Content uploaded by Rute C. Sofia
Author content
All content in this area was uploaded by Rute C. Sofia on Oct 08, 2020
Content may be subject to copyright.
Digital Object Identifier
A Discussion on Context-awareness to
Better Support the IoT Cloud/Edge
Continuum
DANIEL MANIGLIA AMANCIO DA SILVA1, RUTE C. SOFIA 2,3
1COPELABS - University Lusofona, Campo Grande 388, 1700-097 Lisboa, Portugal (e-mail: daniel.maniglia@hotmail.com
2Research Institute of the Free State Bavaria associated with the Technical University of Munich. Guerickerstr. 25, 80634 München (e-mail: sofia@fortiss.org)
3University Lusofona, Campo Grande 376, 1700-097 Lisboa, Portugal
Corresponding author: Daniel M. Silva (e-mail: daniel.maniglia@hotmail.com).
ABSTRACT This paper debates on notions of context-awareness as a relevant asset of networking and
computing architectures for an Internet of Things (IoT), in particular in regards to a smoother support of the
the networking operation between Cloud and Edge. Specifically, the paper debates on notions of context-
awareness and goes over different types of context-awareness indicators that are being applied to Edge
selection algorithms, covering the approaches currently used, the role of the algorithms applied, their scope,
and contemplated performance metrics. Lastly, the paper provides guidelines for future research in the
context of Cloud-Edge and the application of context-awareness to assist in a higher degree of automation
of the network and, as consequence, a better support of the Cloud to Edge continuum.
INDEX TERMS Context-awareness, Internet of Things (IoT), Edge/Fog Computing.
I. INTRODUCTION
The daily routines of regular citizens integrate a wide variety
of highly heterogeneous Internet of Things (IoT) systems.
Such systems integrate simple sensors and actuators, net-
working devices, and more complex cyber-physical systems,
such as smart sensors and mobile personal devices (e.g.,
smartphones) which further integrate a large number of
sensing interfaces. For instance, in personal mobile devices,
sensors such as accelerometer, GPS, microphone, or camera,
bring in the possibility of exploiting new types of data coined
as smart data or small data, derived from the track and trace
process of different aspects of the routine of citizens, e.g.,
roaming habits; application usage; location preferences [1]–
[3]. While such sensorial capability is giving rise to new types
of data and services, it brings in additional computational and
data exchange challenges. Firstly, the datasets are richer, even
though data is fine-grained, and often polled more frequently,
thus resulting in larger volumes of data to be analysed [4],
[5]. Secondly, the IoT communication architectural models
that are being applied to support such data transmission
cannot cope with the properties of such traffic (e.g., high
volumes of small data packets). This is both due to the
increasingly larger number of devices being interconnected
to the Internet and to a higher heterogeneity of the hardware
and software involved [6]. Thirdly, the processing of the
richer and more complex data sets require support from
computationally heavy Artificial Intelligence (AI) engines
supported by the Cloud. While the Cloud helps in supporting
the required data analytics complex computation, the more
heterogeneous IoT scenarios available today are often not
compatible with the delays derived from pushing all of the
data processing and storage to the Cloud [7].
In the quest to assist smart data computation in IoT sce-
narios, related trends concern a decentralisation of Internet
services and of networking functions across the so-called
Cloud to Edge continuum (Cloud-Edge). The Cloud-Edge
continuum refers to a set of operations that are required to ful-
fil, in an automated way, user and application requirements,
taking into consideration networking features. Today, the
Cloud-Edge continuum relies already on context-awareness
indicators, as shall be debated in section III and IV of the
paper. However, this is limited, often tied to strict network
guarantees, and such indicators are not sufficient, in our opin-
ion, to sustain novel and more dynamic IoT environments,
where the Edge is mobile, highly heterogeneous (e.g., an
embedded device, a smart satellite).
Existing trends attempt to best serve mobility of devices
and users; the need for data and user privacy; the larger vol-
umes of sensitive data to be analysed, and the requirements
to handle such data [8] [9]. This is giving rise to alter-
VOLUME 4, 2016 1
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
native ways to provide data exchange in IoT environments,
as occurs with the paradigm of Edge/Fog computing [10].
Usually, such paradigms take into consideration task, service
and resource offloading, to assist in a better resource manage-
ment. However, to support better dynamic environments, it is
necessary to consider how to best adjust the computational
needs to the respective context and hence, it is relevant to
revisit notions of context-awareness.
This is the motivation for this work. We believe that
context-awareness can assist in a smoother transition of
computational/storage/networking resources, from the Cloud
to the Edges and vice-versa. To assist this debate, the paper
contributions are three-fold: i) the paper provides a thorough
review on work that focuses on context-awareness for IoT; ii)
the paper contributes to the definition of context-awareness
in IoT and debates on specific context-awareness indicators
that can be considered to better support a smooth Cloud-Edge
continuum; iii) the paper provides guidelines concerning the
integration of context-awareness in Cloud-Edge IoT environ-
ments
The review provided in this paper has been based on an
extensive review of papers concerning context-awareness for
IoT environments. This review has comprised an analysis of
papers from 2011 until 2020, based on the paper keywords
"context", "context-awareness", "Edge computing", "behav-
ior inference" and also focused on the area of "networking
architectures", areas of interest of the authors. The selection
took into consideration the following aspects: i) the work
has been described in peer-reviewed publications with a high
Impact Factor; ii) most recent references have been preferred
against older ones.
The paper is organised as follows. Section II goes
over related work, explaining the contributions of this pa-
per towards related literature. Section III describes back-
ground on IoT communication aspects, including notions
for Edge/Fog computing. Section IV discusses the role of
context-awareness for IoT and describes specific indicators
that are being used to assist on a selection of the whereabouts
to store and compute data. Section V specifically focuses
on the integration of context-awareness into IoT Fog/Edge
architectures, detailing existing areas of interest. Section VI
concludes the paper, discussing findings and providing a set
of guidelines for future research.
II. RELATED WORK
Several related work has focused on different categories
of network communication challenges experienced in IoT
scenarios. Specifically focusing on the domain of eHealth,
Islam et al. describe on challenges existing in current IoT
healthcare middleware [11]. Dimitrov et al. delve into issues
concerning data mining, data storage, and data analysis [12]
in IoT eHealth scenarios. Poon et al. focus on sensor commu-
nication [13]. Sensing and big data management have been
debated by Yue Hong et al. [14], and the identification of
key components of an end-to-end IoT has been discussed by
Baker et al. [15]. This line of work identifies and highlights
challenges that IoT faces in Smart Health environments,
including security, privacy, usability, energy awareness. This
line of work is relevant to our work, given that eHealth
scenarios experience specific challenges, in particular con-
cerning data privacy and data sensitivity, challenges which
can be lowered if the underlying networking architectures
assist in handling data locally, within trusted environments.
