, security, and other impor-
tant considerations. Service selection approaches as-
sist organizations in making well-informed judgments
regarding the choice of cloud service providers and the
selection of service plans [37].
•Internet of Things (IoT):
Service selection is essential in IoT systems where
various devices and sensors need to interact with
each other to provide many services such as cameras,
actuators, mobiles, homes, hospitals, etc., where this
Technique helps choose the most appropriate services
for data processing, storage, and analysis in IoT
applications [38].
•Healthcare Systems: In healthcare, service selec-
tion strategies can be employed to choose suitable
medical services to optimize patient care effectively.
These strategies assist healthcare clinicians and ad-
ministrators select the most appropriate services,
treatments, and patient interventions. Such tech-
niques include Health Information Systems, Emer-
gency Medical Services, Personalized Medicine, Medi-
cal Imaging and Diagnosis, Telemedicine Services, etc.
[39] [40].
•Web Services(WS):
Service selection is choosing the most suitable web ser-
vice from a pool of available services to fulfill a specific
activity or demand, known as service selection. Differ-
ent components of the services are evaluated during
this phase to ensure the services satisfy the customer’s
demands. These aspects consist of metrics for Quality
of Service (QoS), security, cost, etc. Based on the
factors above, various web services are evaluated and
compared during the selection process, and select the
best service for consumer demands [41] [42].
•Smart cities:
Smart cities use sophisticated technology to enhance
the overall well-being and standard of living of their
citizens. Smart cities use high-tech to improve resi-
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
5
dents’ lives. The challenge in this environment is find-
ing and selecting data and services in this architecture
to enhance citizen lives and nurture new innovative
companies. These services include transportation, en-
ergy and environment, health and social service, crime
and disaster prevention, education, demographics, lo-
gistics urban development and facility management,
culture, tourism, and city administration [43] [44].
•E-commerce:
E-commerce involves online purchases and sales. It
is available on computers, tablets, smartphones, and
other smart devices in various markets. E-commerce
sells books, plane tickets, etc. E-commerce platforms
use many services to manage product catalogs, pay-
ments, orders, customer support, and more. The ser-
vice selection process is when the consumer selects
the most suitable service from the available options.
Therefore, choosing the best service requires apply-
ing methods and techniques. This process creates a
comprehensive and competitive online platform [45].
Service selection
applications
Healthcare
IoT networks
Web services
E-commerce
Energy management
Telecom
Smart cities
FANET &VANET
networks
Cloud computing
Fig. 2: Service Selection Applications.
III. Existing Service Selection Works
This section reviews the existing service selection sur-
veys in various fields. In addition, we present contributions
and discuss the related work of Service selection in mobile
networks in the second subsection.
A. Existing service selection surveys
Within the scope of mobile network service selection,
the imperative to make informed decisions spans across
an array of diverse domains, underscoring its pervasive
significance in the realm of contemporary telecommunica-
tions. This multifaceted decision-making process is intrin-
sic to the effective utilization of mobile network services
and contributes significantly to the evolving landscape
of mobile communication technologies. A few reviews
have surveyed the service selection problem in different
environments; previous surveys related to this issue are
summarized below (See Table.2). In this subsection, we
discuss the existing literature on service selection. Authors
[46] provide a review of the literature on QoS-aware dy-
namic web service selection. It covers current QoS models
and related calculation methods, examines and evaluates
current service selection strategies and approaches, and
highlights several weaknesses of existing algorithms. The
work suggests several future study possibilities, such as
trusted service selection. This survey also notes the dif-
ficulties of developing web service applications and the
need for in-depth research. By using a new classification
method, the authors [47] surveyed, classified, and ana-
lyzed various service selection algorithms in the Internet
of Things (IoT) environment under Quality of Service
(QoS) constraints. The study identifies two main problems
that need to be addressed by the research community
to propose a suitable solution for the service selection
problem: (i)Identifying the methods to design an appro-
priate IoT environment and determining the methodology
for implementing the proposed solution. (ii) highlights
the importance of an evaluation step for assessing the
efficiency and effectiveness of the proposed solution. The
authors [16] propose a taxonomy to categorize and com-
pare different service discovery and selection techniques in
IoT. The authors discuss the main challenges associated
with service discovery and selection in IoT and how they
differ from traditional service management approaches. It
also explores the key factors that need to be considered
when evaluating the quality of service (QoS) and quality
of experience (QoE) for cloud services in IoT and how
these can be measured and optimized. The authors [48]
aim to Present a Systematic Literature Review (SLR)
and study the current IoT service selection approaches
where they Describe the main challenges of the IoT service
selection and Provide an exact evaluation of discussed
mechanisms using some important QoS metrics. Sun et
al. [14] provide a comprehensive survey of state-of-the-art
Cloud service selection approaches, analyzing them from
seven perspectives: context, purpose, data representation
models, selection techniques, selection parameters, meth-
ods for quantifying qualitative parameters, and criteria
weighting methods. The authors classified Cloud service
selection approaches into three categories: MCDM-based,
multi-criteria optimization-based, and logic-based. In [2],
the authors have highlighted the extensive use of multi-
criteria decision-making (MCDM) techniques in service
selection, specifically in determining the weight of quality
of service (QoS) factors and ranking services provided by
different service providers. The authors present a com-
prehensive classification of service selection schemes based
on Multiple Criteria Decision Making (MCDM) and dis-
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
6
cuss their adaptation and application in various contexts,
including web service selection, web service composition,
and cloud service selection. The primary objective is to
assist decision-makers in assessing the relative importance
of each Quality of Service (QoS) factor and the ranking
of services offered by different service providers. Another
work has presented an exhaustive survey of the optimal
selection and composition of cloud-based services [49]. It
begins with a summary of recent research in cloud man-
ufacturing, followed by a discussion of the SCOS (Service
Composition and Optimal Selection) issue. The authors
discuss the difficulties associated with the authenticity and
dependability of QoS values, the significance of user behav-
ior data, and the need to move towards automated and
integrated service composition. The survey also presents
recent advancements in six main areas of cloud manufac-
turing, including architecture, business modes, enabling
technologies, resource modeling and scheduling, service
virtualization, and task description. A survey for service
selection has been investigated in dynamic environments.
Manqele et al. [50] have analyzed various service selec-
tion approaches in dynamic environments and propose
a method for selecting the most effective one based on
certain factors. The proposed approach was tested by
manipulating the relevant services description of avail-
able services, and the method was evaluated based on
response time, recall, and precision metrics. The experi-
ments showed that the content-based algorithm returned
more relevant services to the user and took a shorter
time. The authors also discuss the QoS constraints that
must be considered when selecting a service, such as
reliability, response time, cost, throughput, integrity, and
platform/API. The authors [15] introduced a systematic
review of existing cloud service selection approaches and
offered a taxonomy based on a thorough literature study
according to eight dimensions: decision-making methods,
context, purposes, cloud service performance parameters,
simulation/language tools, domain, datasets, and valida-
tion methods. The authors identify nine crucial challenges
in selecting cloud services that require additional research.
Comparing the Analytic Hierarchy Process (AHP) and
the Analytic Network Process (ANP), the study concludes
that AHP is superior without feedback and dependencies.
In contrast, ANP is ideal for complex decision-making
involving interdependencies between criteria and alterna-
tives.
