See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/328340296

Interoperability in Internet of Things Infrastructure: Classification,

Challenges, and Future Work: Third International Conference, IoTaaS 2017,
Taichung, Taiwan, September 20–22,...

Chapter in Lecture Notes of the Institute for Computer Sciences · January 2018
DOI: 10.1007/978-3-030-00410-1_2

CITATIONS READS

27 3,696

3 authors:

Mahda Noura Mohammed Atiquzzaman

Siemens University of Oklahoma
27 PUBLICATIONS 941 CITATIONS 585 PUBLICATIONS 13,513 CITATIONS

SEE PROFILE SEE PROFILE

Martin Gaedke
Chemnitz University of Technology
285 PUBLICATIONS 3,390 CITATIONS

SEE PROFILE

All content following this page was uploaded by Mahda Noura on 26 October 2018.

The user has requested enhancement of the downloaded file.

 Interoperability in Internet of Things Infrastructure:
Classification, Challenges, and Future Work

Mahda Noura1, Martin Gaedke1, Mohammed Atiquzzaman2

1Technische Universität Chemnitz, Chemnitz, Germany
{mahda.noura,martin.gaedke}@informatik.tu-chemnitz.de
2 University of Oklahoma, Norman, OK-73109

atiq@ou.edu

Abstract. The Internet of Things (IoT) is an important research area, and sub-
stantial developments for a wide range of devices and IoT platforms is evident.
However, one of the critical issues in IoT is that the different proprietary IoT
platforms and systems are still not interoperable; unable to talk with each other.
In this paper, we survey the state-of-the-art on interoperability in IoT. First, we
provide a classification of techniques and schemes looking at IoT interoperability
from different perspectives. For each category, we present the approaches pro-
posed in the papers. Second, we use the interoperability classification as a base-
line to compare some of the existing IoT research projects and identify gaps in
the existing solutions. Our findings will help domain experts and professionals
to get an overview and categorization of existing interoperability solutions in IoT
and select an appropriate approach to help increase the number of interoperable
IoT products.

Keywords: Fragmentation, Internet of Things. Web of Things. Interoperability.

IoT Platforms.

1 Introduction
In the past decade, an abundance of IoT devices and platforms have been integrated
into a wide range of applications like the market, healthcare, agriculture, utilities, energy,
transportation, industrial control, and buildings, etc. Numerous studies forecast the
substantial development of the IoT in the coming years. E.g., International Data Center
(IDC) predicts that by 2020 the IoT solutions market will grow to $7.1 trillion [1], which
will include 50 billion Internet-connected devices [2]. The European project Unify-IoT1,
lately identified that there are more than 300 IoT platforms in the current market.
Those studies are encouraging, since they suggest a tremendous impact of the IoT
over the coming years. However, a new McKinsey analysis [3] points out a substantial
threat to the predicted economic value: missing interoperability. Particularly, the authors
state that 40% of the potential benefits of IoT can be obtained with the interoperability
between IoT systems, i.e. two or more dissimilar systems are able to work together.

1 http://unify-iot.eu
 The current IoT market is fragmented due to the extreme degree of heterogeneity in
terms of device protocols, controllers, network connectivity methods, application
protocols, standards, data formats and so on. The absence of interoperability in IoT is
due to a lack of standardisation [4][5]. Vendors are intentionally defining different IoT
platforms, proprietary protocols and interfaces which are incompatible with other
solutions. Therefore, these vendors create different verticals and mostly closed
ecosystems, which are sometimes called stove pipes or silos. To be precise, the
components in one silo cannot talk to the components in another silo. For example,
currently, before customers can access different IoT things they generally need a
dedicated application for that particular thing preloaded onto the smartphone, such as
the Philips Hue or the Belkin WeMo switch. This way the customer will have many
devices, each with their own application, that work independently of each other. Also,
there are data interoperability issues when developers want to create an innovative IoT
application exploiting resources from different IoT applications and or/services (such as
Oral-B or the Apple HealthKit) in heterogeneous domains (e.g., smart health, smart
home, etc.). These issues ultimately lead to vendor lock-in of end-users.
Considering the importance of interoperability in IoT, first we need to understand
interoperability and the existing solutions to analyze what is needed and identify the
platforms that are ahead to help increase the number of interoperable IoT products. A
classification of IoT interoperability is provided in Section 2. Then, based on the
classification, a survey of the existing H2020 IoT research projects is presented in
Section 3. Finally, the paper concludes in Section 4.

