This article has been accepted for publication in IEEE Access.

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/ACCESS.2022.3229185

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2017.DOI

Cloud Continuum: The Definition

SERGIO MORESCHINI1, FABIANO PECORELLI1, XIAOZHOU LI1, SONIA NAZ1,
DAVID HÄSTBACKA1, AND DAVIDE TAIBI.1,2
1
Tampere University, Tampere, Finland (e-mail: [email protected])
2
University of Oulu, Oulu, Finland (e-mail: [email protected])
Corresponding author: Sergio Moreschini (e-mail: [email protected]).

ABSTRACT The cloud continuum concept has drawn increasing attention from practitioners, academics,
and funding agencies and been adopted progressively. However, the concept remains mired in various
definitions with different studies providing contrasting descriptions. Therefore, to understand the concept
of cloud continuum and to provide its definition, in this work we conduct a systematic mapping study of
the literature investigating the different definitions, how they evolved, and where does the cloud continue.
The main outcome of this work is a complete definition that merges all the common aspects of cloud
continuum, which enables practitioners and researchers to better understand what cloud continuum is.

INDEX TERMS Cloud Continuum, Edge, Fog.

I. INTRODUCTION
Via the comparison amongst the different definitions,
The adoption of service-oriented architecture in cloud com- we shall observe the changes from the earliest to the
puting has profoundly changed the way how software, es- latest. In this way, we shall identify what are the new
pecially large-scale distributed systems, are built [24]. The aspects taken into account regarding “cloud continuum".
cloud is often viewed as an endless pool of resources, on • RQ3: Where does the cloud continue?
which we build and scale applications for various purposes. As cloud is “continued" into other infrastructures, we
Modern cloud systems, however, are inherently complex expect to find cloud-to-* extensions, where * could be
spanning public cloud to private cloud, possibly co-located on premise servers, but also edge, or other infrastruc-
across different regions, and may also include components tures. In this RQ we aim at understanding which are
and compute resources at the edge of the network. these extensions, so as to clarify where the cloud could
Cloud continuum is one of the most recent hypes in the be continued.
cloud computing domain and has raised interests of funding
agencies of EU and US [1], [3], [2]. However, while the The remainder of this paper is structured as follows.
hype is increasing, its definition is still not clear, and var- Section 2 presents related reviews. Section 3 describes the
ious papers are describing the concept of cloud continuum research method adopted. Section 4 presents the results
inconsistently. answering the RQs. Section 5 discusses the results while
Section 6 draws the conclusion and highlights future works.
In order to understand the differences between the dis-
parate definitions of cloud continuum, we propose a
system- atic mapping study of the literature. II. BACKGROUND AND RELATED WORK
In this work, we investigate the existing definitions and A. CLOUD, FOG, EDGE, AND MORE
common characteristics of “cloud continuum" as well as Cloud computing builds on the promise of economies of
their evolution through the time. scale in leveraging scalability and reliability. Scaling up is
We formulate three main Research Questions (RQs) as made possible by creating multiple compute instances and
follows. distributing them. Containers have long been the basis for
implementing microservices based architectures but recent
• RQ1: What are the definitions of cloud continuum?
advancement towards serverless and Functions as a Service
With this RQ we aim at understanding whether there
further emphasize the role of the cloud as a platform
are different definitions of cloud continuum.
abstract- ing underlying infrastructure resources [18] [6].
• RQ2: How has the definition of cloud continuum
Fog computing can be simplified as the cloud brought
evolved?
closer to the use case applications. Fog nodes minimize
load

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185
VOLUME 4, 2016 1

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

on the cloud and are able to host some services from the
cloud, and thus respond faster and also reduce networking
to the cloud [9]. [10] define that “fog is inclusive of cloud,
core, metro, edge, clients, and things,” and “fog seeks to
realize a seamless continuum of computing services from
the cloud to the things" instead of independent application FIGURE 1: The Search and Selection
resource pools. Edge computing takes place at the edge of
the network close to IoT devices, however, not
necessarily on the IoT devices themselves but as close as Process for the cloud continuum.
one hop to them [25]. Edge computing has been pushed Svorobej et al. [23] reviewed the orchestration mecha-
heavily by the telecommunica- tion industry but it has also nisms along the cloud-to-thing continuum with a focus on
emerged from the need to perform computation closer the container-based orchestration and orchestration
applications or with independence from cloud computing. architectures tailored for fog.
Edge computing is characterised by short latency in Asim et al. [5] provided a summary of research issues in
contrast to cloud computing where transmis- sion of data, Cloud computing and Edge computing, as well as current
allocation of resources typically includes delays. For developments in resolving them with CI approaches.
applications where large amounts of data needs to be Ghobaei-Arani et al. [13] provided a literature analysis
processed both fog and edge computing can introduce bene- aiming to identify the state-of-the-art mechanisms on re-
fits as cost savings in transfer, storage and processing. This source management approaches in the fog computing envi-
includes, for example, data from thousands of sensors, ronments.
audio and video streams, and emerging machine learning Kampars et al. [14] reviewed application layer protocols
(ML) based solutions. In VR and AR edge computing that can be used for the communication between the IoT, edge
together with low latency communication is claimed to and cloud layers.
enable cutting the cord, and it has been shown to achieve Spataru [22] surveyed the applications of Blockchain or
minimum gains of up to 30% reduction in end-to-end delay Smart Contracts for computing resources management, data
and even more for most storage, and services operation in the context of Cloud con-
parts of the communication [11]. tinuum.
Kansal et al. [15] presented a systematic literature re-
B. RELATED WORK view of the resource management approaches in fog/edge
Over the last few years, more and more researchers have been paradigm.
focusing on the cloud continuum paradigm. Therefore, Compared to our work, the previous literature reviews
some surveys/reviews on the subject have already been spent a noticeable effort in understanding technical and
presented. In the following, we report an overview of the man- agerial aspects of the cloud continuum (Table 1).
most relevant works available in the literature and discuss Instead, our work focuses on identifying the definition of
the differences with our work. the cloud continuum, how it evolved, and where the cloud
Al-Sharafi et al. [4] presented a literature review on the continues.
adoption of cloud computing services at the organizational
level, with a focus on the elements that contribute to long- III. RESEARCH METHODOLOGY
term adoption. In this study, we conducted a systematic mapping study of the
Pahl et al. [19] performed a literature review to identify, literature, by taking into account the guidelines proposed by
catalog, and compare the corpus of existing research on Petersen et al. [20]. The main aim was to systematically and
containers, their orchestration, and particularly the use of impartially summarize and classify the collected
this technology in the cloud. information regarding the research questions. Specifically
Bittencourt et al. [8] presented a literature review on IoT- herein, we aimed to not only characterize all the existing
Fog-Cloud continuum with the aim of understanding (i) definitions of the “cloud continuum" and other relevant
what are the best types of infrastructures to deploy the concepts, but also to investigate the evolution of such
entire ecosystem, (ii) what are the required mechanisms to definitions through time. The process of the study included
allow orchestration, data exchange, and resource four main steps. Firstly, we established the research
management, and questions. Secondly, we defined the search strategy.
(iii) what are the types of applications that can benefit most Thirdly, we defined the data extraction strategy. Fourthly,
from this ecosystem. we synthesized and visualized the ob-
Nguyen et al. [17] surveyed the current landscape of the tained results.
existing approaches and tools that attempt to cope with this
edge and cloud heterogeneity, scalability and dynamicity. A. SEARCH STRATEGY
Bendechache et al. [7] surveyed the list of suitable meth- The aim as well as the challenge for a systematic mapping
ods, algorithms, and simulation approaches for resource study was to define the search query that enables the
management in cloud-to-thing continuum. retrieval of a complete set of studies that contain the
Ramanathan et al. [21] conducted a survey to retrieve all definitions [16]. For such a purpose, the search strategy
the resource allocation techniques that have been developed encompassed a set of
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