In the context of IoT for Smart Cities environments, where
smart applications are used to collect and to exchange dif-
ferent types of data, Sholl et al. propose a Smart City archi-
tecture that harnesses the power of semantic technologies to
allow machines and people to understand the relationships
among data in a context-aware manner, and to extract knowl-
edge [16]. Choi et al. propose a software architecture to assist
efficient middleware deployment in Smart Cities, by relying
on semantic technologies [17].
Context-awareness is also highly relevant to data mining
and classification as, for instance, debated in the context of
vehicular networks by Ruta et al. [18].
Chen et al. surveyed Edge computing resource-efficient
offloading mechanisms [19]. Still in regards to Edge/Fog
computation offloading, Wang et al. [20] collected and inves-
tigated key issues, methods related to the offloading problem
in Cloud to Edge environments.
Another category of related work focuses on the under-
standing and definition of context and context-awareness,
which are central points in this review paper. Some au-
thors [21] define context in association with parameters such
as location, neighbour identity, time-based indicators such as
visit duration, environmental characteristics such as season,
temperature. Ryan et al. define context as the user’s location,
environment, identity, and the time [22]. Dey et al. states
that context is the user’s emotional state, focus of attention,
location and orientation, date and time, objects and people
in the user’s environment [23]. Schilit et al. argue that the
only important aspects of context are user location, the user’s
neighbours, and resources near the user [24]. They define
context to be subject to the constantly changing execution
environment and the environment is thus three-fold: com-
puting environment, user environment and physical environ-
ment. Sofia et al. [25] define context indicators based on the
network layers, derived from roaming patterns of users.
Our work differs from the described related work in that it
debates on research that applied context-awareness to assist
in automating the IoT Cloud-Edge operation, surveying the
use of context data to improve network performance in
Edge/Fog Computing for environments exhibiting variability,
such as occurs today in IoT environments that involve Thing-
to-Thing and People-to-Thing interactions.
III. IOT COMMUNICATION BACKGROUND
IoT environments can be broadly grouped into two cat-
egories, related with the specific requirements and ex-
pected benefits: Consumer IoT (CIoT) and Industrial IoT
(IIoT) [26] [27]. Both IoT categories rely on computational
architectures that integrate four main functional blocks: data
2
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
capture; data storage; data analysis; data exchange. However,
the requirements on these different environments introduce
different challenges.
IIoT [28] [29] focuses on how smart machines, networked
sensors, people, and data analytics can improve aspects such
as productivity, service efficiency. IIoT is applied to different
vertical markets, e.g., Industrial Automation, Smart Cities,
Smart Factory, Logistics. Moreover, specific IIoT markets
include also Smart Health, Smart Energy, or People-at-Work
markets.
IIoT is expected to support both Machine to Machine
(M2M) and People-to-Machine interaction, either for appli-
cation monitoring, control, for instance, or as part of a self-
organised system, with a distributed control which does not
necessarily require human intervention. IIoT often implies
higher data rates and larger data volumes. Moreover, applica-
tions are often mission and/or safety critical requiring strict
and bounded guarantees, such as low delay, low jitter, or zero
packet congestion.
CIoT concerns the use of IoT in aspects related to the
daily living of people and aims at increasing usefulness of
technology in such context. It involves scenarios focused on
the interconnection of consumer and devices, as well as of
anything involving the users’ environments such as homes,
offices, and cities [30]. Vertical markets of CIoT comprise,
for instance, Smart Cities, Connected Mobility, Smart Health.
Personal IoT (PIoT) is a sub-category of CIoT focused on
the application of smart systems based on personal devices,
as well as based on sets of sensors and actuators applied to
improve quality of living. The most popular form of PIoT
concerns fitness solutions aiming to bring awareness and to
improve physical health of users [11], [31]. Currently, these
systems are more commonly used in the context of Ambient
Assisted Living (AAL). AAL encompasses technical systems
to support people with special needs in their daily routine,
e.g., elderly [32], temporarily disabled people, or anyone that
needs supportive monitoring. [33] [34].
A. SUPPORTING ASYNCHRONOUS AND
MANY-TO-MANY COMMUNICATION
From a protocol perspective, the interconnection of IoT
Things and applications, be it directly to a controller or to
the Cloud-Edge, has been traditionally deployed by having
sensors harvesting information and sending such information
to a specific device/system, for instance, an IoT gateway, an
IoT broker. Hence, initially the point-to-point communica-
tion model provided by TCP/IP was enough to support the
requirements of IoT data exchange.
With the increase of IoT devices, as well as with the new
software-based and open-source approaches being explored,
IoT services are becoming more complex, thus introducing
additional requirements. Firstly, several, if not most of the
devices in IoT scenarios are mobile. Secondly, the integration
of the different hardware and software solutions that compose
IoT environments is often provided by third-parties. Thirdly,
IoT scenarios often accommodate hundreds or thousands of
devices, often communicating across large distances.
To cope with these changes, data exchange in IoT needs
to be supported by mechanisms capable of accommodating
aspects such as mobility, security, large distances, intermit-
tent connectivity. For this, it is necessary to support two main
communication requirements: asynchronous communication
support, and many-to-many service distribution support.
Internet communication protocols are therefore evolving,
in the context of IoT, to support the 2 main mentioned
requirements. For instance, the procotocols that support
IoT data exchange (IP-based messaging protocols) usually
rely on a broker-based publish/subscribe communication
model [27]. The broker is a mediating functional entity that
handles data being exchanged between producers and con-
sumers in an asynchronous way. First, consumers subscribe
specific data interests. Then they get the matching infor-
mation provided by producers [35]. Broker models create
an abstraction layer as well, and can protect the identity
of producers and subscribers. Nevertheless, they are still
focused on reaching hosts (machines), and not really focused
on the content.
The most recent evolution of publish/subscriber models
is embodied in the Information-centric Networking publish-
subscriber paradigm [36]. Information-centric Network
(ICN) is a networking architectural paradigm that is focused
on data reachability, instead of host reachability. In the con-
text of IoT, ICN models seem to be promising as the network
semantics that ICN automatically supports aspects such as
consumer mobility [37], security, as well as address abstrac-
tion by design. There are today several ICN architectural
proposals such as the Data-Oriented Network Architecture
(DONA) [38]; the Network of information (NETINF) [39]; the
Content-Centric Networking (CCN) [40]; the Named Data
Network (NDN) [36] . Out of these, the networking archi-
tecture most suitable for IoT is the NDN architecture [41].
The NDN architecture defines a simple and robust data-
centric, pull-based and receiver-driven communication model
based on the exchange of two packets types, Interest and
Data packets. Interest packets are sent by consumers willing
to express interest on specific content and contain hierarchi-
cal, global content names [42]. Data packets are sent by
producers upon the reception of Interest packets, and carry
chunks of signed data.
B. THE ROLE OF EDGE/FOG COMPUTING
Fog Computing [10], also known as Edge computing [43],
extends the Cloud Computing paradigm to the "Edges" of the
network, bringing in new opportunities to explore applica-
tions and services. By assisting the placement of storage and
data processing closer to the data sources, Edge Computing
brings in benefits in terms of latency and energy consump-
tion [44]–[48], for instance.