B. Service selection in mobile networks
This section discusses the existing literature on service
selection in mobile networks (See Table 3). It reviews
various approaches and techniques for improving the pro-
cess of service selection in mobile networks, such as game
theory models, machine learning, and fuzzy logic. The
strengths and limitations of each approach are evaluated,
providing insights into the evolution of service selection
mechanisms. In the UAV context, a new Game Theoretic
approach was proposed for Selection Services in UAV
Clouds (GTSSUC) [53] to enable normal users to select
the most suitable UAV-Service-Provider. The proposed
approach considers the user requirements and the qualities
of the UAV provider to find the most adequate provider.
The simulation advocates for the proposed method’s effi-
ciency in both network QoS and service gain.
In the same context, Brahimi et al. [54] investigated a
Game Theory approach for Cloud Services(GTCS)in MEC-
and UAV-enabled networks, enabling ordinary end-users
to select the most suitable UAV service provider based
on specific features, limitations, and prices. Hadjkouider
et al. [8] introduced a Stackelberg game-based solution
for service discovery and selection in the context of UAV-
based mobile edge computing. The proposed game is effi-
cient in terms of price and QoS metrics. It allows clients to
determine the most appropriate provider while efficiently
considering consumer preferences and service constraints.
Two Linear Integer Problem (LIP)-based optimization
techniques [55] are proposed for selecting the optimal set of
UAVs: Energy Aware Selection of UAVs (EAS) focuses on
minimizing energy consumption, while Delay Aware Selec-
tion of UAVs (DAS) aims to reduce response times. DAS is
particularly effective for optimizing operation time, while
EAS is geared towards minimizing energy consumption.
In the context of Vehicular Ad hoc Networks (VANETs),
Brik et al. [56] introduced a novel game theory-based
approach for managing service provisioning in vehicular
cloud computing. This approach takes into account the
benefits of each player and allows Consumer Vehicles
(CVs) to select the most suitable Provider Vehicle (PV)
based on their probability of interaction. CVs can choose
to either Consume (C) or Not Consume (NC), while PVs
can decide whether to Offer (O) or Not Offer (NO). How-
ever, this approach operates exclusively on a Vehicle-to-
Vehicle (V2V) architecture and does not consider Vehicle-
to-Infrastructure (V2I) interactions.
In [57], authors introduced a flexible approach for se-
lecting a vehicular cloud service provider based on a link
stability metric and linguistic quantifiers. The system effi-
ciently handles user preferences and reduces latency while
considering consumer preferences and service provider
features. The ranking was refined by defining two unique
operators: Least Satisfactory Proportion (LSP) and Most
Satisfactory Proportion (MSP). Additionally, the work
presented in [58] proposed a multi-agent/multi-objective
interaction game system to manage on-demand service
provision in a vehicular cloud based on a game theoretic
approach and a Quality of Experience (QoE) framework.
Three game approaches were developed to help drivers
minimize service costs and latency while maximizing their
privacy. Furthermore, the authors investigated the QoE
framework for service provision in a vehicular cloud for
various types of users.
Zhang et al. [59] investigated a Stackelberg game-based
approach to address the challenge of service pricing and
selection for IoT applications offloading in a multi-MEC
system. The proposed method comprises two stages: in
the first stage, the cloud service broker sets service prices
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
7
TABLE 2: Comparison of the reviewed works about service selection.
Ref. Year Domain Approach criteria
Han et al. [46] 2011 Web service Analyze and discuss the strategies of services selection
systematically
Abosaif et al. [47] 2020 Internet of Things (IoT) Classify and analyze algorithms for service selection un-
der QoS constraints.
Li et al. [48] 2020 Internet of Things (IoT) Centralized, decentralized, and hybrid taxonomy
Achir et al. [16] 2022 Internet of Things (IoT) Description methods, selection algorithms, and architec-
tural aspects
Sun et al. [14] 2014 Cloud services Analyze the five perspectives: decision-making tech-
niques, data, parameters representation models, and
characteristics of Cloud services.
Bouzary et al. [49] 2018 Cloud services Selection criteria, objective function, algorithms, map-
ping approaches, and correlation awareness
Manqele et al. [50] 2017 Dynamic environments Based on multi-agent approach, QoS, ontology, and func-
tional requirements
Mhakur et al. [15] 2022 Cloud services Eight dimensions considered: context, decision-making
methods, cloud service performance parameters, pur-
poses, simulation/language tools, datasets, domain, and
experimental methods
Hosseinzadeh et al. [2] 2020 Cloud service, web service MCDM-based service selection, classification service se-
lection techniques according to the applied MCDM meth-
ods
thakur et al. [51] 2023 Cloud services,Internet of Things (IoT) Description and investigation the Quality of Service
based Cloud Service Composition (QoSCSC) from the
perception of AI
Ghafouri et al. [52] 2022 Web Service Classify the QoS prediction methods to select best service
in web
Our work 2024 Internet of Things (IoT), Vanet,UAV Decision-making and description methods, categories for
service selection, provide a Taxonomy of service selection
approaches in mobile networks
and load-balancing strategies to maximize revenue, while
in the second stage, IoT users determine the number
of their generated application requests. The method of
backward induction is employed to solve the game, and
the effectiveness of the proposed scheme is validated
through simulation results. In the same context, Singh
et al. [60], analyzed the issue of IoT service selection,
considering the IoT framework as a composition of three
major components: things, communication, and comput-
ing. Subsequently, the quality-of-service (QoS) parameters
associated with each of these components are discussed.
Furthermore, the authors investigate a framework that
presents a hybrid strategy for selecting IoT-based services,
which combines the Analytic Hierarchy Process (AHP)
and Technique for Order of Preference by Similarity to
Ideal Solution (TOPSIS) approaches.
Process management in resource-constrained WSNs is a
crucial topic that warrants significant attention to intelli-
gently select services with optimal QoS values and residual
energy. In [61], authors proposed an energy-aware QoS-
guaranteed workflow management mechanism, a novel
QoS model, an efficient service selection schema, and an
adaptation mechanism for balanced energy usage in WSNs
using Linear Programming (LP). The study also addresses
the necessity for runtime modifications of component ser-
vices based on residual energy and explores the trade-off
between execution process changes and network lifetime.
C. Service discovery in mobile IoT networks
Service discovery in the Internet of Things (IoT) is crit-
ical for rapidly discovering and accessing services across
several IoT networks. Multiple approaches have been pre-
sented to improve the service discovery processes in IoT
contexts. Demir et al. [62] introduced a service discov-
ery technique called Peer-Assisted Multi-Service Discovery
(PAMSer) that is designed to be both highly scalable and
reliable. The system aims to identify and pick services that
fulfill the specified Quality of Service (QoS) criteria. This
algorithm utilizes a peer-to-peer (P2P) research discov-
ery technique to give real-time quality of service (QoS)
information for services. It eliminates the requirement
for periodic monitoring of services or flooding of search
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
8
query messages, as seen in current systems. Furthermore, a
selection optimization algorithm called Optimal Selection
(OptSel) is proposed to enhance the selection process.
The authors propose a deep reinforcement learning-based
approach for selecting and composing moving IoT services
[63]. The primary focus is on optimizing the composition
process through learning algorithms. The project seeks
to address key difficulties in IoT service composition by
focusing on improving accuracy, scalability, and efficiency.
The suggested method is evaluated using a detailed com-
parison to a parallel flock-based service discovery algo-
rithm, which provides useful insights into its accuracy and
effectiveness. Furthermore, the study contributes to the
formal modeling of crowdsourced moving services by em-
ploying a spatiotemporal model to ease service discovery
and composition, minimizing dependency on traditional
indexing processes. Through these efforts, the study hopes
to advance the state-of-the-art in IoT service composition,
opening up new options for future research and devel-
opment in this field. The study [64] proposes a latent-
based recommendation system optimized for the Social
Internet of Things (SIoT). It uses multi-modal features to
generate latent item graphs targeted at increasing service
recommendations. Extensive experimental evaluations are
carried out, using real-world datasets and performance
metrics like MAE, RMSE, Precision, Recall, and NDCG.