2 VOLUME 4, 2016

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

TABLE 1: Summary of the related literature

reviews
Source Target Year
Al-Sharafi et al. [4] Adoption of cloud computing services at the organizational level 2017
Pahl et al. [19] Containers orchestration and usage in the Cloud 2017
Bittencourt et al. [8] IoT-Fog-Cloud continuum infrastructures, orchestration, data exchange, and resource management 2018
Nguyen et al. [17] Existing approaches and tools supporting edge and cloud development 2019
Bendechache et al. [7] Resource management in cloud-to-thing continuum 2020
Ramanathan et al. [21] Resource allocation techniques for the Cloud continuum 2020
Svorobej et al. [23] Orchestration mechanisms along the cloud-to-thing continuum 2020
Asim et al. [5] Research issues in Cloud computing and Edge computing 2020
ghobaei2020resource Resource management approaches in the fog computing environments 2020
Kampars et al. [14] Communication protocols between the IoT, edge and cloud layers 2021
Spartaru [22] Blockchain usage in Cloud continuum 2021
Kansal et al. [15] Resource management in fog/edge paradigm 2022
Our study Definition of Cloud continuum 2022

TABLE 2: Initial Search Results by Sources

VOLUME 4, 2016
Library Scopus IEEE ACM WoS Non-Duplicates
Count 271 148 61 102 378

steps, namely, defining search string, identifying key sources,

selecting primary studies, extracting data and synthesizing
the results.
The search strategy involved the outline of the most rele-
vant bibliographic sources and search terms, the definition
of the inclusion and exclusion criteria, and the selection
process relevant for the inclusion decision. Our search
strategy can be depicted in Fig. 1.
As for the search terms, we included cloud concepts, Fog,
Edge, and Continuum:
( cloud AND ( edge OR fog ) AND continuum )
We searched for scientific literature in four bibliographic
sources: Scopus,1 IEEEXplore Digital Library2, the ACM
Digital Library,3 and Web of Science4. The adoption of four
databases ensured the completeness of the search results.
We conducted our search on March 1st 2022, retrieving
378 non-duplicated papers from the four sources. The num-
ber of papers retrieved for each source is reported in Table
2.

B. PRIMARY STUDIES SELECTION

In order to select the primary studies from the preliminary
search results, we defined the inclusion and exclusion
criteria shown in Table 3. We included the research papers
published in journals or conferences, defining Cloud
Continuum. On the other hand, we excluded the research
papers that are not in English, duplicated, not discussing the
topic connected to the defined research questions.
Furthermore, we also excluded the papers that are not peer-
reviewed, as well as the work plans or roadmap, posters and
vision papers.
With the inclusion and exclusion criteria defined, we se-
lected the primary studies via two steps. Firstly, two of the
authors read the title and abstract of each paper separately
1
SCOPUS, https://www.scopus.com.
2
IEEEXPLORE DIGITAL LIBRARY https://ieeexplore.ieee.org/.
3
ACM DIGITAL LIBRARY: https://dl.acm.org.
4
WEB OF SCIENCE database: https://www.webofscience.com/.

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185
TABLE 3: Inclusion/Exclusion Criteria
Inclusion Papers defining the concept of cloud continuum
Inc./Exc.Criteria
ExclusionNot in English
Duplicated (post summarizing other websites) Out of topic (using the terms for other purp
Work plans, roadmaps, vision papers, posters
to determine whether it should be excluded or be read
fully. Whenever there was disagreement between them, a
third person assert the decision by the inclusion and
exclusion. Out of 378 papers screened, we had 93
disagreement with a Cohen’s kappa coefficient of 0.51,
indicating a moderate agreement [12]. As a result, we
identified 181 papers that need to be considered for the
next step.
We then ran a snowballing process including all the papers
referenced by the 181 papers. We then followed the same
process by applying inclusion and exclusion criteria to
their titles and abstracts. As a result, we included two more
papers: one peer-reviewed, and one grey literature [SP1].
The reason for including this specific non-peer-reviewed
work [SP1] is due to its large amount of citations;
especially when many of our selected papers referred to it
as the first definition of cloud continuum. Though
belonging to the gray literature, this study represents an
important milestone for the definition of cloud continuum
that has evolved over time with the addition/removal of
other keywords. It is also important to notice that no other
grey literature works are mentioned by the selected studies.
Each of the 183 papers (181 from the initial search,
and 2 from snowballing), was fully read by one of the
authors independently and evaluated by the inclusion and
exclusion criteria. As a result, we selected 36 papers.