In this paper Fog and Edge are used indistinctly, as we
consider the most recent evolution of Edge, where the Edge
is elastic in terms of whereabouts or even system composition
3
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
(e.g., an Edge can be a smart sensor, a satellite, a smartphone,
or an eNodeB) [49]. However, other views provide a stricter
perspective on Edge computing, derived from a telecom-
munications perspective. This is the case, for instance, of
the Mobile Edge Computing (MEC) architecture, where the
Edge is still within the control of the operator and consists
of a specific computational unit, working in isolation or
being complementar to the Cloud. While in Fog computing,
the notion of Edge is more elastic, covering, for instance,
field-level and end-user devices (e.g., smartphones, smart
sensors) [50].
For IoT, and due to aspects such as security (e.g., the
need to have in-plant security and resilient communication
in IIoT scenarios), large distances, as well as large sets
of frequent data lead to an insufficiency of the Cloud to
satisfy the Quality of Service (QoS) requirements (e.g., low
latency) of different IoT applications. Fog computing aims to
overcome some limitations of Cloud-centric IoT-models by
taking advantage of Edge network resources [51].
Fog/Edge network architectures integrate mechanisms to
better distribute data computation and data storage across a
specific infrastructure. Figure 1 illustrates such a networking
architecture, where Layers represent Tier levels.
FIGURE 1. Fog/Cloud Computing Architecture.
Tier 1 integrates IoT field-level devices, such as sensors
and actuators. These are data sources, devices that capture
and distribute data to other Tier devices, same Tier, or next
Tier level. Tier 2 (FOG) integrates IoT devices coined as
Fog nodes [52]. IoT hubs and gateways that gather data
and process information fall into this category. The Tier 2
level includes also devices such as routers and Access Points
(AP). Fog nodes are arranged in a hierarchical way and
communication is only possible between a parent-child pair
in the hierarchy. Given that these devices are in the edges of
the network, often located in Customer Premises, Fog nodes
often have limited resources. Tier 3 (CLOUD) devices often
have a significantly higher amount of resources. These are,
for instance, virtual machines in data centers.
IV. CONTEXT-AWARENESS IN IOT
Context-aware computing has been used over the last decade
in desktop applications, Web applications, mobile com-
puting, and pervasive/ubiquitous computing. Context-aware
computing is a computing paradigm in which applications
can discover and take advantage of context information such
as user location, time of day, neighbouring users and devices,
user activity [53]. Context is "any information that can be
used to characterise the situation of an entity. An entity can
be a person, place, or object that is considered relevant to the
interaction between a user and an application" [54].
Hence, there is a significant difference between context
information and raw data sent by IoT devices. Raw data
concerns unprocessed data that is directly retrieved from data
sources. Context information is generated by processing raw
sensor data. Such data is validated, checked for consistency,
and often annotated with meta-data [55]. For instance, GPS
sensor readings can be considered as raw sensor data. Once
it represents a geographical location, it becomes context.
IoT environments comprise a large number of devices
and large volumes of data to be transmitted and processed.
Understanding how to use and how to process that data to
generate relevant knowledge is therefore dependent on the
type of context of services, users, as well as networking
architectures. Hence, context-awareness plays a critical role
in assisting decisions in terms of what data needs to be
processed, where that data should be processed, and when.
In regards to IoT environments, context-awareness is being
applied to improve different computational aspects, as sum-
marised in Table 1. The table categorizes related work first by
area of application, explaining the purpose (column 2), and
where the related work applies such improvements (column
3). The context-awareness indicators used are presented in
column 4, while the applicability domain (vertical market)
is provided in column 5. The related literature is placed in
column 6.
A first area of related work (row 1) applies context-
awareness to authentication and control in untrusted environ-
ments. Context-aware access control mechanisms are being
used to provide system access, using the user personal data
context and not personal data.
A second area of related work (row 2) concerns the
application of context-awareness for resource management
and orchestration. Such line of work focuses on improving
the overall computational and networking performance by
exploring context-awareness to reduce energy consumption;
reduce overall latency; message overload. Context-awareness
is relevant to assist in deciding when and where to process
data, thus contributing to latency reduction, for instance [69],
[79].
In regards to forwarding/routing applications (row 3), one
example of the work being pursued is to take into consid-
eration, at a network level, the context that surrounds users
and that can assist in better defining opportunities for data
transmission over time, and space, i.e., context-awareness at
the network layers [25]. Context-awareness can also assist
4
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
TABLE 1. Context-awareness Application in IoT Computational and Networking Architectures.
Area Purpose Where Indicators Domain References
Authentication
and Control
To facilitate secure
authentication
and control of IoT devices in
untrusted environments
End-user, Edge,
Cloud
Physical context
(light, temperature, noise);
computing context
(app usage, touch
patterns),
User context (roaming
patterns, neighbor cluster,
location, etc.)
PIoT, SmartHome,
SmartHealth,
Automation
[20], [56]–[65]
Resource
management
and orchestration
To improve the overall IoT
computational platform and
services providing control
based on context
Edge, Cloud
Device usage and
resources, time, location,
user behavior
SmartHome,
SmartCities [66]–[74]
Forwarding/routing
To introduce context-awareness
to the network
layers in order to provide better
chances to forward, in particular
in highly heterogeneous scenarios
End-user, Edge,
Cloud
Location, roaming data,
accelerometer, speed,
device usage (e.g., battery)
Opportunistic IoT
environments,
SmartCities, PIoT
[25], [75]–[78]
Offloading To assist in deciding where
to store data and to compute Edge, Cloud
Location, battery, latency
and other network
measurement indicators
Smart Logistics,
SmartCities,
SmartMobility
[71], [79]–[82]
Semantic
interoperability
To assist in a
smoother operation in large-scale
heterogeneous environments
Edge
Application delay
requirements, request
history, service/user
similarity
Smart Logistics,
SmartCities,
SmartMobility
[83]–[88]
Multi-layer
interoperability
To provide IoT interoperable
ecosystems by using
context-awareness
in a bottom-up approach
Edge, Cloud
Users,
device usage,
environmental
indicators
Data discovery,
management and
communication
[87]–[91]
in a better distribution of in-network caching; more efficient
naming aggregation, as well as in a more efficient data
transmission in the context of large-scale scenarios [68], [92].
Another category of work focuses on applying context-
awareness to offloading(row 4), i.e., to decide where to store
data, and also where to compute such data. For this purpose,
parameters such as location, residual energy of the device are
being applied.
A fifth category of related work focuses on semantic
interoperability aspects, including related work that has been
delving on improving data sharing on upper layers via se-
mantic modelling. Once the information can be collected
from a range of sources and some information must be
explicitly supplied by users, context-awareness can be ap-
plied to identify the relationships level between people, the
ownership of devices and communication channels providing
a seamless approach to the interconnection of devices and
their data exchange, by providing automated support to the
interconnection of, for instance, different data models derived
from different applicability domains(row 5). In this context,
indicators derived from the application layer (such as delay
requirements), or even similarity between used services is
being applied to assist in an automated interconnection.