The results show that the suggested system outperforms
the baseline approaches in terms of accuracy and precision.
In addition, the experimental setup includes comparisons
with existing methodologies, a complete examination of
item graphs formed from multi-modal features, and sen-
sitivity analysis on important hyperparameters, which
provide extensive insights into the system’s performance
and robustness. In [65] authors present a configurable
distributed network service discovery system that em-
ploys stateless scanning technology and random destina-
tion address approaches to improve efficiency. This text
discusses the use of the operating system protocol stack to
thoroughly examine and establish connections, resulting
in more precise service discovery. Moreover, the paper
provides a collection of patterns that may be used to
convert user-customized service descriptions into standard
syntax. This helps to simplify the process of integrating
these descriptions into the scanning tool ecosystem. A
service discovery solution specifically designed for dis-
tributed embedded systems based on edge choreography
has been proposed [66]. This solution utilizes Raspberry Pi
machines at the IoT edge layer. The application assesses
the use of the CPU and the duration of delays in various
methods of exchanging messages, emphasizing the influ-
ence of a regular expression interpreter on the efficiency
of the system. This study presents illustrations of message
body structure for different delivery patterns, highlighting
the incorporation of new services with a limited under-
standing of meaning. The paper [67] introduces a web
service selection method specifically developed to facilitate
the worldwide optimization of Quality of Service (QoS)
and dynamic replanning. This algorithm utilizes location
matrix coding and integrates user feedback to improve
the precision of service selection. The study conducts a
thorough evaluation to determine the efficacy of several
QoS measurement methods and ranking algorithms. User
satisfaction metrics are used as the criteria for evaluation.
Simulation studies are performed to verify the effectiveness
of QoS-aware service discovery in satisfying consumers’
requirements. Moreover, the paper discusses the difficulties
related to harmonizing the comprehension of QoS indica-
tors among consumers and service providers, highlighting
the significance of this comprehension in enhancing service
quality and satisfaction. A specialized Graph Neural Net-
work (GNN) method designed for service discovery [68] is
presented in large-scale Social Internet of Things (SIoT)
networks, using the social connections between devices.
The suggested approach’s usefulness in facilitating fast
service discovery across SIoT networks is demonstrated
by simulated tests done on real-world datasets, evaluating
various embedding strategies and clustering methodolo-
gies. The authors present a new technique, as described
in [69], that focuses on clustering objects to enable the
grouping of entities that have similar services and capabil-
ities, with the goal of promoting reciprocal collaboration.
The suggested approach improves the efficiency of service
search by clustering objects based on shared criteria and
establishing relationships between service kinds and inter-
object collaboration. This methodology enhances both
the efficiency of service discovery and the effectiveness
of community identification inside the Social Internet of
Things (SIoT) ecosystem.
IV. Proposed taxonomy for service selection
approaches in mobile networks
The following section explains the proposed classifica-
tion approach for service selection in mobile networks, as
depicted in Fig. 3. Our taxonomy classifies the methods
into five categories based on the state-of-the-art research
approaches identified in our survey.
•The first category comprises mobile devices that func-
tion as service providers or sources. They may be
equipped with various sensors (e.g., cameras, thermal
sensors, LiDAR) to wirelessly collect and transmit
data in real-time, such as UAV, VANET, MANET,
and IoT applications.
•In the second category, architecture specifies the
structure, behavior, and perspectives necessary to
meet all technical and operational requirements while
optimizing common quality attributes such as per-
formance and security. It illustrates how processes
interact with one another and how network com-
ponents are distributed. The architecture structure
comprises centralized, decentralized, hierarchical, and
hybrid elements.
•The third category describes computing devices that
allocate computing resources, such as servers, stor-
age systems, and databases, to provide services, ap-
plications, and data storage. The properties above
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
9
TABLE 3: Comparison of the different works about service selection in mobile networks
Ref. Year Technology Technique used Attributes Dynamicity/Static (D/S) Environment Architecture
Hadjkouider [8] 2023 UAVs Stackelberg game Qos, price D MEC Decentralized
Bousbaa [53] 2021 UAVs Nash Equilibrium game Delay, cost D,S Cloud Computing Hyrarchical
Brahimi [54] 2023 UAVs Nash Equilibrium game Throughput,Delay, cost, D,S MEC Decentralized
energy
Motlagh [55] 2016 UAVs Nash Equilibrium game Delay,energy D,S —- Hyrarchical
Brik [56] 2018 VANETs Nash Equilibrium game QoS , costs D,S Cloud Computing Hybrid
Tamani [57] 2017 VANETs Fuzzy quantified, QoS, user preferences D Cloud Computing Decentralized
Linguistic quantifiers
Aloqaily [58] 2016 VANETs Nash Equilibrium Latency, cost , privacy D Cloud Computing Hyrarchical
Yang [38] 2020 IoT Algorithm QoS D —- Decentralized
Zhang [59] 2020 IoT Stackelberg game Price D,S MEC Decentralized
Demir [62] 2021 IoT Algorithm(OptSel,PAMSer) QoS , energy D —- Decentralized
Singh [60] 2020 IoT MCDM method Qos D —- Decentralized
Tong [61] 2017 WSN Linear Programming (LP) Qos,energy S —- Decentralized
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
10
encompass scalability and flexibility, such as cloud
computing, edge computing, and cloudlet.
•The metrics in The fourth category refer to the
specific criteria or parameters employed in assess-
ing and selecting various mobile services or applica-
tions. These indicators assist consumers and service
providers in making well-informed decisions regarding
selecting services that align with their requirements
and preferences, such as cost, security, and user pref-
erence.
•The fifth category identifies service selection methods
that encompass a range of techniques and tactics
employed to ascertain the most suitable service or
application for fulfilling a specified task or function.
Given the restricted resources and connectivity alter-
natives available on mobile devices, efficient service
selection is paramount to offer an excellent user expe-
rience. These methods include Game Theory, Machine
Learning, Fuzzy Logic, and MCDM (Multi-Criteria
Decision Making). fig3
V. Service selection solutions’ methods and
approach
A. Service selection’ methods
The techniques of Service selection methods in mobile
IoT networks, such as Game Theory, Machine Learn-
ing, Fuzzy Logic, and Multi-Criteria Decision Making
(MCDM), have been applied in the context of mobile
IoT networks. We discuss this issue below for detailed
experimental results and case studies.
•Game Theory: Game Theory is employed to rep-
resent interactions among different agents in a net-
work, where each actor strives to maximize its own
utility. Within the realm of mobile IoT networks,
these agents encompass devices, services, or users.
Game Theory enables the examination of strategic
choices in situations involving competition or co-
operation, especially in cases with scarce resources,
where gadgets or services vie for these restricted
resources [70]. Game-theoretic approaches frequently
result in equilibrium solutions, where no actor may
gain a benefit by unilaterally changing their strategy.
These equilibrium points ensure that resource allo-
cation and service selection are optimized fairly and
equally. Implementations sometimes involve the use of
auction-based models or cooperative gaming scenarios
to allocate resources and provide services efficiently.
These advanced techniques improve decision-making
processes in mobile IoT networks, guaranteeing eq-
uitable allocation of resources and optimizing overall
network performance.