C. DATA EXTRACTION STRATEGY

From the 36 Selected Papers (SP), we extracted the
data that answers our research questions. Importantly, we
extract the definitions on “continuum", the year of the
publication, and the information on where the cloud is
“continued". In addition to the key data mentioned above,
we also extracted the type of publication (e.g.
conference paper, or journal
3

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

article). continuum of resources available from the network edge to

TABLE 4: The information extracted from the selected pa- the cloud/datacenter" while Chiang et al. [SP2] defined
pers cloud continuum explicitly mentioning computational-
related as- pects, for instance, where and how the
RQInformation ExtractedMotivation computation is per- formed.
RQ1 Definition of Continuum We identify three main groups of definitions, with
RQ2 Publication Year To understand the chronological
respect to their main aspects. Each group is represented by
a block
RQ3 Cloud-to-* continuum evolution of the definition.
Identify the possible extensions of of a different colour in Fig. 4. The first and larger group
the cloud (*) mentioned in the SPs contains all those sources defining cloud continuum as an
aggregation/combination of different elements such as IoT
5
The description of the information extracted, together (The raw data is temporarily blind for peer-review. It will be uploaded
with their motivation and the mapping to the RQs, is into a permanent repository in case of acceptance of this paper).
reported in Table 3.

D. KEYWORDING
The different definitions were written in natural language.
Therefore, we needed to run a qualitative analysis among
the authors, to identify similar definitions and different
ones.
For this purpose, we applied a collective coding process
to answer our RQs:
The manual identification of the aforementioned informa-
tion was extracted collaboratively. From each paper, we
first extracted the definition and print to a post-it note
(RQ1). Then, one author attached it to a whiteboard, and
the other authors read all the other definitions proposed by
the papers. All the authors discussed one by one the
similarities and differences of each of the definitions, so as
to decide whether to group them into a single definition or
to create a new one. Finally, the authors re-position the
post-it notes reporting groups of similar definitions, and
their key differences. For each definition, all the authors
follow the same process to
identify common aspects.
Last, authors highlighted with different colors the contin-
uum extension to the cloud (RQ3)

E. REPLICABILITY
In order to allow replication and extension of our work by
other researchers, we prepared a replication package5 for
this study with the complete results obtained.

IV. RESULTS
As expected, publications on Cloud Continuum are contin-
uously growing in the recent years. The first definitions of
cloud continuum were presented in [SP1] and [SP2] in
2016. For the following three years only four papers are
identified as related to the definition of cloud continuum.
The interest in the topic started to grow in 2020. As
depicted in Fig. 3 the majority of paper identified are from
2021. In the remainder of this Section, we answer our RQs.

A. THE DEFINITIONS OF CLOUD CONTINUUM (RQ1)

The first definitions of Cloud Continuum were both presented
in 2016. Gupta et al. [SP1] defined cloud continuum as “a

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185
devices, fog and edge nodes. In this case, cloud defying the concept of cloud continuum in the same
contin- uum only refers to the continuum of resources, fashion. It is important to notice that each cluster is year-
but not of the computation. The second block contains all based as the definition evolved during the years (even
the sources defining cloud continuum with a particular when the author is the same). The highest amount of cluster
focus on the processing/computation. Finally, we group can be found in the first group of work, those related to the
together all those sources that do not belong to these two distribution of resources.
blocks. Within this group we can find 3 different clusters. The
Fig. 4 also shows that the definition of cloud continuum first one includes 5 different work agreeing on the same
has two different origins. Both of the papers which gave definition which puts the concept of continuum strictly
origin to the definition, as presented previously, have been related to the concept of fog. The second cluster is
published in 2016 but each of these focused on a different composed of 3 works which stress the importance of
aspect. While the definition in [SP1] focused on the having a combination of multiple edge and fog devices.
elements composing the system, the one proposed in [SP2] The third cluster defines the cloud continuum as an
was cen- tered around the concept of "where happens aggregation of heterogeneous resources from the Edge to
what". the cloud. The latter even tho it is composed of only two
The definition provided in [SP1] has been extended in works, has a definition that focuses on the data path with a
2019 from Kahvazadeh et al. [SP6] where the continuum bottom-up design.
of resources has been extended to ’the whole set of The other two clusters can be found one per each group.
resources from the edge up to the cloud’. In parallel to this, The first one, in the group "extension of the processing",
Balouek- Thomert et al. [SP5], centered their definition on includes two works defying cloud continuum as a Set of
the concept of "distributed resources services on demand". processing units located between the IoT and the Cloud.
Within these groups we can identify some clusters. Each The other one, also including two works, focuses on the
cluster combine multiple work within the same year different services across multiple infrastructures.

4 VOLUME 4, 2016

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

FIGURE 2: Selected Papers by Types FIGURE 3: Selected Papers by

Years TABLE 5: Initial Search Results by Sources

Reference Year Architecture Performance
Multi-Cloud Fog IoT Micro Datacenters Anywhere Simultaneous
[SP1] 2016 ✓
[SP2] 2016 ✓ ✓
[SP3] 2017 ✓
[SP4] 2018 ✓ ✓
[SP5] 2019 ✓
[SP6] 2019 ✓
[SP7] 2020
[SP8] 2020 ✓
[SP9] 2020
[SP10] 2020
[SP11] 2020 ✓ ✓ ✓
[SP12] 2020 ✓ ✓ ✓
✓ ✓
[SP13] 2020
[SP14] 2020 ✓
[SP15] 2021
[SP16] 2021 ✓
[SP17] 2021 ✓ ✓
[SP18] 2021 ✓ ✓
[SP19] 2021 ✓ ✓
[SP20] 2021 ✓
[SP21] 2021
✓ ✓
[SP22] 2021
✓
[SP23] 2021
[SP24] 2021
[SP25] 2021
[SP26] 2021
✓ ✓
[SP27] 2021
✓
[SP28] 2021
✓
[SP29] 2021
✓ ✓
[SP30] 2021
✓ ✓
[SP31] 2021
✓
[SP32] 2021
✓
[SP33] 2021
✓ ✓
[SP34] 2021 ✓ ✓
[SP35] 2022
[SP36] 2022