The last row (row 6) covers work related with multi-layer
interoperability. This work focuses on discovery, manage-
ment and high-level communication of IoT devices in het-
erogeneous IoT platforms, defining, for instance, component-
based methods for middleware interoperability.
V. CONTEXT-AWARENESS AND SELECTION
ALGORITHMS
Edge selection algorithms provide a smoother operation in
Cloud-Edge environments, in particular when considering
services and applications that might require very short re-
sponse times, or applications that might produce a large
quantity of data to be processed. Sending such data to the
Cloud may result in large delays, or excessive energy con-
sumption by the network devices, for instance. An example
of technological solutions that require adaptation on the go
are Mobile Pervasive Augmented Reality (MPARS) [93]. As
stated by Pascoal et al., context-awareness derived from the
surrounding environment, as well as from the user’s habits,
and computational preferences can assist a better aggregation
and placement of data. This also assists in extending the reach
of computational and networking architectures, considering
Edges that are mobile and resource constrained.
Current Edge placement algorithms are often focused on
aspects such as latency and energy improvement, as sum-
marised in Table 2, which summarises Edge selection algo-
rithms, categorizing them by context information considered
(column 2), scope (column 3), as well as performance metrics
relied by the algorithm (column 4).
Wattenhofer and Zollinger [94] propose XTC (1), a topol-
ogy control algorithm to select the nearest Edge in ad-
hoc wireless networks. The algorithm has three steps: 1-
Neighbour ordering; 2- neighbour order exchange, and 3-
Edge selection. It also has the advantage of not requiring
full knowledge of the topology, or prior status on the node
whereabouts. It therefore applies heuristics that take into con-
5
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
TABLE 2. Edge Selection Algorithms.
Nr. Algorithm Context Scope Metrics
1XTC [94] Neighbor ordering, neighbor order
exchange, Edge selection (no need
for node location or global topology
knowledge)
Network control
algorithm for ad-hoc environments,
node selection
Latency
2 AR Edge Selection [95] Application requirements
and traffic load Edge Selection Latency threshold
3Opportunistic
Routing (ENS_OR) [96]
Distance to sink
and residual energy
on both neighboring nodes
Opportunistic relaying algorithm,
node selection
Distance to sink
and residual energy
4 RNST [97] Node location via trilateration Wireless indoor location,
node selection
relative direct distance
radio range
5 Latency Bounded Mini-
mum Influential Node Se-
lection [98]
Diffusion, influential nodes impact
on the speed of diffusion Node Selection Propagation speed
of influential nodes
6
Computation
and networking load
node selection [99]
Application requirements
node resources
link quality resources
Node Selection Node and link availability
7 Branch-and-bound
algorithm [80]
Computational overhead
Joint node
selection and
resource allocation
Computation
overhead threshold
8 Threshold based
policy [100]
Network policies Edge/Cloud selection Delay threshold
9 ThriftyEdge [19] Task resource usage Resource-efficient computation
Node selection
Latency,
minimum
node usage
(e.g., CPU, memory)
10 Edge Selected MPA [101] Topology, neighborhood status Edge Selection Channel quality
and proximity
sideration the direct neighborhood of the node, at different
instant in times.
Sumit et al., propose an Edge selection algorithm for AR
applications (2) [95]. Their algorithm takes into consider-
ation both application requirements and traffic load. The
algorithm scans the state of neighboring edges to find a "best"
Edge which can serve the user within a specified latency
threshold.
An Energy Saving via Opportunistic Routing algorithm
(3) is proposed by Luo et al. [96]. This algorithm is applied
in wireless networks and has two steps to select nodes: 1-
selects a set of nodes with higher centrality; and 2- considers
the status provided by other nodes it encounters. In terms of
indicators, it considers the node’s distance to the data sink,
and the residual energy on both the parent and successor
nodes.
The RNST algorithm (4) [97] was developed to support
mobile nodes for indoor wireless networks. It provides node
location via trilateration, considering four steps: 1- A mobile
node broadcasts a location message to its neighboring ref-
erence nodes, then the reference nodes return a confirmed
location message; 2- The mobile node calculates the dis-
tances between each pair of nodes and judges if any of the
three reference nodes can form almost equilateral triangle; 3-
Compute the estimated locations of the mobile node using
each of the possible equilateral triangles; 4- The mobile node
calculates the average location value.
ALatency-bounded Minimum Influential Node Selection
Algorithm (5), proposed by Zou et al. [98], provides a se-
lection of the most influential nodes on a (social) network,
where most influential relates with the speed of diffusion,
and not with connectivity. The algorithm steps are: 1- find
a 1-hop dominating set for the rest of the nodes that are
INACTIVE 2- the vertices that could be influenced by the 1-
hop Latency-Bounded Minimum Influential Node Selection
in Social Networks.
A computation and networking load node selection algo-
rithm (6) [99] is one of the first works, to our knowledge,
that realises the need to meet, in an integrated way, both
application and networking requirements. It relies on node
resources such as CPU, and link resources, and considers as
selection metrics node availability derived from a node and
link QoS perspective.
The branch-and-bound algorithm (7) [80] proposed by
Pham et al. addresses both node selection and resource
allocation. It is a Divide and conquer algorithm that uses
computation overhead to select a node.
Zhao et al. provide a threshold-based policy mechanism
(8) [100] which finds an optimal local node to run delay-
tolerant applications in mobile Cloud computing, designing
a scheduling scheme to realize the cooperation between the
local Cloud and the Internet Cloud.
Xu Chen et al. introduce ThriftyEdge (9) [19], a resource-
efficient IoT task offloading algorithm. The authors rely on a
hybrid approach to exploit the hierarchical resources across
local nodes, nearby helper nodes, and the Edge-Cloud in
proximity. They propose a topology-sorting-based task graph
partition algorithm in order to reduce the Edge resource
6
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
occupancy (usage).
Yudan Wang and Ling Qiu propose MPA (ES-MPA)
(10) [101], a low complexity Edge discovery and selection
approach to better support the massive connectivity of cellu-
lar IoT.
Summarising, most of the existing algorithms that provide
support for node selection usually consider a minor set of
network or node requirements, e.g., latency, residual energy.
Less common is the attempt to combine application/task
and network requirements. Moreover, out of the analysis
performed, we did not find algorithms that took into consid-
eration behavior inference (node, link, service, and user), for
instance.