•Machine Learning: Machine Learning: By employ-
ing machine learning methodologies, it becomes feasi-
ble to predict the most advantageous services for IoT
(Internet of Things) devices through the examination
of previous information [72]. This predictive method-
ology is extremely beneficial in adjusting to chang-
ing network circumstances and user needs. Training
machine learning models on historical data can lead
to high accuracy in predicting service requirements,
enabling efficient allocation of services through dy-
namic optimization. Different machine learning tech-
niques, like as neural networks, decision trees, and re-
inforcement learning, are used as centralized methods
for allocating services in real-time. These algorithms
take into account both network conditions and de-
vice behavior. In addition, techniques such as feder-
ated learning or distillation, which function based on
teacher-student paradigms, successfully tackle issues
related to scalability and flexibility [73]. These in-
novative methodologies enable systems to constantly
optimize the distribution of services, ensuring the
effective use of resources and enhancing overall system
performance.
•Fuzzy Logic: Fuzzy Logic deals with the process of
thinking that is characterized by being approximate
rather than strictly rigid and precise [74]. Fuzzy Logic
is highly beneficial in mobile IoT networks for making
decisions when faced with ambiguity and inaccurate
information, which is common owing to the dynamic
nature of these networks. Fuzzy Logic systems are
highly effective in managing ambiguity and uncer-
tainties that are naturally present in IoT situations.
They provide strong decision-making processes for
selecting services. Fuzzy Logic implementations may
use fuzzy rule-based systems that assess different
factors, such as battery life, network bandwidth,
and service latency, in order to find the most suit-
able service options [71]. By integrating Fuzzy Logic
into decision-making procedures, mobile IoT networks
may proficiently handle ambiguous circumstances,
guaranteeing adaptable and effective service selection
customized to the current network dynamics. Possible
implementations could involve the utilization of fuzzy
rule-based systems to select services depending on
factors like as battery life, network bandwidth, and
service latency.
•MCDM: Multi-criteria decision-making (MCDM)
approaches are essential for assessing and resolv-
ing conflicts between numerous criteria in decision-
making processes. In the context of mobile IoT net-
works, these criteria often include aspects like as
service quality, cost, energy consumption, and latency
[60]. MCDM methodologies, such as the Analytic
Hierarchy Process (AHP) or Technique for Order
Preference by Similarity to Ideal Solution (TOPSIS),
have demonstrated their efficacy in systematically
assessing and prioritizing many criteria to pick the
best appropriate services for IoT devices. MCDM is
utilized in the selection of IoT services for smart
cities, where many criteria such as service response
time, dependability, and energy efficiency need to be
carefully evaluated. By utilizing Multiple Criteria De-
cision Making (MCDM) procedures, decision-makers
may effectively navigate intricate decision landscapes,
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
11
Service selection approaches in mobile networks
Environment TypeArchitecture
Method Metric
Centralized
Decentralized
Hybrid
Hierarchical
Manet
UAV
IoT
Vanet
Cost
Security
Energy
User
preference
QoS
Requirements
Cloud
computing
Cloullet
Mobile edge
computing
Game theory
Machine
learning
Algorithm
MCDM
Fuzzy logic
Fig. 3: Taxonomy of service selection approaches in mobile networks
TABLE 5: Summary of Service Selection Methods in Mobile Networks
Method Goal Pros Cons Types
Multi-Criteria
Decision Making
(MCDM) [6]
To evaluate and prior-
itize services based on
multiple criteria.
– Allows for a system-
atic comparison of ser-
vices.
– Can consider a wide
range of criteria.
- Can be complex to
implement.
– Requires accurate
weighting of criteria.
– Analytic Hierarchy Process
(AHP)
– Technique for Order of Pref-
erence by Similarity to Ideal
Solution (TOPSIS)
– Multi-Attribute Utility The-
ory (MAUT)
Machine Learning
Algorithms [7]
To predict service per-
formance and make se-
lection decisions based
on data.
– Can improve with
more data over time.
– Adapts to changing
patterns in user behav-
ior and network condi-
tions.
– Requires significant
amounts of data for
training.
– May be complex
to understand and
explain.
– Neural Networks
– Decision Trees
– Reinforcement Learning
Game Theory [70] To analyze and predict
interactions between
decision-makers
for optimal service
selection.
– Provides insights into
competitive and coop-
erative scenarios.
– Can lead to equitable
resource distribution.
– Theoretical and may
not always reflect real-
world complexities.
– Requires comprehen-
sive understanding of
all players’ strategies.
– Cooperative games
– Non-cooperative games.
Fuzzy Logic [71] To make better
informed service
selection decisions
under conditions of
uncertainty, vagueness,
or imprecision
– Flexibility and intu-
itive decision making
optimization and con-
trol
– Complexity in imple-
mentation
– Computationally in-
tensive
– Takagi-Sugeno Fuzzy model
– Fuzzy Multi-Criteria Deci-
sion Making
Algorithms based To combine multiple
services to meet com-
plex user requirements
– Allows for flexible
and customized service
offerings.
– Can leverage the best
features of each service
– Composition can be
complex to orchestrate.
– May introduce la-
tency or compatibility
issues.
– Genetic Algorithms (GAs).
– Dynamic Programming.
– Greedy Algorithms.
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
12
taking into account a wide range of criteria and
making well-informed choices that are in line with
their intended objectives.
The existing methods for service selection solutions can
be categorized into two main groups, Fig.4:
•Optimization-based techniques:
In service selection, optimization-based techniques
aim to find the optimal allocation or assignment of
services to specific demands or requirements. The
objective is to optimize a particular performance
metric, such as minimizing cost, maximizing
efficiency, or maximizing resource utilization. This
involves employing techniques such as dynamic
programming, integer programming, greedy
algorithms, etc. The focus is on determining the
best configuration or combination of services that
satisfy the given constraints and optimize the defined
objective [55].
•Comparison-based techniques:
refers to a specific method of selecting services that
involves comparing different services to identify the
optimal combination that can effectively fulfill a given
set of needs. These techniques consider various factors
and criteria, such as cost, quality, reliability, customer
satisfaction, and other relevant criteria. The focus is
on identifying the best alternative(s) based on the
overall performance across multiple criteria [75]. One
of the most common comparison-based service selec-
tion approaches is Multiple Criteria Decision Making
(MCDM) [76], which includes methods such as An-
alytic Hierarchy Process/Analytic Network Process
(AHP/ANP), Technique for Order of Preference by
Similarity to Ideal Solution (TOPSIS), outranking,
Multi-Attribute Utility Theory (MAUT), and Simple
Additive Weighting (SAW), among others.
B. Service selection-based approach
•QoS-based service selection: Quality of Service
(QoS) involves implementing a mechanism that as-
signs varying priority levels to different applications,
users, or data flows. This mechanism also ensures a
specific performance for a given data flow. QoS-based
service selection entails selecting the most suitable
service by considering the user’s quality of service
(QoS) prerequisites and the range of services that
are accessible. This methodology is employed within
wireless networks to enhance the service rendered to
users. The suggested method entails the reception
of user requests and subsequent analysis of service
patterns by identifying comparable services, locations,
and the requisite data values. The primary objective
of this initiative is to optimize the quality of service
(QoS) advancements in the realm of mobile networks,
as stated in reference [77].
•Cost-based service selection: Cost-aware service
selection in mobile environments involves choosing
services from various providers based on cost consider-
ations. Wireless networks offer connectivity and data
services for both personal and corporate use, thus ne-
cessitating a balance between cost and quality. Cost-
aware service selection aims to find the most cost-
effective solution for users, taking into account factors
such as data usage patterns, coverage, dependability,
scalability, and value-added services. Additionally, it
requires knowledge of service providers’ pricing and
agreements. Conducting research, comparing services,
and analyzing cost-quality trade-offs within budget
constraints can help users select a reliable and afford-
able service [78].