B. THE EVOLUTION OF THE CLOUD

• Multi-Cloud: definitions referring to multiple cloud
CONTINUUM DEFINITION (RQ2)
entities;
• Fog: definitions explicitly referring to Fog;
• IoT: definitions referring to internet of things, IoT,
In order to answer RQ2, we firstly extract the commonly
things;
adopted keywords of the cloud continuum definitions of the
• Anywhere: definitions explicitly reporting that the
selected papers. Herein, based on the opinions of two
computation can be executed everywhere;
domain experts, we extract six different keywords that
• Micro Datacenters: definitions explicitly reporting
delineate the characteristics (i.e., how, when and where) of
the use of micro datacenters to the goal of providing
cloud con- tinuum and specify the entities (i.e., what) it
low- latency access to data processing and data
connects. The keywords include:
storage.
VOLUME 4, 2016 5
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

2016 2017 2018 2019 2020 2021 2022

LEGEND Fog continuum expands the

computational capabilities from the
Fog and Edge edge network to the cloud layer
Multi-cloud
Combination [SP23]
of the cloud and
IoT including fog and edge computing
Keywords [SP15]
Extends the cloud computing
[SP29]
sources including fog and
computing
[SP20]
Extreme geographic distribution of [SP25]
infrastructure from the cloud to
the device The extension of the Cloud with
distributed micro-data centers
[SP13] and mobile Edge servers

[SP22]
An infrastructure where computing
resources are distributed from
endpoint devices at the edge of the
network to data centers or HPC
systems at its core
Complex collective of
[SP26]
components that varies in
capabilities and numbers Multi-cloud resources with local
Fluid ecosystem where distributed devices, including resource-
resources and services are [SP14]
constrained (mobile) edges and
aggregated on demand to fogs
support emerging data-driven
application workflows [SP19]

Aggregation of heterogeneous
[SP5]
resources along the data path from
theThe
Edgeaggregation
to the Cloud of

[SP21]
along the [SP24]
data path
the Edge to the
Next evolutionary step of cloud Combination of several edge and
Continuum of resources available The whole set of resources applications, incorporating other fog devices, with multi-cloud
from the network edge to the from the edge up to the cloud, compute facilities such as data- infrastructure
A landscapeandof platform
infrastructure
cloud/datacenter coined as IoT continuum generating nodes (IoT) and services
including gateway servers, local
[SP1] [SP6] intermediaries (edges, fogs) [SP18]
premisestoand
platforms public
process data
[SP16]
[SP11] captured at the
[SP17]

Hierarchical network where service Set of processing units located

providers can place compute between the IoT and the Cloud,
resources anywhere in the optimize response times and
set of processing units, such as
network bandwidth consumption in time-
Fog and cloud complement each fog servers and edge devices,
[SP12] sensitive applications
other to form a service continuum located between the IoT and
between the cloud and the endpoints Digital infrastructure jointly used by [SP31]
by providing mutually beneficial and complex application workflows [SP30]
interdependent services to make typically combining real-time
computing, storage, control, and data generation, processing Large digital ecosystem comprising
communication possible anywhere and computation IoT, Edge, Fog, and Cloud
along the continuum Computing, data cycles from
data gathering, processing
[SP10] and analysis to knowledge
[SP2]
Data processing and storage generation and decision making
may be local to an end-device [SP27]
at the edge of a network, located in
the cloud, or somewhere in
between, in “the fog”

[SP7]

Continuum that runs from

specialized embedded devices to Novel abstraction layer to express a
highly capable, standards-based continuous range of capacities
individual terminals
[SP36]
[SP8]

Digital services across multiple physical

infrastructures and administrative
boundaries
Digital services across
The continuum collaboration of administrative boundaries
devices from fog to servers [SP32]

[SP3] [SP33]

Set of operations that are required

to fulfil, in an automated way, user Systems that are simultaneously
and application requirements, taking executed on the Edge, Fog, and
The Fog and Cloud are a natural into consideration networking Cloud computing tiers
continuum of one another; thus, the features
marriage of these two killer [SP35]
technologies would offer an ideal [SP9]
IoT data provisioning of resources
Enables the deployment,
[SP4] upgrading, and migration of fog
services running on various nodes
located between IoT devices and
the cloud
[SP34]
Sensor devices deployed in the
Industrial Internet of Things (IIoT)

[SP28]

FIGURE 4: Definitions of cloud computing grouped by year and concepts. Each column represent a different year while the
coloured blocks represents different aspects. Arrows between two blocks indicate that there is a direct citation to the
definition.