2 Interoperability Classification in IoT

Interoperability is a major topic in many different domains and there are several dis-
tinct definitions of this term in the literature. Between the diverse definitions for in-
teroperability, we quote the most noteworthy one in our context. The IEEE defines in-
teroperability as “the ability of two or more systems or components to exchange infor-
mation and to use the information that has been exchanged” [6]. According to this def-
inition, there are many scientific challenges: the ability to get the data, to exchange
information, and the ability to use the information once it has been received.
Standard organizations and open source communities have been working to address
interoperability issues in different parts or levels. We divide the existing interoperabil-
ity solutions in the literature according to the level of interoperability that has been
achieved between IoT platforms or systems: device level, networking level, syntactic
level, semantic level, cross-platform level, and cross-domain interoperability. The cat-
egories are described in the following subsections.

Device Level Interoperability

Various communication technologies such as: WiFi, 3G/4G, ANT+, ZigBee have
emerged since only one wireless technology cannot support the different requirements
of IoT markets. However, in the absence of a de-facto communication standard(s), not
all smart devices implement all these communication technologies. Device Level in-
teroperability refers to enabling the integration of such heterogeneous communication
technologies and standards supported by different IoT devices. This layer should focus
on accessing devices through unifying interfaces and the ability to integrate new de-
vices into any IoT platform. For example, consider a smart home scenario where the
light bulbs and thermostats use ZigBee, speakers communicate with Bluetooth, and
switches communicate through WiF. Interoperability in this example enables different
devices to understand and translate between these disparate communication technolo-
gies. An ideal IoT platform would offer a pool of standardized communication proto-
cols where the device manufacturers may select the appropriate protocols (e.g. CoAP
for constrained devices). In the literature device level interoperability relies either on a
gateway solution (sometimes called protocol converters) that can be extended using
plug-ins, to support new communication protocols or by instructing the device vendors
to only use the protocols that are supported (such as Fosstrak2). For example, the Apple
HomeKit 3 , If-This-Then-That (IFTTT) 4 and Eclipse Ponte 5 , Light-weight M2M 6
(LWM2M) are some of the gateway solutions in the literature.

2.1 Network Level Interoperability

Network level interoperability deals with mechanisms to exchange messages between

systems through different networks (networks of networks) to provide end-to-end com-
munication. To make systems interoperable, each system should be able to exchange
messages with other systems through various types of networks. In this level, protocol
interoperability is the main focus. At the standardization level, the IETF has developed
a set of standards for routing including RPL, CORPL, and CARP and solutions for
encapsulation including 6LowPAN, 6TiSCH, 6Lo, and Thread [7]. In addition, the
cloud has been used as a medium to address interoperability at this level. This is called
Fog of Things [8], where the computing, storage and networking services are placed at
the edge of the network rather than centralized cloud servers. Fog of Things aims for
providing value to the data before making it available on the web facilitating the in-
teroperability of the devices at the edge and preparing the managed data for further
applications to be interoperable. Another new solution to address interoperability in
this level is software-based approaches such as Software Defined Networking (SDN)
which hides all the control and management operations from the IoT devices by setting
them inside a middleware layer [9], which alleviates the dependency from vendors.

2 https://fosstrak.github.io/
3 www.apple.com/ios/home
4 https://ifttt.com
5 http://www.eclipse.org/proposals/technology.ponte/
6
http://technical.openmobilealliance.org/Technical/technical-information/omna/lightweight-
m2m-lwm2m-object-registry
2.2 Syntactic Level Interoperability

Syntactic level interoperability refers to interoperation of the format as well as the data
structure used in any exchanged information or service between heterogeneous IoT sys-
tem entities. This level of interoperability is important to enable smooth message tran-
sition between different IoT systems. Web technologies such as HTTP, JSON, REST
and SOAP architecture of the World Wide Web, an approach referred to as the Web of
Things (WoT) is proposed to provide greater interoperability. The WoT enables devel-
opers to connect things using web technologies and tools to create new applications and
mashups. The use of the web provides a one-for-all solution for providing higher degree
of interoperability, since there is no need to install/develop specific software and drivers
for various devices, enabling the connection of heterogenous devices in dissimilar do-
mains. The Web supports different content types which resolve the challenge of work-
ing with different data formats in different applications across multiple platforms. Some
of the most common web-based representations of the resources are plain text, JSON,
XML and EXI. XML helps achieve syntactic interoperability by encoding syntactic
information into XML documents, providing platform and language independence,
vendor neutrality, and extensibility, which are all crucial to interoperability. In addition,
JSON is becoming popular in the IoT market, as it is lightweight, simple and offers
capabilities close to the XML ones without requiring the overhead (e.g. schema) and
processing requirements of XML. Also, the Sensor Web Enablement7 (SWE) frame-
work provide a standard set of web service interfaces towards making it easier to share
sensor data. Moreover, there are many efforts for IoT/Cloud convergence [10], and
several IoT cloud-enabled platforms (ThingWorx 8 , OpenIoT 9 , Xively 10 , and
ThingsSpeak11) are available at the syntactic level to facilitate the aggregation of data
and services from heterogeneous IoT devices.