•Energy consumption-based service selection:
The issue of limited battery capacity is becoming
more relevant in energy consumption within mo-
bile environments of service selection due to the
advancements in mobile networks [79]. Conventional
approaches endeavor to attain this goal by employing
a straightforward method of selecting services based
solely on their minimal reaction time. However, the
challenge is determining the most optimal service for
each activity to minimize the overall energy usage.
Using these strategies can effectively minimize the
reaction time of composite services, as demonstrated
in the study by Feeney et al. [80]. Consequently,
adopting these methods can ensure the lowest possible
standby energy consumption. In wireless networks, it
is imperative to minimize standby energy consump-
tion and optimize data transmission during favorable
network circumstances to mitigate energy consump-
tion associated with data transmission.
•Security-based service selection: In the current
digital landscape, security-based services have
emerged as pivotal components, prioritizing the
safety and protection of various entities, particularly
individuals. This paradigm emphasizes selecting
services that bolster essential security elements
[81], including authentication, trust, privacy, object
access control, and data integrity. The selection
process aims to ensure that chosen components, be
they services or objects, align seamlessly with the
security requirements defined by users and ongoing
programs, thereby guaranteeing an optimal level of
responsiveness [82]. A viable solution to enhancing
security in service selection involves integrating
pseudonym techniques [83]. This strategy offers a
promising approach to address privacy concerns in the
digital age. By leveraging pseudonyms, personalized
services can be delivered without compromising the
imperative of privacy preservation. This integration
heralds a future where privacy and service quality
can coexist harmoniously. The following case studies
serve as illustrations [84]:
– Online retail: Within the domain of e-commerce,
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
13
Optimization-based
Comparison based
Fuzzy Logic
Genetic Algorithms(GA)
Simple Additive Weighting (SAW)
Analytic Network Process (ANP)
Analytic Hierarchy Process (AHP)
TOPSIS
Game Theory
Integer Linear Programming (ILP)
Machine Learning
Category of service selection
technique
Fig. 4: Service selection technique tree
online retailers are implementing pseudonym
strategies to protect customer purchase records.
Through the process of anonymizing transactional
data, these systems guarantee the confidentiality and
protection of sensitive consumer information
– Healthcare services: Healthcare providers are using
pseudonyms for patient records, which allows for
discreet treatment and medical care delivery while
maintaining strict privacy regulations.
Despite the effectiveness of pseudonym techniques in
strengthening privacy, it’s crucial to recognize their
inherent vulnerabilities. Potential risks encompass
linkability, inference attacks, data breaches, and ad-
ditional cybersecurity threats like phishing attacks
[85], all of which present substantial challenges in
service selection. This necessitates vigilance against
such malicious activities and the implementation of
robust security measures to safeguard sensitive data
and uphold service integrity.
•Combined metrics based service selection: This
particular set of methodologies aims to identify the
most appropriate services and objects by considering
a range of characteristics [2]. The selection method
is commonly employed to integrate services or prod-
ucts to construct a more intricate framework that
enhances capabilities and overall performance. The
various metrics under consideration encompass all
network performance characteristics, including en-
ergy consumption, security, dependability, through-
put, and other relevant factors. Therefore, the se-
lection method requires establishing a harmonious
balance between these specified metrics, which may
occasionally conflict with each other, to determine
the most advantageous services from the array of
choices available. Regularly assessing metrics for each
selectable element and maintaining their updates are
crucial to ensure the utmost accuracy of outcomes
throughout the selection process [86].
VI. Challenges and future research directions
The service selection problem has been extensively ex-
plored by researchers, resulting in the proposal of several
study approaches. However, it is imperative to acknowl-
edge that specific unresolved challenges within academic
and industrial domains require attention and resolution
in the forthcoming period. This section focuses on the
challenges and future research directions related to pro-
moting and developing competent and efficient service
selection solutions. Figure 5 presents the challenges and
future research directions.
A. Challenges
•Battery life and energy efficiency: battery capacity is
a constraint for mobile devices. Extending battery life
and enhancing energy efficiency continue to be signif-
icant obstacles. Discovering and selecting appropriate
services from the geographically closest cloud data
centers can reduce energy consumption in a mobile
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
14
environment. Additionally, addressing this challenge
involves optimizing service selection to utilize fewer
device resources.
•Quality of Service (QoS): QoS presents several signifi-
cant challenges. It is crucial to ensure efficient, reliable
communication and provide users with a satisfactory
and consistent level of service, especially in scenarios
where there may be variations in network conditions,
service providers, and user requirements.
•Security and privacy: security and privacy measures
for service selection in mobile environments and net-
works are essential to safeguard the integrity and
protection of data. It is a major concern for cus-
tomers to ensure confidentiality, authentication, data
integrity, and privacy when they require any service
from a service provider. Several research efforts have
been conducted to address security and confidentiality
concerns and ensure secure service delivery.
•Scalability and resource management: It can be de-
fined as managing the growing number of mobile de-
vices and efficiently allocating resources. The system
must be capable of accommodating new devices, ser-
vices, or operations and handling increasing requests
without compromising the quality and performance
of existing services. Therefore, scalability refers to
the ability to manage a large volume of connected
devices in a mobile environment, which is considered
a challenge for the network.
Challenges Future
directions
Battery
life
Blockchain
technology
Scalability
Quality
of
Service Edge AI
Self
quantify
Security
& privacy
5G and
beyond
Fig. 5: Challenges and future research directions.
B. Future research directions
As the technological landscape evolves, several key do-
mains emerge as potential focal points for academic and
industry-based exploration:
•5G and beyond: selecting a service provider and plan
in the context of 5G (fifth generation) wireless tech-
nology involves considering various factors to ensure
you get the best value and performance, and with
increased connectivity compared to previous genera-
tions, which required offering faster data speeds and
lower latency to minimize a gap between the possible
transformation in terms of performance and efficiency
and the emerging new service requirements.
•Edge AI: it revolutionizes service selection by imbuing
it with intelligence, dynamism, and responsiveness
to evolving conditions and user preferences. This in-
tegration enables the system to effectively optimize
and select the most suitable service, while also ana-
lyzing and comparing the Quality of Service (QoS)
offered by diverse providers. This capability proves
especially valuable in scenarios necessitating real-time
decision-making. Furthermore, Edge AI facilitates the
customization of services based on preferences and
content availability across multiple platforms, thereby
enhancing the overall user experience.
•Self quantity: Self-quantification on mobile devices
and apps can significantly influence personal data
analytics and self-improvement, including service se-
lection. It enhances decision-making and overall well-
being through data collection and tracking, health
and fitness tracking, data privacy and security, afford-
ability, and guaranteed service availability.
•Blockchain technology: it is a chain of blocks in which
transaction information is recorded and maintained
in a distributed public ledger across several peer-to-
peer-connected virtual computers. The selection of
a provider is typically based on the values provided
and customer feedback. These services are offered to
various organizations and utilized by multiple con-
sumers. This underscores the need for a trustworthy
environment to evaluate the parametric values for
service selection using blockchain technology.
VII. Conclusion
As a promising and dynamic field, service selection
in mobile networks has garnered significant attention
across various domains such as healthcare, education,
e-commerce, transportation, etc., aiming to select the
most suitable services for users based on multiple factors.
However, the mobility of networks introduces a range of
challenges that must be addressed to make the service
selection process more seamless and efficient.