6 VOLUME 4, 2016

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

• Simultaneous: definitions explicitly reporting that the

etc., where IoT is the individual entity/device providing ser-
computation can be simultaneously executed on multi-
vices; Zeiner and Unterberger [SP28] defines edge-to-cloud
ple nodes.
continuum as a data-driven Internet of Things combines
Therefore, by summarizing the adoptions of these key- the physical world with the world of information, where
words by the selected papers in chronological order IoT is referred to as the assembly instead of the individual.
(reported in TABLE 5), we can observe the evolution of the Specially, Mehran et al. [SP22] define cloud continuum as the
cloud continuum definition. extension of the cloud with distributed micro-datacenters
The two earliest definitions, [SP1] and [SP2] in 2016, and mobile edge servers, which is the first and only time
both anchored the concept of fog between cloud and edge, when micro-datacenter is used.
where the term “continuum" was firstly used by its literal Until February 2022, two studies also provided
meaning in this context. Specially, Chiang and Zhang [SP2] definitions to cloud continuum. Dustdar et al. [SP35] define
empha- sized that within such continuum, services like it by em- phasizing it is the system that is “simultanously"
computing, storage, control and communication could be executed on the edge, fog, and cloud computing tiers.
provided any- where between cloud and edge. From Similarly, in 2021, Risco et al. [SP17] also mentioned the
2017 to 2018, the two studies, [SP3] and [SP4] continued term “simultanously" by indicating cloud continuum
adopting the term “continuum" describing the combination “simultaneously involves both on-premises and public
of fog and cloud, when Peng et al. [SP4] indicated that the Cloud platforms to process data cap- tured at the edge". The
continuum of fog and cloud could provide ideal IoT data other definition given by Spillner et al. did not specify the
provisioning. In 2019, Balouek-Thomert et al. [SP5] also entities that cloud continuum aggregating but emphasize it
mentioned “computing continuum" as a fluid ecosystem is an “novel abstraction layer to express a continuous range
with aggregated resources and services but didn’t of capacities".
emphasized its positioning between cloud and edge.
Meanwhile, also in 2019, Kahvazadeh et al. [SP6] C. WHERE DOES THE CLOUD CONTINUE (RQ3)
proposed the term “IoT continuum" but similarly coined the Among the 36 SPs, nine of them mention the continuum
definition as a whole set of resources between edge and as “cloud-to-thing(s) continuum”. Therein, these studies in-
cloud. dicate that cloud continuum connects or is placed between
Since 2020, the number of studies that provided definitions cloud(s) and the IoT-connected devices (i.e., things). Spe-
to cloud continuum has been growing sharply. In 2020, cially, Kassir et al. [SP12] state that “cloud-to-thing(s) con-
eight studies mentioned the concept of “continuum" and tinuum" is equivalent to "Fog-to-Cloud continuum". Mean-
similarly placed the concept as the services between cloud while, two studies, [SP22] and [SP23], use “Cloud-fog con-
and the end-devices (i.e., edge). However, though five tinuum" or “fog continuum" indicating the continuum ex-
studies, [SP7], [SP8], [SP11], [SP12], and [SP14], tends the cloud towards fog, which could either refer to fog
mentioned “fog" when defining continuum, none of the nodes (i.e., also things) or fog in general.
studies have clearly distin- guish them; when some studies, On the other hand, seven papers amongst the 36 SPs
e.g., [SP8], [SP12], indi- cate continuum is between cloud use the term “Edge-to-Cloud continuum" (or Cloud-edge
and fog. Meanwhile, four studies mentioned IoT when continuum, or edge/cloud continuum) indicating the cloud
defining continuum [SP11]- [SP14]; however, the relation “continues" towards edge nodes. Kahvazadeh et al. [SP6]
between continuum and IoT is not clearly delineate either. use the term “IoT continuum" but describe the same
On the other hand, Kassir et al. [SP12] also indicate that connection between cloud and edge. Three studies use
compute resources can be placed anywhere in the network directly the term “cloud continuum" but also define it as
when citing [SP2]. Furthermore, Spillner et al. [SP11] combination of cloud and edge.
emphasize that continuum is more than simply a “multi- Furthermore, ten studies use “Computing Continuum" to
cloud" but incorporating other compute facilities, e.g., emphasize the computing capability instead of the connec-
mobile devices, IoT sensor nodes, edges and fogs, which is tion of entities. Within these definitions, the “continuum"
the first time continuum is connected with the notion of can be used connecting any entities, e.g., edge, fog, local
“multi-cloud". devices (i.e., IoT or things), data centers, etc. Specially,
In 2021, nine studies mentioned “fog" as a critical entity Balouek-Thomert et al. [SP5] do not describe the specific
in the definition of continuum. Different from previously, nodes being connected by continuum but defines
many of these studies, e.g., [SP18]-[SP20], [SP22], [SP23], “computing continuum" as “a digital infrastructure jointly
have anchored the continuum concept as the combination or used by com- plex application workflows". Beckman et al.
aggregation of several fog, edge, IoT devices or services, or [SP14] provide a similar definition as “a collective of
the extension of the cloud. Meanwhile, four studies [SP16]- components with various capabilities and numbers in