VI. CONCLUSIONS AND GUIDELINES FOR FUTURE
RESEARCH
This paper reviews work concerning the relevancy of inte-
grating context-awareness to improve the IoT data exchange
across Edge and Cloud, in particular regarding the needs
of IoT services and applications. The paper provides an
overview on the needs of different IoT environments and re-
vises proposals which consider context-awareness indicators
to provide operational improvements, e.g., latency reduction,
lower energy consumption. The review shows that the role
of context-awareness in IoT environments is acknowledged,
but that its integration to support more dynamic IoT environ-
ments is still limited, often being defined simply as location
to assist traffic locality, or node resources, as described in
section IV. As also debated in section IV, there are sev-
eral opportunities to improve the Edge-Cloud continuum, by
considering different levels of context-awareness indicators,
derived from application requirements and from networking
requirements, and also derived from the behaviour learning
of inference of user activities and habits (e.g., roaming pat-
terns; preferred network locations). It is therefore relevant to
consider some of the findings, to derive guidelines for future
research. A summary of such guidelines is:
•IoT applications are becoming more and more dis-
tributed across the Cloud and Edge, as addressed in
section III-B. Edge selection mechanisms (cf. section
V) consider a limited integration of context-awareness.
Other indicators which may better support more dy-
namic environments (e.g., indicators derived from mo-
bility patterns) can be considered, thus being a relevant
area of future work.
•The support of many-to-many asynchronous commu-
nication is today based on publish/subscribe models,
as described in section III-A, is relevant to better sup-
port the needs of IoT data exchange. It provides the
opportunity to scale better in comparison to the tradi-
tional client/server communication models, through par-
allel operation, message caching, tree-based or network-
based routing. In addition to the IP-based messaging
protocols commonly used in IoT environments, it is
relevant to further delve on the relevancy of paradigms
such as ICN, and focus on the integration of ICN ar-
chitectures, such as NDN, into IoT. A relevant research
area, which has been initiated but still requires much
more exploration be it in terms of performance mea-
surement or in terms of network architectures evolution
is the applicability of ICN paradigms into IoT environ-
ments.
•Variable and heterogeneous IoT scenarios, such as the
ones embodied in PIoT, will benefit from bringing data
processing closer to the end-user, as discussed in section
V. Context-awareness therefore plays a relevant role, be
it in terms of better defining traffic and computational
locality, or to assist in a more automated behavior of
IoT networking architectures, end-to-end.
•To promote feedback in close-to-realtime, context-
awareness can assist the network in making decisions
that improve the network operation and, as conse-
quence, can also improve data processing.
REFERENCES
[1] H. Zheng and J. Jumadinova, “Monitoring the well-being of a person
using a robotic-sensor framework,” in 2016 AAAI Spring Symposium
Series. Association for the Advancement of Artificial Intelligence, 2016.
[2] M. Miloševi´
c, M. T. Shrove, and E. Jovanov, “Applications of smart-
phones for ubiquitous health monitoring and wellbeing management,”
JITA-JOURNAL OF INFORMATION TECHNOLOGY AND APPLI-
CATIONS, vol. 1, no. 1, 2011.
[3] K. Hänsel, “Wearable and ambient sensing for well-being and emotional
awareness in the smart workplace,” in Proceedings of the 2016 ACM
International Joint Conference on Pervasive and Ubiquitous Computing:
Adjunct. ACM, 2016, pp. 411–416.
[4] Y. Sun, H. Song, A. J. Jara, and R. Bie, “Internet of things and big data
analytics for smart and connected communities,” IEEE access, vol. 4, pp.
766–773, 2016.
[5] R. Kitchin and T. P. Lauriault, “Small data in the era of big data,”
GeoJournal, vol. 80, no. 4, pp. 463–475, 2015.
[6] I. Yaqoob, E. Ahmed, I. A. T. Hashem, A. I. A. Ahmed, A. Gani,
M. Imran, and M. Guizani, “Internet of things architecture: Recent
advances, taxonomy, requirements, and open challenges,” IEEE wireless
communications, vol. 24, no. 3, pp. 10–16, 2017.
[7] V. Cozzolino, A. Y. Ding, J. Ott, and D. Kutscher, “Enabling fine-grained
edge offloading for iot,” in Proceedings of the SIGCOMM Posters and
Demos. ACM, 2017, pp. 124–126.
[8] A. Hosseinian-Far, M. Ramachandran, and C. L. Slack, “Emerging trends
in cloud computing, big data, fog computing, iot and smart living,” in
Technology for Smart Futures. Springer, 2018, pp. 29–40.
[9] P. Sethi and S. R. Sarangi, “Internet of things: architectures, protocols,
and applications,” Journal of Electrical and Computer Engineering, vol.
2017, 2017.
[10] F. Bonomi, R. Milito, J. Zhu, and S. Addepalli, “Fog computing and its
role in the internet of things,” in Proceedings of the first edition of the
MCC workshop on Mobile cloud computing. ACM, 2012, pp. 13–16.
[11] S. R. Islam, D. Kwak, M. H. Kabir, M. Hossain, and K.-S. Kwak, “The
internet of things for health care: a comprehensive survey,” IEEE Access,
vol. 3, pp. 678–708, 2015.
[12] D. V. Dimitrov, “Medical internet of things and big data in healthcare,”
Healthcare informatics research, vol. 22, no. 3, pp. 156–163, 2016.
[13] C. C. Poon, B. P. Lo, M. R. Yuce, A. Alomainy, and Y. Hao, “Body sensor
networks: In the era of big data and beyond,” IEEE reviews in biomedical
engineering, vol. 8, pp. 4–16, 2015.
[14] Y. Yuehong, Y. Zeng, X. Chen, and Y. Fan, “The internet of things in
healthcare: An overview,” Journal of Industrial Information Integration,
vol. 1, pp. 3–13, 2016.
[15] S. B. Baker, W. Xiang, and I. Atkinson, “Internet of things for smart
healthcare: Technologies, challenges, and opportunities,” IEEE Access,
vol. 5, pp. 26 521–26544, 2017.
[16] S. Sholla, R. Naaz, and M. A. Chishti, “Semantic smart city: Context
aware application architecture,” in 2018 Second International Conference
7
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
on Electronics, Communication and Aerospace Technology (ICECA).
IEEE, 2018, pp. 721–724.
[17] C. Choi, C. Esposito, H. Wang, Z. Liu, and J. Choi, “Intelligent power
equipment management based on distributed context-aware inference in
smart cities,” IEEE Communications Magazine, vol. 56, no. 7, pp. 212–
217, 2018.
[18] M. Ruta, F. Scioscia, F. Gramegna, S. Ieva, E. Di Sciascio, and R. P.
De Vera, “A knowledge fusion approach for context awareness in vehic-
ular networks,” IEEE Internet of Things Journal, vol. 5, no. 4, pp. 2407–
2419, 2018.
[19] X. Chen, Q. Shi, L. Yang, and J. Xu, “Thriftyedge: Resource-efficient
edge computing for intelligent iot applications,” IEEE network, vol. 32,
no. 1, pp. 61–65, 2018.
[20] R. Wang and D. Tao, “Context-aware implicit authentication of smart-
phone users based on multi-sensor behavior,” IEEE Access, vol. 7, pp.
119 654–119 667, 2019.
[21] P. J. Brown, J. D. Bovey, and X. Chen, “Context-aware applications:
from the laboratory to the marketplace,” IEEE personal communications,
vol. 4, no. 5, pp. 58–64, 1997.