This paper reviews proposed solutions that leverage
mobile networks for service selection, employing differ-
ent techniques and application services while safeguard-
ing privacy using pseudonym techniques. We categorize
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
15
these solutions based on the metrics used, including QoS-
based selection, cost-based, security-based, etc., and clas-
sify them into two main groups: optimization-based and
comparison-based techniques. Through this classification,
we identify the integration of some new technologies for
service selection to better serve users’ needs, such as deep
learning and game theory, along with addressing recent
research advancements. Finally, we highlight the main
challenges and future research directions in this field,
aiming to inspire further studies and engage researchers
interested in advancing this area of research.
Acknowledgements
This work has been partially funded by R&D
projects TED2021-130890B-C22 and PID2021-122580NB-
100, from MCIN/AEI/10.13039/501100011033, by Euro-
pean Union NextGenerationEU/PRTR, and ”ERDF A
way of making Europe”.
References
[1] A. I. Hentati and L. C. Fourati, “Comprehensive survey of uavs
communication networks,” Computer Standards & Interfaces,
vol. 72, p. 103451, 2020.
[2] M. Hosseinzadeh, H. K. Hama, M. Y. Ghafour, M. Masdari,
O. H. Ahmed, and H. Khezri, “Service selection using multi-
criteria decision making: a comprehensive overview,” Journal
of Network and Systems Management, vol. 28, pp. 1639–1693,
2020.
[3] L. Huang and S. Deng, “Service selection for mobile service or-
chestration,” in 2014 IEEE International Conference on Mobile
Services. IEEE, 2014, pp. 147–148.
[4] S. Zaman, K. Alhazmi, M. A. Aseeri, M. R. Ahmed, R. T. Khan,
M. S. Kaiser, and M. Mahmud, “Security threats and artificial
intelligence based countermeasures for internet of things net-
works: a comprehensive survey,” Ieee Access, vol. 9, pp. 94668–
94 690, 2021.
[5] H. K. Apat, R. Nayak, and B. Sahoo, “A comprehensive review
on internet of things application placement in fog computing
environment,” Internet of Things, p. 100866, 2023.
[6] P. K. D. Pramanik, S. Biswas, S. Pal, D. Marinkovi´c,
and P. Choudhury, “A comparative analysis of multi-criteria
decision-making methods for resource selection in mobile crowd
computing,” Symmetry, vol. 13, no. 9, p. 1713, 2021.
[7] K. Baek and I.-Y. Ko, “Spatio-cohesive service selection
using machine learning in dynamic iot environments,” in
Web Engineering: 18th International Conference, ICWE 2018,
C´aceres, Spain, June 5-8, 2018, Proceedings 18. Springer, 2018,
pp. 366–374.
[8] A. M. Hadjkouider, C. A. Kerrache, A. Korichi, Y. Sahraoui, and
C. T. Calafate, “Stackelberg game approach for service selection
in uav networks,” Sensors, vol. 23, no. 9, p. 4220, 2023.
[9] D. H. Devi, K. Duraisamy, A. Armghan, M. Alsharari,
K. Aliqab, V. Sorathiya, S. Das, and N. Rashid, “5g technology
in healthcare and wearable devices: A review,” Sensors, vol. 23,
no. 5, p. 2519, 2023.
[10] H. Yu, T. Taleb, and J. Zhang, “Deterministic latency/jitter-
aware service function chaining over beyond 5g edge fab-
ric,” IEEE Transactions on Network and Service Management,
vol. 19, no. 3, pp. 2148–2162, 2022.
[11] S. Deng, H. Wu, D. Hu, and J. L. Zhao, “Service selection
for composition with qos correlations,” IEEE Transactions on
Services Computing, vol. 9, no. 2, pp. 291–303, 2014.
[12] Z. Ding, J. Liu, Y. Sun, C. Jiang, and M. Zhou, “A transaction
and qos-aware service selection approach based on genetic algo-
rithm,” IEEE Transactions on Systems, Man, and Cybernetics:
Systems, vol. 45, no. 7, pp. 1035–1046, 2015.
[13] H. Ma, A. Wang, and M. Zhang, “A hybrid approach
using genetic programming and greedy search for qos-
aware web service composition,” Transactions on large-scale
data-and knowledge-centered systems XVIII: special issue on
database-and expert-systems applications, pp. 180–205, 2015.
[14] L. Sun, H. Dong, F. K. Hussain, O. K. Hussain, and E. Chang,
“Cloud service selection: State-of-the-art and future research
directions,” Journal of Network and Computer Applications,
vol. 45, pp. 134–150, 2014.
[15] N. Thakur, A. Singh, and A. Sangal, “Cloud services selection:
A systematic review and future research directions,” Computer
Science Review, vol. 46, p. 100514, 2022.
[16] M. Achir, A. Abdelli, L. Mokdad, and J. Benothman, “Service
discovery and selection in iot: A survey and a taxonomy,”
Journal of Network and Computer Applications, vol. 200, p.
103331, 2022.
[17] X. Fang, “Semantic clustering analysis for web service discovery
and recognition in internet of things,” Soft Computing, vol. 27,
no. 3, pp. 1751–1761, 2023.
[18] Y. Dongre and R. Ingle, “An investigation of qos crite-
ria for optimal services selection in composition,” in 2020
2nd International Conference on Innovative Mechanisms for
Industry Applications (ICIMIA). IEEE, 2020, pp. 705–710.
[19] F. Ahmad, Z. Ahmad, C. A. Kerrache, F. Kurugollu, A. Adnane,
and E. Barka, “Blockchain in internet-of-things: Architecture,
applications and research directions,” in 2019 International
conference on computer and information sciences (ICCIS).
IEEE, 2019, pp. 1–6.
[20] G. Ruggeri, V. Loscr´ı, M. Amadeo, and C. T. Calafate, “The
internet of things for smart environments,” p. 51, 2020.
[21] J. Li and S. Zhu, “Service composition considering energy
consumption of users and transferring files in a multicloud
environment,” Journal of Cloud Computing, vol. 12, no. 1, pp.
1–12, 2023.
[22] A. Sunyaev and A. Sunyaev, “Cloud computing,” Internet
Computing: Principles of Distributed Systems and Emerging
Internet-Based Technologies, pp. 195–236, 2020.
[23] H. Wu, S. Deng, W. Li, J. Yin, X. Li, Z. Feng, and A. Y. Zomaya,
“Mobility-aware service selection in mobile edge computing sys-
tems,” in 2019 IEEE International Conference on Web Services
(ICWS). IEEE, 2019, pp. 201–208.
[24] G. Zou, Z. Qin, S. Deng, K.-C. Li, Y. Gan, and B. Zhang,
“Towards the optimality of service instance selection in mo-
bile edge computing,” Knowledge-Based Systems, vol. 217, p.
106831, 2021.
[25] C. A. Kerrache, C. T. Calafate, N. Lagraa, J.-C. Cano, and
P. Manzoni, “Rita: Risk-aware trust-based architecture for col-
laborative multi-hop vehicular communications,” Security and
Communication Networks, vol. 9, no. 17, pp. 4428–4442, 2016.
[26] A. M. Abdullah, “Mobile ad hoc networks: A survey of existing
mobility models and routing protocols,” 2021.
[27] S. G. Datey and T. Ansari, “Mobile ad-hoc networks its advan-
tages and challenges,” International Journal of Electrical and
Electronics Research, vol. 3, no. 2, pp. 491–496, 2015.
[28] Z. Sun, Y. Liu, J. Wang, G. Li, C. Anil, K. Li, X. Guo,
G. Sun, D. Tian, and D. Cao, “Applications of game theory in
vehicular networks: A survey,” IEEE Communications Surveys
& Tutorials, vol. 23, no. 4, pp. 2660–2710, 2021.