[SP19] also indicate that cloud continuum is a “multi- aggregate". Spillner et al. [SP36] provide a high-level
cloud" infrastructure. On the other hand, eight studies abstracted definition of computing con- tinuum as “novel
indicate that IoT is a crucial part of the cloud continuum abstraction layer to express a continuous range of
concept when, however, the interpretation of the term capacities"
differs slightly. For example, Xhafa and Krause [SP27] Comparatively, early studies, e.g., [SP1] and [SP2], did
define cloud continuum as a large digital ecosystem not try to provide distinguishable terms but only use the
comprising IoT, Edge, Fog, and
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185
VOLUME 4, 2016 7

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

([SP16]) and game theory ([SP20]), are proposed as promis-

ing solutions for application deployment and orchestration.
The robustness of the cloud continuum is also
considered a critical aspect for the future. For example, [SP11]
highlights the complexity of the awareness of application
deployment towards the adaptation for higher resilience.
[SP23] and [SP15] also indicate that tolerant IoT services
and self- healing components shall serve for the future
steps towards structural and behavioral optimization of
cloud continuum system.
Furthermore, security of the cloud continuum systems
FIGURE 5: Architecture of Cloud Continuum. ([SP2], [SP4]) is also a key aspect when specific
techniques, e.g., Information-Centric Network integration
([SP9]), and Hybrid key distribution ([SP6]) are seen as
future works.
term “continuum" literally trying to describe the conceptual
Other performance characteristics, e.g., scalability
idea. Similarly, these two studies also place the
([SP28]), mobility ([SP23]) and consistency ([SP2]),
“continuum" between cloud to edge or cloud to fog.
together with the corresponding ways of acquisition
([SP11]), comparison ([SP29]) and benchmark ([SP26]) are
V. DISCUSSION also mentioned as the challenges.
Several definitions of Cloud Continuum have been Meanwhile, other future challenges include high-level
proposed in the last six years. However, only few have been ab- straction models and architectural trade-off ([SP2],
used or extended. [SP10]), [SP14]), interfaces and user experience ([SP1],
It is interesting to notice the two main types of [SP2]), posi- tioning and localization, ([SP4]) and the
definitions, one considering the continuum as distribution Incentives of device participation ([SP2]). The researchers
of resources in different network elements, including IoT, shall consider con- tributing to the solutions to the above
Fog, Edge, but also HPC, while the other definition challenges in order to enrich the domain knowledge of
considering the contin- uum as an extension of the cloud continuum research.
processing power to different nodes, often mentioning the
possibility of executing also AI. The investigation of the B. THREATS TO VALIDITY
different cloud continuum con- cepts allowed us to draw an We are aware that our work is subject to threats to valid-
ity. The terms Cloud, Fog, IoT, and Edge are sufficiently
overall architecture of the cloud
continuum (Figure 5) stable to be used as search strings. In order to assure the
Based on the analysis conducted in this work, we can retrieval of all papers on the selected topic, we searched
propose a new definition of cloud continuum, combining broadly in general publication databases, which index most
the most frequently mentioned aspects. well-reputed publications. To improve the reliability of this
work, we defined search terms and applied procedures that
Cloud Continuum is an extension of the traditional Cloud towards
can bemultiple entities
replicated by (e.g.,
others.Edge,
SinceFog,
thisIoT)
is athat providestudy
mapping analysis, proc
and no systematic review, the inclusion/exclusion criteria
are only related to whether the topic of Cloud Continuum is
present in a paper or not, as suggested by [20]. As for the
analysis procedure, since our analysis only uses descriptive
statistics, the threats are minimal. However, we are aware
A. FUTURE CHALLENGES that the synthesis of the definition might be subjective. To
The results of this work enabled us to distill a set of chal- mitigate this threat, the analysis was done collaboratively,
lenges for the cloud continuum. Therein, the majority of using a collecting coding methods, and discussing with all
the SPs point out the challenges concerning the dynamic the authors about inconsistencies. The Kohen K index about
allocation of the computation([SP31][SP22]), and in par- our disagreement also confirms the quality of the qualitative
ticular of the execution of the AI, and the related resource analysis performed.
orchestration, network partitioning ([SP30]) and support for
context-awareness ([SP9]). VI. CONCLUSION
As part of the resource orchestration, job scheduling is In this work, we proposed a systematic mapping study on
also identified as one of the most common challenges that the definition of Cloud Continuum to obtain an overview of
need to be addressed in the future ([SP4], [SP13], [SP14], its existing definitions and how the concept has been
[SP16], [SP24]). Tools such as Kafka-ML ([SP30]) and evolved.
network virtualization ([SP1]) are proposed towards such We identified 36 studies which proposed definitions to
an end. Furthermore, other techniques, e.g., adopting APIs Cloud Continuum dated from 2016. All these definitions are
summarized in Figure 4. We organized all the 36 existing
definitions in chronological order.
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