[22] N. Ryan, J. Pascoe, and D. Morse, “Enhanced reality fieldwork: the
context aware archaeological assistant,” Bar International Series, vol.
750, pp. 269–274, 1999.
[23] A. K. Dey, “Context-aware computing: The cyberdesk project,” in Pro-
ceedings of the AAAI 1998 Spring Symposium on Intelligent Environ-
ments, 1998, pp. 51–54.
[24] B. Schilit, N. Adams, and R. Want, “Context-aware computing applica-
tions,” in Mobile Computing Systems and Applications, 1994. Proceed-
ings., Workshop on. IEEE, 1994, pp. 85–90.
[25] R. C. Sofia, I. Santos, J. Soares, S. Diamantopoulos, C.-A. Sarros,
D. Vardalis, V. Tsaoussidis, and A. d’Angelo, “Umobile d4. 5: Report
on data collection and inference models,” UMOBILE Consortium, Tech.
Rep., 2017.
[26] D. Bandyopadhyay and J. Sen, “Internet of things: Applications and
challenges in technology and standardization,” Wireless Personal Com-
munications, vol. 58, no. 1, pp. 49–69, 2011.
[27] R. C. Sofia, “An overview on the evolution of iot communication ap-
proaches (pre-print),” MDPI pre-print, 2019.
[28] E. Sisinni, A. Saifullah, S. Han, U. Jennehag, and M. Gidlund, “Indus-
trial internet of things: Challenges, opportunities, and directions,” IEEE
Transactions on Industrial Informatics, vol. 14, no. 11, pp. 4724–4734,
2018.
[29] H. Geng, “The industrial internet of things (iiot),” 2017.
[30] R. Williams, E. McMahon, S. Samtani, M. Patton, and H. Chen, “Identi-
fying vulnerabilities of consumer internet of things (iot) devices: A scal-
able approach,” in 2017 IEEE International Conference on Intelligence
and Security Informatics (ISI). IEEE, 2017, pp. 179–181.
[31] A. S. Yeole and D. Kalbande, “Use of internet of things (iot) in health-
care: A survey,” in Proceedings of the ACM Symposium on Women in
Research 2016. ACM, 2016, pp. 71–76.
[32] L. Mainetti, L. Patrono, A. Secco, and I. Sergi, “An iot-aware aal
system for elderly people,” in Computer and Energy Science (SpliTech),
International Multidisciplinary Conference on. IEEE, 2016, pp. 1–6.
[33] M. Hassanalieragh, A. Page, T. Soyata, G. Sharma, M. Aktas, G. Mateos,
B. Kantarci, and S. Andreescu, “Health monitoring and management
using internet-of-things (iot) sensing with cloud-based processing: Op-
portunities and challenges,” in 2015 IEEE international conference on
services computing (SCC). IEEE, 2015, pp. 285–292.
[34] P. P. Ray, “Home health hub internet of things (h 3 iot): An architec-
tural framework for monitoring health of elderly people,” in Science
Engineering and Management Research (ICSEMR), 2014 International
Conference on. IEEE, 2014, pp. 1–3.
[35] N. Naik, “Choice of effective messaging protocols for iot systems: Mqtt,
coap, amqp and http,” in 2017 IEEE international systems engineering
symposium (ISSE). IEEE, 2017, pp. 1–7.
[36] L. Zhang, A. Afanasyev, J. Burke, V. Jacobson, P. Crowley, C. Pa-
padopoulos, L. Wang, B. Zhang et al., “Named data networking,” ACM
SIGCOMM Computer Communication Review, vol. 44, no. 3, pp. 66–73,
2014.
[37] R. C. Sofia, “Guidelines towards information-driven mobility
management,” Future Internet, vol. 11, no. 5, 2019. [Online]. Available:
https://www.mdpi.com/1999-5903/11/5/111
[38] T. Koponen, M. Chawla, B.-G. Chun, A. Ermolinskiy, K. H. Kim,
S. Shenker, and I. Stoica, “A data-oriented (and beyond) network archi-
tecture,” in ACM SIGCOMM Computer Communication Review, vol. 37,
no. 4. ACM, 2007, pp. 181–192.
[39] B. Ahlgren, M. D’ambrosio, C. Dannewitz, A. Eriksson, J. Golic,
B. Grönvall, D. Horne, A. Lindgren, O. Mämmelä, M. Marchisio et al.,
“Second netinf architecture description,” 4WARD EU FP7 Project, De-
liverable D-6.2 v2. 0, 2010.
[40] V. Jacobson, D. K. Smetters, J. D. Thornton, M. F. Plass, N. H. Briggs,
and R. L. Braynard, “Networking named content,” in Proceedings of the
5th international conference on Emerging networking experiments and
technologies. ACM, 2009, pp. 1–12.
[41] M. Meddeb, “Information-centric networking, a natural design for iot
applications?” Ph.D. dissertation, INSA de Toulouse, 2017.
[42] M. Amadeo, C. Campolo, A. Iera, and A. Molinaro, “Named data
networking for iot: An architectural perspective,” in Networks and Com-
munications (EuCNC), 2014 European Conference on. IEEE, 2014, pp.
1–5.
[43] W. Shi, J. Cao, Q. Zhang, Y. Li, and L. Xu, “Edge computing: Vision and
challenges,” IEEE Internet of Things Journal, vol. 3, no. 5, pp. 637–646,
2016.
[44] J. Gedeon, F. Brandherm, R. Egert, T. Grube, and M. Mühlhäuser, “What
the fog? edge computing revisited: Promises, applications and future
challenges,” IEEE Access, vol. 7, pp. 152847–152 878, 2019.
[45] D. Sarddar, S. Barman, P. Sen, and R. Pandit, “Refinement of resource
management in fog computing aspect of qos,” International Journal of
Grid and Distributed Computing, vol. 11, no. 5, pp. 29–44, 2018.
[46] M. Iorga, L. Feldman, R. Barton, M. J. Martin, N. S. Goren, and C. Mah-
moudi, “Fog computing conceptual model,” NIST: National Institute of
Standards and Technology, 2018.
[47] T. Choudhari, M. Moh, and T.-S. Moh, “Prioritized task scheduling in fog
computing,” in Proceedings of the ACMSE 2018 Conference. ACM,
2018, p. 22.
[48] P. Maiti, J. Shukla, B. Sahoo, and A. K. Turuk, “Qos-aware fog nodes
placement,” in 2018 4th International Conference on Recent Advances in
Information Technology (RAIT). IEEE, 2018, pp. 1–6.
[49] A. Yousefpour, C. Fung, T. Nguyen, K. Kadiyala, F. Jalali, A. Niakan-
lahiji, J. Kong, and J. P. Jue, “All one needs to know about fog computing
and related edge computing paradigms: A complete survey,” Journal of
Systems Architecture, vol. 98, pp. 289–330, 2019.