[29] H. Amirat, N. Lagraa, C. A. Kerrach, and Y. Ouinten,
“Fuzzy clustering for misbehaviour detection in vanet,” in 2018
International Conference on Smart Communications in Network
Technologies (SaCoNeT). IEEE, 2018, pp. 200–204.
[30] S. Djeradi, T. Bendouma, N. Lagraa, C. A. Kerrache, A. Lakas,
and B. Brik, “Toward the integration of uavs’ services into
the cloud,” IEEE Communications Standards Magazine, vol. 5,
no. 4, pp. 25–32, 2021.
[31] C. A. Kerrache, C. Calafate, A. Lakas, and M. Lahby, Internet
of Unmanned Things (IoUT) and Mission-based Networking.
Springer Nature, 2023.
[32] M. Elshrkawey, S. M. Elsherif, and M. E. Wahed, “An enhance-
ment approach for reducing the energy consumption in wireless
sensor networks,” Journal of King Saud University-Computer
and Information Sciences, vol. 30, no. 2, pp. 259–267, 2018.
[33] K. Muteba, K. Djouani, and T. Olwal, “5g nb-iot: Design,
considerations, solutions and challenges,” Procedia Computer
Science, vol. 198, pp. 86–93, 2022.
[34] K. Shafique, B. A. Khawaja, F. Sabir, S. Qazi, and M. Mus-
taqim, “Internet of things (iot) for next-generation smart
systems: A review of current challenges, future trends and
prospects for emerging 5g-iot scenarios,” Ieee Access, vol. 8, pp.
23 022–23 040, 2020.
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
16
[35] U. N. Kar and D. K. Sanyal, “A critical review of 3gpp standard-
ization of device-to-device communication in cellular networks,”
SN Computer Science, vol. 1, no. 1, p. 37, 2020.
[36] L. Wang, Q. He, D. Gao, J. Wan, and Y. Zhang, “Temporal-
perturbation aware reliability sensitivity measurement for adap-
tive cloud service selection,” IEEE Transactions on Services
Computing, vol. 15, no. 4, pp. 2301–2313, 2020.
[37] M. Rahimi, N. Jafari Navimipour, M. Hosseinzadeh, M. H.
Moattar, and A. Darwesh, “Toward the efficient service selection
approaches in cloud computing,” Kybernetes, vol. 51, no. 4, pp.
1388–1412, 2022.
[38] W. Yang and F. Deng, “A service selection method based on qos
in iot,” in 2020 5th International Conference on Computer and
Communication Systems (ICCCS). IEEE, 2020, pp. 791–795.
[39] A. Lakhan, M. A. Dootio, A. H. Sodhro, S. Pirbhulal, T.-M.
Grønli, M. S. Khokhar, and L. Wang, “Cost-efficient service se-
lection and execution and blockchain-enabled serverless network
for internet of medical things,” 2021.
[40] A. H. Sodhro, F. K. Shaikh, S. Pirbhulal, M. M. Lodro, and
M. A. Shah, “Medical-qos based telemedicine service selec-
tion using analytic hierarchy process,” Handbook of large-scale
distributed computing in smart healthcare, pp. 589–609, 2017.
[41] M. Daaji, A. Ouni, M. M. Gammoudi, S. Bouktif, and M. W.
Mkaouer, “Multi-criteria web services selection: Balancing the
quality of design and quality of service,” ACM Transactions on
Internet Technology (TOIT), vol. 22, no. 1, pp. 1–31, 2021.
[42] W. Serrai, A. Abdelli, L. Mokdad, and Y. Hammal, “Towards an
efficient and a more accurate web service selection using mcdm
methods,” Journal of computational science, vol. 22, pp. 253–
267, 2017.
[43] P. Sirohi, F. N. Al-Wesabi, H. M. Alshahrani, P. Maheshwari,
A. Agarwal, B. K. Dewangan, A. M. Hilal, and T. Choudhury,
“Energy-efficient cloud service selection and recommendation
based on qos for sustainable smart cities,” Applied Sciences,
vol. 11, no. 20, p. 9394, 2021.
[44] T. K. Soltvedt, “A distributed-to-centralized cost model for
service selection in smart cities,” Master’s thesis, NTNU, 2019.
[45] K. Cao, Y. Xu, Q. Wu, J. Wang, and C. Liu, “Optimal channel
and logistics service selection strategies in the e-commerce con-
text,” Electronic Commerce Research and Applications, vol. 48,
p. 101070, 2021.
[46] X. Han, Y. Liu, B. Xu, and G. Zhang, “A survey on qos-
aware dynamic web service selection,” in 2011 7th International
Conference on Wireless Communications, Networking and
Mobile Computing. IEEE, 2011, pp. 1–5.
[47] A. N. Abosaif and H. S. Hamza, “Quality of service-aware ser-
vice selection algorithms for the internet of things environment:
A review paper,” Array, vol. 8, p. 100041, 2020.
[48] Y. Li, Y. Huang, M. Zhang, and L. Rajabion, “Service selection
mechanisms in the internet of things (iot): a systematic and
comprehensive study,” Cluster Computing, vol. 23, pp. 1163–
1183, 2020.
[49] H. Bouzary and F. Frank Chen, “Service optimal selection
and composition in cloud manufacturing: a comprehensive sur-
vey,” The International Journal of Advanced Manufacturing
Technology, vol. 97, pp. 795–808, 2018.
[50] L. Manqele, M. Dlodlo, L. Coetzee, and G. Sibiya, “A survey for
service selection approaches in dynamic environments,” in 2017
IEEE AFRICON. IEEE, 2017, pp. 1049–1054.
[51] V. Swathi, G. S. Kumar, and A. V. Vathsala, “Cloud service
selection system approach based on qos model: A systematic
review,” International Journal on Recent and Innovation Trends
in Computing and Communication, vol. 11, no. 2, pp. 05–13,
2023.
[52] S. H. Ghafouri, S. M. Hashemi, and P. C. Hung, “A survey
on web service qos prediction methods,” IEEE Transactions on
Services Computing, vol. 15, no. 04, pp. 2439–2454, 2022.
[53] F. Z. Bousbaa, A. Lakas, A. E. Rezigat, H. S. Benguettache,
N. Lagraa, C. A. Kerrache, and A. E. K. Tahari, “Gtss-uc: a
game theoretic approach for services’ selection in uav clouds,” in
2021 IEEE/ACM 25th International Symposium on Distributed
Simulation and Real Time Applications (DS-RT). IEEE, 2021,
pp. 1–8.
[54] S. Y. Brahimi, F. Mouffak, F. Z. Bousbaa, C. A. Kerrache,
N. Lagraa, and A. Lakas, “Cloud service selection in ioft-
enabled multi-access edge computing: a game theoretic ap-
proach,” Annals of Telecommunications, pp. 1–12, 2023.
[55] N. H. Motlagh, M. Bagaa, and T. Taleb, “Uav selection for
a uav-based integrative iot platform,” in 2016 IEEE Global
Communications Conference (GLOBECOM). IEEE, 2016, pp.
1–6.
[56] B. Brik, J. A. Khan, Y. Ghamri-Doudane, N. Lagraa, and
A. Lakas, “Gss-vc: A game-theoretic approach for service selec-
tion in vehicular cloud,” in 2018 15th IEEE Annual Consumer
Communications & Networking Conference (CCNC). IEEE,
2018, pp. 1–6.