8 VOLUME 4, 2016

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

In conclusion, we propose to complement existing defini- [15] Kansal, P., Kumar, M., Verma, O.P.: Classification of resource manage-
tions into a common one that merges explicitly two aspects: ment approaches in fog/edge paradigm and future research prospects: a
the continuum as extension of the resources, and as exten- systematic review. The Journal of Supercomputing pp. 1–60 (2022)
[16] Kitchenham, B., Charters, S.: Guidelines for performing systematic litera-
sion of computational capabilities. ture reviews in software engineering (2007)
As a result, we formulated the definition of cloud con- [17] Nguyen, P., Ferry, N., Erdogan, G., Song, H., Lavirotte, S., Tigli, J.Y.,
tinuum as ”an extension of the traditional Cloud towards Solberg, A.: Advances in deployment and orchestration approaches for
iot- a systematic review. In: 2019 IEEE international congress on Internet
multiple entities (e.g., Edge, Fog, IoT) that provide of Things (ICIOT). pp. 53–60. IEEE (2019)
analy- sis, processing, storage, and data generation [18] Nupponen, J., Taibi, D.: Serverless: What it is, what to do and what not
capabilities.” The new definition enables both practitioners to do. In: 2020 IEEE International Conference on Software Architecture
Companion (ICSA-C). pp. 49–50 (2020).
and re- searchers to better understand the concept of cloud [19] Pahl, C., Brogi, A., Soldani, J., Jamshidi, P.: Cloud container
contin- uum and to gain insights into the potential advance in technologies: a state-of-the-art review. IEEE Transactions on Cloud
Computing 7(3), 677–692 (2017)
service- [20] Petersen, K., Feldt, R., Mujtaba, S., Mattsson, M.: Systematic mapping
oriented computing. studies in software engineering. In: 12th International Conference on
As regards future work, we are planning to extend this Evaluation and Assessment in Software Engineering (EASE) 12. pp. 1–
10 (2008)
work in the context of cognitive continuum. [21] Ramanathan, S., Shivaraman, N., Suryasekaran, S., Easwaran, A., Borde,
E., Steinhorst, S.: A survey on time-sensitive resource allocation in the
cloud continuum. it-Information Technology 62(5-6), 241–255 (2020)
REFERENCES
[22] Spataru, A.: A review of blockchain-enabled fog computing in the cloud
[1] Cloud computing: towards a smart cloud computing continuum (2020), continuum context. Scalable Computing: Practice and Experience 22(4),
https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/ 463–468 (2021)
opportunities/topic-details/ict-40-2020, accessed: 2022-07-07 [23] Svorobej, S., Bendechache, M., Griesinger, F., Domaschka, J.: Orchestra-
[2] Cloud computing and high-throughput computing resources for tion from the cloud to the edge. The Cloud-to-Thing Continuum pp. 61–
collaborative research in computational neuroscience (crcns) 77 (2020)
grantees (2021), https://beta.nsf.gov/funding/opportunities/ [24] Wei, Y., Blake, M.B.: Service-oriented computing and cloud computing:
cloud-computing-and-high-throughput-computing-resources-collaborative, Challenges and opportunities. IEEE Internet Computing 14(6), 72–75
accessed: 2022-07-07 (2010)
[3] Cognitive cloud: Ai-enabled computing continuum from cloud to edge [25] Yousefpour, A., Fung, C., Nguyen, T., Kadiyala, K., Jalali, F.,
(ria) (2022), https://ec.europa.eu/info/funding-tenders/opportunities/ Niakanlahiji, A., Kong, J., Jue, J.P.: All one needs to know about fog
portal/screen/opportunities/topic-details/horizon-cl4-2022-data-01-02, computing and related edge computing paradigms: A complete survey.
accessed: 2022-07-07 Journal of Systems Architecture 98, 289 – 330 (2019).
[4] Al-Sharafi, M.A., Arshah, R.A., Abu-Shanab, E.A.: Factors influencing
the continuous use of cloud computing services in organization level.
In: Proceedings of the international conference on advances in image APPENDIX A: THE SELECTED PAPERS
processing. pp. 189–194 (2017) [SP1] Gupta, H. et al., 2016. SDFog: A software defined computing
[5] Asim, M., Wang, Y., Wang, K., Huang, P.Q.: A review on computational architecture for QoS aware service orchestration over edge
intelligence techniques in cloud and edge computing. IEEE Transactions devices. arXiv preprint arXiv:1609.01190.
on Emerging Topics in Computational Intelligence 4(6), 742–763 (2020)
[6] Aslanpour, M.S., Toosi, A.N., Cicconetti, C., Javadi, B., Sbarski, P., [SP2] Chiang, M. and Zhang, T., 2016. Fog and IoT: An overview
Taibi, D., Assuncao, M., Gill, S.S., Gaire, R., Dustdar, S.: Serverless of research opportunities. IEEE Internet of things journal,
edge computing: Vision and challenges. In: 2021 Australasian Computer 3(6), pp.854-864.
Science Week Multiconference. ACSW ’21, Association for Comput-
[SP3] Coughlin, T., 2017. Convergence through the cloud-to-thing
ing Machinery, New York, NY, USA (2021). , https://doi.org/10.1145/
3437378.3444367
consortium [future directions]. IEEE Consumer Electronics
[7] Bendechache, M., Svorobej, S., Takako Endo, P., Lynn, T.: Simulating Magazine, 6(3), pp.14-17.
resource management across the cloud-to-thing continuum: A survey and [SP4] Peng, L. et al., 2018. Toward integrated Cloud–Fog networks
future directions. Future Internet 12(6), 95 (2020) for efficient IoT provisioning: Key challenges and solutions.
[8] Bittencourt, L., Immich, R., Sakellariou, R., Fonseca, N., Madeira, E., Future Generation Computer Systems, 88, pp.606-613.
Curado, M., Villas, L., DaSilva, L., Lee, C., Rana, O.: The internet of
things, fog and cloud continuum: Integration and challenges. Internet of [SP5] Balouek-Thomert, D. et al., 2019. Towards a computing con-
Things 3, 134–155 (2018) tinuum: Enabling edge-to-cloud integration for data-driven
[9] Bonomi, F., Milito, R., Zhu, J., Addepalli, S.: Fog computing and its role workflows. The International Journal of High Performance
in the internet of things. In: Proceedings of the first edition of the MCC Computing Applications, 33(6), pp.1159-1174.
workshop on Mobile cloud computing. pp. 13–16 (2012)
[SP6] Kahvazadeh, S. et al., 2019. Securing combined fog-to-cloud
[10] Chiang, M., Ha, S., I, C., Risso, F., Zhang, T.: Clarifying fog computing
and networking: 10 questions and answers. IEEE Communications Maga-
systems: challenges and directions. In FTC 2019 (pp. 877-
zine 55(4), 18–20 (2017). 892). Springer.
[11] Elbamby, M.S., Perfecto, C., Bennis, M., Doppler, K.: Edge computing [SP7] Domaschka, J. et al., 2020. Towards an architecture for reliable
meets millimeter-wave enabled vr: Paving the way to cutting the cord. capacity provisioning for distributed clouds. In Managing
In: 2018 IEEE Wireless Communications and Networking Conference Distributed Cloud Applications and Infrastructure (pp. 1-25).
(WCNC). pp. 1–6 (2018). Palgrave Macmillan, Cham.
[12] Emam, K.E.: Benchmarking kappa: Interrater agreement in software pro-
cess assessments. Empirical Software Engineering 4(2), 113–133 (1999) [SP8] Milojicic, D., 2020. The edge-to-cloud continuum.
[13] Ghobaei-Arani, M., Souri, A., Rahmanian, A.A.: Resource management Computer, 53(11), pp.16-25.
approaches in fog computing: a comprehensive review. Journal of Grid [SP9] Da Silva, D.M.A. and Sofia, R.C., 2020. A discussion on
Computing 18(1), 1–42 (2020)
context-awareness to better support the IoT cloud/edge con-
[14] Kampars, J., Tropins, D., Matisons, R.: A review of application layer
communication protocols for the iot edge cloud continuum. In: 2021
tinuum. IEEE Access, 8, 193686-193694.
62nd International Scientific Conference on Information Technology and [SP10] Rosendo, D. et al., 2020. E2clab: Exploring the computing
Management Science of Riga Technical University (ITMS). pp. 1–6. continuum through repeatable, replicable and reproducible
IEEE (2021)