[50] M. Chiang, S. Ha, F. Risso, T. Zhang, and I. Chih-Lin, “Clarifying fog
computing and networking: 10 questions and answers,” IEEE Communi-
cations Magazine, vol. 55, no. 4, pp. 18–20, 2017.
[51] M. Aazam, S. Zeadally, and K. A. Harras, “Deploying fog computing
in industrial internet of things and industry 4.0,” IEEE Transactions on
Industrial Informatics, vol. 14, no. 10, pp. 4674–4682, 2018.
[52] E. M. Tordera, X. Masip-Bruin, J. Garcia-Alminana, A. Jukan, G.-J. Ren,
J. Zhu, and J. Farré, “What is a fog node a tutorial on current concepts
towards a common definition,” arXiv preprint arXiv:1611.09193, 2016.
[53] G. W. Musumba and H. O. Nyongesa, “Context awareness in mobile
computing: A review,” International Journal of Machine Learning and
Applications, vol. 2, no. 1, p. 5, 2013.
[54] G. D. Abowd, A. K. Dey, P. J. Brown, N. Davies, M. Smith, and
P. Steggles, “Towards a better understanding of context and context-
awareness,” in International symposium on handheld and ubiquitous
computing. Springer, 1999, pp. 304–307.
[55] L. Sanchez, J. Lanza, R. Olsen, M. Bauer, and M. Girod-Genet, “A
generic context management framework for personal networking envi-
ronments,” in Mobile and Ubiquitous Systems: Networking & Services,
2006 Third Annual International Conference on. IEEE, 2006, pp. 1–8.
[56] A. Kaur, G. Rai, A. Malik et al., “Authentication and context awareness
access control in internet of things: A review,” in 2018 8th International
Conference on Cloud Computing, Data Science & Engineering (Conflu-
ence). IEEE, 2018, pp. 630–635.
[57] R. V. Reddy, D. Murali, and J. Rajeshwar, “Context-aware middleware
architecture for iot-based smart healthcare applications,” in Innovations
in Computer Science and Engineering. Springer, 2019, pp. 557–567.
[58] M. Loske, L. Rothe, and D. G. Gertler, “Context-aware authentication:
State-of-the-art evaluation and adaption to the iiot,” in 2019 IEEE 5th
World Forum on Internet of Things (WF-IoT). IEEE, 2019, pp. 64–69.
[59] F. Kausar, W. Aman, and D. Al-Abri, “Mutually authenticated group key
management protocol for healthcare iot networks,” in Proceedings of the
Future Technologies Conference. Springer, 2019, pp. 1–12.
[60] S. D. Anton, D. Fraunholz, C. Lipps, K. Alam, and H. D. Schotten,
“Putting things in context: Securing industrial authentication with context
information,” arXiv preprint arXiv:1905.12239, 2019.
8
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
[61] M. Trnka, F. Rysavy, T. Cerny, and N. Stickney, “Using wi-fi enabled
internet of things devices for context-aware authentication,” in Interna-
tional Conference on Information Science and Applications. Springer,
2018, pp. 635–642.
[62] M. Smith-Creasey, F. Albalooshi, and M. Rajarajan, “Context awareness
for improved continuous face authentication on mobile devices,” in 2018
IEEE 16th Intl Conf on Dependable, Autonomic and Secure Computing,
16th Intl Conf on Pervasive Intelligence and Computing, 4th Intl Conf on
Big Data Intelligence and Computing and Cyber Science and Technology
Congress (DASC/PiCom/DataCom/CyberSciTech). IEEE, 2018, pp.
644–652.
[63] K. Habib and W. Leister, “Context-aware authentication for the internet
of things,” in Elev. Int. Conf. Auton. Auton. Syst. fined, 2015, pp. 134–
139.
[64] C. Doukas and I. Maglogiannis, “Bringing iot and cloud computing
towards pervasive healthcare,” in Innovative Mobile and Internet Services
in Ubiquitous Computing (IMIS), 2012 Sixth International Conference
on. IEEE, 2012, pp. 922–926.
[65] H. Zhang, Z. Wang, C. Scherb, C. Marxer, J. Burke, L. Zhang, and
C. Tschudin, “Sharing mhealth data via named data networking,” in
Proceedings of the 3rd ACM Conference on Information-Centric Net-
working. ACM, 2016, pp. 142–147.
[66] A. Garcia-de Prado, G. Ortiz, J. Boubeta-Puig, and D. Corral-Plaza,
“Air4people: a smart air quality monitoring and context-aware notifica-
tion system,” Journal of Universal Computer Science, vol. 24, no. 7, pp.
846–863, 2018.
[67] R. Dobrescu, D. Merezeanu, and S. Mocanu, “Context-aware control
and monitoring system with iot and cloud support,” Computers and
Electronics in Agriculture, vol. 160, pp. 91–99, 2019.
[68] E. Ba¸stu˘
g, M. Bennis, and M. Debbah, “Proactive caching in 5g small cell
networks,” in Towards 5G: Applications, Requirements and Candidate
Technologies. Wiley Online Library, 2016, pp. 78–98.
[69] E. Zeydan, E. Bastug, M. Bennis, M. A. Kader, I. A. Karatepe, A. S. Er,
and M. Debbah, “Big data caching for networking: Moving from cloud to
edge,” IEEE Communications Magazine, vol. 54, no. 9, pp. 36–42, 2016.
[70] X. Zhao, P. Yuan, S. Tang et al., “Collaborative edge caching in context-
aware device-to-device networks,” IEEE Transactions on Vehicular Tech-
nology, vol. 67, no. 10, pp. 9583–9596, 2018.
[71] J. Lee and J. Lee, “Hierarchical mobile edge computing architecture
based on context awareness,” Applied Sciences, vol. 8, no. 7, p. 1160,
2018.
[72] B. Afzal, S. A. Alvi, G. A. Shah, and W. Mahmood, “Energy efficient
context aware traffic scheduling for iot applications,” Ad Hoc Networks,
vol. 62, pp. 101–115, 2017.
[73] H. Flores, V. Kostakos, S. Tarkoma, P. Hui, and Y. Li, “Evidence-aware
mobile cloud architectures,” in Mobile Big Data. Springer, 2018, pp.
65–84.
[74] L. Tang and S. He, “Multi-user computation offloading in mobile edge
computing: A behavioral perspective,” IEEE Network, vol. 32, no. 1, pp.
48–53, 2018.
[75] A. Dhumane, S. Guja, S. Deo, and R. Prasad, “Context awareness in
iot routing,” in 2018 Fourth International Conference on Computing
Communication Control and Automation (ICCUBEA). IEEE, 2018,
pp. 1–5.
[76] B. Guo, D. Zhang, Z. Wang, Z. Yu, and X. Zhou, “Opportunistic iot:
Exploring the harmonious interaction between human and the internet of
things,” Journal of Network and Computer Applications, vol. 36, no. 6,
pp. 1531–1539, 2013.