[57] N. Tamani, B. Brik, N. Lagraa, and Y. Ghamri-Doudane, “Ve-
hicular cloud service provider selection: A flexible approach,”
in GLOBECOM 2017-2017 IEEE Global Communications
Conference. IEEE, 2017, pp. 1–6.
[58] M. Aloqaily, B. Kantarci, and H. T. Mouftah, “Multia-
gent/multiobjective interaction game system for service provi-
sioning in vehicular cloud,” IEEE Access, vol. 4, pp. 3153–3168,
2016.
[59] W. Zhang, X. Li, L. Zhao, X. Yang, T. Liu, and W. Yang,
“Service pricing and selection for iot applications offloading in
the multi-mobile edge computing systems,” IEEE Access, vol. 8,
pp. 153 862–153 871, 2020.
[60] M. Singh, G. Baranwal, and A. K. Tripathi, “Qos-aware se-
lection of iot-based service,” Arabian Journal for Science and
Engineering, vol. 45, no. 12, pp. 10 033–10 050, 2020.
[61] E. Tong, L. Chen, and H. Li, “Energy-aware service selection
and adaptation in wireless sensor networks with qos guarantee,”
IEEE Transactions on Services Computing, vol. 13, no. 5, pp.
829–842, 2017.
[62] K. Demir, “A qos-aware service discovery and selection mech-
anism for iot environments,” S¯adhan¯a, vol. 46, no. 4, p. 242,
2021.
[63] A. G. Neiat, A. Bouguettaya, and M. Bahutair, “A deep rein-
forcement learning approach for composing moving iot services,”
IEEE Transactions on Services Computing, vol. 15, no. 5, pp.
2538–2550, 2021.
[64] A. Khelloufi, H. Ning, A. Naouri, A. B. Sada, A. Qammar,
A. Khalil, L. Mao, and S. Dhelim, “A multimodal latent-
features-based service recommendation system for the social
internet of things,” IEEE Transactions on Computational Social
Systems, 2024.
[65] X. Yu, Z. Hu, and Y. Xin, “A new approach customiz-
able distributed network service discovery system,” Wireless
Communications and Mobile Computing, vol. 2021, pp. 1–10,
2021.
[66] S. Blanc, J.-L. Bayo-Mont´on, S. Palanca-Barrio, and N. X.
Arreaga-Alvarado, “A service discovery solution for edge
choreography-based distributed embedded systems,” Sensors,
vol. 21, no. 2, p. 672, 2021.
[67] N. Zhang, “Service discovery and selection based on dynamic
qos in the internet of things,” Complexity, vol. 2021, pp. 1–12,
2021.
[68] A. Hamrouni, H. Ghazzai, and Y. Massoud, “Service discovery
in social internet of things using graph neural networks,” in 2022
IEEE 65th International Midwest Symposium on Circuits and
Systems (MWSCAS). IEEE, 2022, pp. 1–4.
[69] M. Malekshahi Rad, A. M. Rahmani, A. Sahafi, and N. N.
Qader, “Community detection and service discovery on social
internet of things,” International Journal of Communication
Systems, vol. 36, no. 11, p. e5501, 2023.
[70] C. Chi, Y. Wang, X. Tong, M. Siddula, and Z. Cai, “Game
theory in internet of things: A survey,” IEEE Internet of Things
Journal, vol. 9, no. 14, pp. 12 125–12 146, 2021.
[71] Z. Du, C. Wu, T. Yoshinage, L. Zhong, and Y. Ji, “On-device
federated learning with fuzzy logic based client selection,” in
Proceedings of the Conference on Research in Adaptive and
Convergent Systems, 2022, pp. 64–70.
[72] M. A. Alsheikh, D. T. Hoang, D. Niyato, D. Leong, P. Wang,
and Z. Han, “Optimal pricing of internet of things: A ma-
chine learning approach,” IEEE Journal on Selected Areas in
Communications, vol. 38, no. 4, pp. 669–684, 2020.
[73] Y. Sahraoui, C. A. Kerrache, C. T. Calafate, and P. Man-
zoni, “Fedrx: Federated distillation-based solution for prevent-
ing hospitals overcrowding during seasonal diseases using mec,”
in 2024 IEEE 21st Consumer Communications & Networking
Conference (CCNC). IEEE, 2024, pp. 558–561.
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
17
[74] S. Singarapu, D. Siri, and D. V. Nemova, “Study of a fuzzy
logic-based approach to supplier selection,” in MATEC Web of
Conferences, vol. 392. EDP Sciences, 2024, p. 01018.
[75] P. Bagga, R. Hans, and A. Joshi, “Qos based web service
selection and multi-criteria decision making methods,” 2019.
[76] R. K. Gavade, “Multi-criteria decision making: An overview
of different selection problems and methods,” International
Journal of Computer Science and Information Technologies,
vol. 5, no. 4, pp. 5643–5646, 2014.
[77] S. Rajasekar and C. Palanisamy, “Optimized location-based
service selection for qos improvement in wireless networks,”
Computers & Electrical Engineering, vol. 64, pp. 277–287, 2017.
[78] W. Li, H. Faragardi, M. Ozger, C. Cavdar, and B. Skubic, “Cost
aware service selection in a mobile edge marketplace,” Computer
Networks, vol. 205, p. 108680, 2022.
[79] S. Deng, H. Wu, W. Tan, Z. Xiang, and Z. Wu, “Mobile service
selection for composition: An energy consumption perspective,”
IEEE Transactions on Automation Science and Engineering,
vol. 14, no. 3, pp. 1478–1490, 2015.
[80] L. M. Feeney and M. Nilsson, “Investigating the energy con-
sumption of a wireless network interface in an ad hoc net-
working environment,” in Proceedings IEEE INFOCOM 2001.
Conference on computer communications. Twentieth annual
joint conference of the IEEE computer and communications
society (Cat. No. 01CH37213), vol. 3. IEEE, 2001, pp. 1548–
1557.
[81] M. S. Abdalzaher and O. Muta, “A game-theoretic approach
for enhancing security and data trustworthiness in iot appli-
cations,” IEEE Internet of Things Journal, vol. 7, no. 11, pp.
11 250–11 261, 2020.
[82] C. Muralidharan and R. Anitha, “Edsac–an efficient dempster
shafer algorithm for classification to estimate the service, se-
curity and privacy risks with the service providers,” Wireless
Personal Communications, pp. 1–21, 2022.
[83] A. P. Mdee, M. T. R. Khan, J. Seo, and D. Kim, “Security
compliant and cooperative pseudonyms swapping for location
privacy preservation in vanets,” IEEE Transactions on Vehicular
Technology, 2023.
[84] M. Ghafoorian and M. Nikooghadam, “An anonymous and
secure key agreement protocol for nfc applications using
pseudonym,” Wireless Networks, vol. 26, no. 6, pp. 4269–4285,
2020.
[85] Y. Sahraoui, C. A. Kerrache, A. Korichi, A. M. Vegni, and
M. Amadeo, “Learnphi: A real-time learning model for early
prediction of phishing attacks in iov,” in 2022 IEEE 19th Annual
Consumer Communications & Networking Conference (CCNC).
IEEE, 2022, pp. 252–255.
[86] M. G. A. Malik, S. K. Ray, Z. Bashir, and A. Mughal, “Se-
lecting ubiquitous services in future heterogenous wireless net-
works using multi-attributes decision making,” in 2018 Eleventh
International Conference on Mobile Computing and Ubiquitous
Network (ICMU). IEEE, 2018, pp. 1–4.
This article has been accepted for publication in IEEE Open Journal of the Communications Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJCOMS.2024.3400981
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