VOLUME 4, 2016 9

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

edge-to-cloud experiments. In 2020 CLUSTER (pp. 176- [SP31] Torres, D.R. et al., 2021. An open source framework based
186). IEEE. on Kafka-ML for Distributed DNN inference over the Cloud-
[SP11] Spillner, J. et al., 2020. Rule-based resource matchmaking to-Things continuum. Journal of Systems Architecture, 118,
for composite application deployments across IoT-fog-cloud p.102214.
continuums. In 2020 IEEE/ACM UCC (pp. 336-341). IEEE. [SP32] Alberternst, S. et al., 2021. From Things into Clouds–and
[SP12] Kassir, S. et al., 2020. Service placement for real-time back. In 2021 CCGrid (pp. 668-675). IEEE.
applica- tions: Rate-adaptation and load-balancing at the [SP33] Alberternst, S. et al., 2021. Orchestrating Heterogeneous De-
network edge. In CSCloud 2020/EdgeCom 2020 (pp. 207- vices and AI Services as Virtual Sensors for Secure Cloud-
215). IEEE. Based IoT Applications. Sensors, 21(22), p.7509.
[SP13] Bendechache, M. et al., 2020. Simulating resource manage- [SP34] Cˇ ilic´, I. et al., 2021. Towards Service Orchestration
ment across the cloud-to-thing continuum: A survey and future for the Cloud-to-Thing Continuum. In 2021 SpliTech (pp.
directions. Future Internet Journal, 12(6), p.95. 01-07). IEEE.
[SP14] Beckman, P. et al., 2020. Harnessing the computing continuum [SP35] Dustdar, S. et al., 2022. On distributed computing continuum
for programming our world. Fog Computing: Theory and systems. IEEE Transactions on Knowledge and Data Engi-
Practice, pp.215-230. neering.
[SP15] Alonso, J. et al., 2021. Optimization and Prediction Tech- [SP36] Spillner, J. et al., 2022. Intent-Based Placement of Microser-
niques for Self-Healing and Self-Learning Applications in a vices in Computing Continuums. Future Intent-Based Net-
Trustworthy Cloud Continuum. Information, 12(8), p.308. working, 38-50
[SP16] Luckow, A. et al., 2021. Exploring task placement for edge-
to- cloud applications using emulation. In 2021 ICFEC (pp.
79- 83). IEEE.
[SP17] Risco, S. et al., 2021. Serverless workflows for containerised
applications in the cloud continuum. Journal of Grid
Comput- ing, 19(3), pp.1-18. SERGIO MORESCHINI is a Ph.D. candidate in
[SP18] Spillner, J., 2021. Self-balancing architectures based on liquid the Faculty of Information Technology and Com-
functions across computing continuums. In UCC (pp. 1-6). munication Sciences, Tampere University, Fin-
land. He is a researcher at Cloud Software Evolu-
[SP19] Hass, D. and Spillner, J., 2021. Interactive application tion and Assessment (CloudSEA) research group.
deploy- ment planning for heterogeneous computing His main research interest focuses on tools for
continuums. In AINA (pp. 551-560). Springer, Cham. MLOps. He also contributes actively to the do-
[SP20] Kimovski, D. et al., 2021. Cloud, Fog, or Edge: Where to mains of empirical software engineering, open-
Compute?. IEEE Internet Computing, 25(4), pp.30-36. source software quality, data-driven software en-
[SP21] Balouek-Thomert, D. et al., 2021. Evaluating policy-driven gineering, etc.
adaptation on the Edge-to-Cloud Continuum. In 2021 Urgen-
tHPC (pp. 11-20). IEEE.
[SP22] Mehran, N. et al., 2021. A Two-Sided Matching Model for
Data Stream Processing in the Cloud–Fog Continuum. In
2021 CCGrid (pp. 514-524). IEEE.
FABIANO PECORELLI is a postdoctoral re-
[SP23] Nezami, Z. et al., 2021. Decentralized edge-to-cloud load searcher in Cloud Software Evolution and As-
balancing: Service placement for the Internet of Things. sessment (CloudSEA) research group at Tampere
IEEE Access, 9, pp.64983-65000. University, Finland. He received a bachelor’s,
[SP24] Balouek-Thomert, D. et al., 2021. MDSC: modelling dis- master’s, and Ph.D. degree in Computer Science
tributed stream processing across the edge-to-cloud contin- from the University of Salerno, Italy. His
uum. In UCC 2021 (pp. 1-6). research interests include software code and test
quality, predictive analytics, mining software
[SP25] Kimovski, D. et al., 2021. Mobility-Aware IoT Applications
repositories, software maintenance and evolution,
Placement in the Cloud Edge Continuum. IEEE Transactions and empiri- cal software engineering. He serves
on Services Computing. and had served
[SP26] Rosendo, D. et al., 2021. Reproducible performance as a referee for various international journals in the field of software
optimiza- tion of complex applications on the edge-to-cloud engineering (e.g., TSE TOSEM, EMSE, JSS).
continuum. In CLUSTER (pp. 23-34). IEEE.
[SP27] Xhafa, F. and Krause, P., 2021. IoT-Based Computational
Modeling for Next Generation Agro-Ecosystems: Research
Issues, Emerging Trends and Challenges. In IoT-based
Intelli- gent Modelling for Environmental and Ecological
Engineering (pp. 1-21). Springer, Cham. XIAOZHOU LI is a postdoctoral researcher in
[SP28] Zeiner, H. and Unterberger, R., 2021. Time-aware Data the Faculty of Information Technology and
Spaces-A key Computing Unit in the Edge-to-Cloud Contin- Commu- nication Sciences, Tampere University,
Finland. He is a researcher at Cloud Software
uum. In 2021 FiCloud (pp. 250-255). IEEE.
Evolution and Assessment (CloudSEA) research
[SP29] Dizdarevic´, J. and Jukan, A., 2021. Experimental group. His research interests include open-source
Benchmark- ing of HTTP/QUIC Protocol in IoT Cloud/Edge software quality, software maintenance and
Continuum. In ICC 2021 (pp. 1-6). IEEE. evolution, user review opinion mining, data-
[SP30] Carnero, A. et al., 2021. Managing and Deploying Distributed driven empirical soft- ware engineering,
and Deep Neural Models Through Kafka-ML in the Cloud- computational game studies, gamification design,
to- Things Continuum. IEEE Access, 9, pp.125478-125495. etc.

10 VOLUME 4, 2016

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

SONIA NAZ is a Ph.D. student at the Faculty

of Information Technology and Communication
Sciences, Tampere University, Finland. She is
part of the Cloud Software Evolution and
Assessment (CloudSEA) research group. Her
research area is Empirical Software Engineering
with a particular focus on Agile processes for
software develop- ment. Sonia got her Master of
Computer Science from Fatima Jinnah Women’s
University, Pakistan, and a Master of Software
Engineering from Inter-
national Islamic University, Pakistan. She was a lecturer in the Computer
Science course at the University of Wah, Pakistan (2015 - 2020).

DAVID HÄSTBACKA is an Assistant Professor

(tenure track) at Tampere University, Finland. He
received his D.Sc.(Tech.) degree (with
distinction) in 2013 and M.Sc(Tech.) degree in
2007 at Tam- pere University of Technology. His
research inter- ests are in system and software
architectures, and interoperability of software
systems in production and energy systems
applications.

DAVIDE TAIBI is full Professor at the

University of Oulu (Finland) where he head the
M3S Cloud research group. His research is
mainly focused on Empirical Software
Engineering applied to cloud- native systems,
with a special focus on the migra- tion from
monolithic to cloud-native applications. He is
investigating processes, and techniques for
developing Cloud Native applications,
identifying cloud-native specific patterns and anti-
patterns. He is member of the International
Software Engineer-
ing Network (ISERN) from 2018. Before moving to Finland, he has been
Assistant Professor at the Free University of Bozen/Bolzano (2015-2017),
post-doctoral research fellow at the Technical University of Kaiserslautern
and Fraunhofer Institute for Experimental Software Engineering - IESE
(2013-2014) and research fellow at the University of Insubria (2007-2011).

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Access. 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/ACCESS.2022.3229185

VOLUME 4, 2016 11

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/