[77] M. Musolesi and C. Mascolo, “Car: Context-aware adaptive routing for
delay-tolerant mobile networks,” IEEE Transactions on Mobile Comput-
ing, vol. 8, no. 2, pp. 246–260, 2009.
[78] A. Dhumane, R. Prasad, and J. Prasad, “Routing issues in internet of
things: a survey,” in Proceedings of the international multiconference of
engineers and computer scientists, vol. 1, 2016, pp. 16–18.
[79] T. X. Tran, A. Hajisami, P. Pandey, and D. Pompili, “Collaborative
mobile edge computing in 5g networks: New paradigms, scenarios, and
challenges,” arXiv preprint arXiv:1612.03184, 2016.
[80] X.-Q. Pham, T.-D. Nguyen, V. Nguyen, and E.-N. Huh, “Joint node
selection and resource allocation for task offloading in scalable vehicle-
assisted multi-access edge computing,” Symmetry, vol. 11, no. 1, p. 58,
2019.
[81] V. Tripathi, “Adaptive computation offloading in mobile cloud comput-
ing.” in CLOSER, 2017, pp. 524–529.
[82] K. Zhang, Y. Zhu, S. Leng, Y. He, S. Maharjan, and Y. Zhang, “Deep
learning empowered task offloading for mobile edge computing in urban
informatics,” IEEE Internet of Things Journal, vol. 6, no. 5, pp. 7635–
7647, 2019.
[83] C. Tu, H. Liu, Z. Liu, and M. Sun, “Cane: Context-aware network
embedding for relation modeling,” in Proceedings of the 55th Annual
Meeting of the Association for Computational Linguistics (Volume 1:
Long Papers), vol. 1, 2017, pp. 1722–1731.
[84] X. Hu, X. Li, E. Ngai, V. Leung, and P. Kruchten, “Multidimensional
context-aware social network architecture for mobile crowdsensing,”
IEEE Communications Magazine, vol. 52, no. 6, pp. 78–87, 2014.
[85] M. Alirezaie, J. Renoux, U. Köckemann, A. Kristoffersson, L. Karlsson,
E. Blomqvist, N. Tsiftes, T. Voigt, and A. Loutfi, “An ontology-based
context-aware system for smart homes: E-care@ home,” Sensors, vol. 17,
no. 7, p. 1586, 2017.
[86] E. Mingozzi, G. Tanganelli, C. Vallati, B. Martínez, I. Mendia, and
M. Gonzalez-Rodriguez, “Semantic-based context modeling for quality
of service support in iot platforms,” in 2016 IEEE 17th International
Symposium on A World of Wireless, Mobile and Multimedia Networks
(WoWMoM). IEEE, 2016, pp. 1–6.
[87] G. Fortino, W. Russo, C. Savaglio, M. Viroli, and M. Zhou, “Modeling
opportunistic iot services in open iot ecosystems.” in WOA, 2017, pp.
90–95.
[88] R. Casadei, G. Fortino, D. Pianini, W. Russo, C. Savaglio, and M. Viroli,
“A development approach for collective opportunistic edge-of-things
services,” Information Sciences, vol. 498, pp. 154–169, 2019.
[89] G. Fortino, C. Savaglio, C. E. Palau, J. S. de Puga, M. Ganzha, M. Pa-
przycki, M. Montesinos, A. Liotta, and M. Llop, “Towards multi-layer
interoperability of heterogeneous iot platforms: The inter-iot approach,”
in Integration, interconnection, and interoperability of IoT systems.
Springer, 2018, pp. 199–232.
[90] G. Aloi, G. Caliciuri, G. Fortino, R. Gravina, P. Pace, W. Russo,
and C. Savaglio, “Enabling iot interoperability through opportunistic
smartphone-based mobile gateways,” Journal of Network and Computer
Applications, vol. 81, pp. 74–84, 2017.
[91] R. Casadei, G. Fortino, D. Pianini, W. Russo, C. Savaglio, and M. Viroli,
“Modelling and simulation of opportunistic iot services with aggregate
computing,” Future Generation Computer Systems, vol. 91, pp. 252–262,
2019.
[92] Z. Chang, L. Lei, Z. Zhou, S. Mao, and T. Ristaniemi, “Learn to cache:
Machine learning for network edge caching in the big data era,” IEEE
Wireless Communications, vol. 25, no. 3, pp. 28–35, 2018.
[93] R. Pascoal, A. D. Almeida, and R. C. Sofia, “Mobile pervasive augmented
reality systems—mpars: The role of user preferences in the perceived
quality of experience in outdoor applications,” ACM Transactions on
Internet Technology (TOIT), vol. 20, no. 1, pp. 1–17, 2020.
[94] R. Wattenhofer and A. Zollinger, “Xtc: A practical topology control algo-
rithm for ad-hoc networks,” in 18th International Parallel and Distributed
Processing Symposium, 2004. Proceedings. IEEE, 2004, p. 216.
[95] S. Maheshwari, D. Raychaudhuri, I. Seskar, and F. Bronzino, “Scal-
ability and performance evaluation of edge cloud systems for latency
constrained applications,” in 2018 IEEE/ACM Symposium on Edge
Computing (SEC). IEEE, 2018, pp. 286–299.
[96] J. Luo, J. Hu, D. Wu, and R. Li, “Opportunistic routing algorithm for
relay node selection in wireless sensor networks,” IEEE Transactions on
Industrial Informatics, vol. 11, no. 1, pp. 112–121, 2015.
[97] G. Han, D. Choi, and W. Lim, “Reference node placement and selection
algorithm based on trilateration for indoor sensor networks,” Wireless
Communications and Mobile Computing, vol. 9, no. 8, pp. 1017–1027,
2009.
[98] F. Zou, Z. Zhang, and W. Wu, “Latency-bounded minimum influential
node selection in social networks,” in International Conference on Wire-
less Algorithms, Systems, and Applications. Springer, 2009, pp. 519–
526.
[99] J. Subhlok, P. Lieu, and B. Lowekamp, “Automatic node selection for
high performance applications on networks,” ACM SIGPLAN Notices,
vol. 34, no. 8, pp. 163–172, 1999.
[100] T. Zhao, S. Zhou, X. Guo, Y. Zhao, and Z. Niu, “A cooperative scheduling
scheme of local cloud and internet cloud for delay-aware mobile cloud
computing,” in 2015 IEEE Globecom Workshops (GC Wkshps). IEEE,
2015, pp. 1–6.
[101] Y. Wang and L. Qiu, “Edge selection-based low complexity detection
scheme for scma system,” in 2016 IEEE 84th Vehicular Technology
Conference (VTC-Fall). IEEE, 2016, pp. 1–5.
9
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
Author et al.: Preparation of Papers for IEEE TRANSACTIONS and JOURNALS
10
This is the author's pre-print version. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising, promotion or for creating new
collective works for resale or for redistribution to thirds must be obtained from the
camera-ready owner. The camera-ready version of this work has been published by IEEE Access, 2020,
and the camera-ready version is property of IEEE.