2.3 Semantic Level Interoperability

Semantic level interoperability deals with the technologies needed for enabling the
meaning of information to be shared by communicating parties. To enable building new
innovative, applications which make use of data from multiple existing vertical IoT
silos these systems must not only be able to exchange information but also have a com-
mon understanding of the meaning of this data. This level is concerned with data and
information models which will describe: the things, application functionalities, data
modeling and service descriptions, in a uniform way to enable machines to read and
understand the data sent and received. For example, consider two smart lightening de-
ployments, which have been planned and implemented independently. There is a need

7 www.opengeospatial.org/ogc/markets-technologies/swe
8 http://www.thingworx.com
9 http://www.openiot.eu
10 https://xively.com
11 http://thingspeak.com
to combine both deployments to calculate the amount of energy gains reached. This is
challenging because each deployment speaks diverse languages at the data level. They
have different data formats as well as different semantics, such as units of measurement,
sensor types and features, mathematic constructs and so on. The technologies from the
Semantic Web have been used to address interoperability in this level. Ontologies are
used to define a common, machine-readable dictionary that is able to express resources,
services, APIs and related parameters (such as Semantic Sensor Network, IoT-Lite, and
Architectural Reference Model). Other semantic web techniques such as Resource De-
scription Framework (RDF), RDF Schema (RDFS), and Web Ontology Language
(OWL), Linked Data (LD) and SPARQL are used for representing web resources in a
uniform form and reasoning over them. In this level, there are issues such as: (1) ontol-
ogy heterogeneity (e.g., ontology designed by different persons differ in the structure),
terms used to describe data (e.g., t, temp and temperature are several terms to describe
temperature), and the meaning of data exchanged according to the context (e.g., body
temperature differs from room temperature). Semantic interoperability can be achieved
through agreed-upon information models of the terms used as part of the interfaces and
exchanged data. Moreover, catalog based approaches such as HyperCat12, allows dis-
tributed data repositories to be used jointly by applications.

2.4 Cross-Platform Interoperability

The Cross-Platform interoperability is the main requirement to have an interoperable

IoT system. This interoperability level enables federation across different IoT platforms
by integrating data from various platforms specific to one vertical domain such as smart
home, smart healthcare, smart garden, etc. For example, assume that a user wants to
use a single application to manage the smart lighting at home and in the office. Cur-
rently, two different applications are required; one for his home automation system, and
the other for the office environment. The cross-platform interoperability level allows
managing devices at both home, in the office, and other place.

2.5 Cross-Domain Interoperability

Cross-Platform solutions focus on specific activities that are limited to one domain. The
Cross-domain interoperability enables the federation of different platforms within
heterogeneous domains to build horizontal IoT applications. This federation will not
try to mandate a specific protocol at any levels of the protocol stack as the only standard
across domains. In contrast, it is essential that IoT platforms can choose the desired
protocols to control the end-to-end communications and data exchange (from sensors
to gateways to cloud-based platforms) based on their requirement and purpose. In the
literature, some IoT solution providers wrap and offer their domain-specific platforms
in a ‘Sensing as a Service’ way [11], which provides third parties useful information

12 www.hypercat.io
 with respect to a single domain. For example, a smart home platform can provide do-
main-specific enablers such as air temperature and the lighting conditions. These ena-
blers can then be exploited by other IoT platforms, such as smart healthcare, to provide
more innovative applications and scenarios.

3 Analysis of Current IoT Interoperability Platforms

To assess the maturity of IoT interoperability, we determine the features discussed in

Section 2 that are supported by state-of-the-art IoT platforms. We analysed some of the
recent H2020 European research projects as shown in Table 1. These projects are de-
veloping interoperability solutions at different interoperability levels. In the following,
we discuss the mappings of the interoperability levels and the method and solutions
provided by the projects. In addition, we discuss some shortcomings.

Table 1. A summary of the IoT platforms supporting interoperability requirements. = sup-

ported; = not supported
TagIt Smart!

FIESTA-IoT
UniversAAL

VICINITY
SymbIoTe

Open-IoT
BUTLER

FIWARE

Inter-IoT
RERUM
bIoTope
Big IoT

AGILE

VITAL

iCore
Device Level               
Network Level               
Syntactic               
Level
Semantic               
Level
Cross-               
Platform Level
Cross-Domain               
Level

3.1 Interoperability among IoT platforms

TagItSmart!13 offers a set of tools and enabling technologies integrated into a plat-
form with open interfaces to make mass-market products connected using smart printed
QR codes, smartphone, and cloud. However, interoperability support is limited to the
device level in this project. Similarly, the AGILE project focuses on the integration of
heterogeneous devices by build a modular IoT gateway, which provides RESTful APIs
to interact with user devices. The configuration of the gateway is performed automati-

13 http://tagitsmart.eu
cally based on the hardware configuration, reducing the gateway setup time. The bIo-
Tope14 provides a platform that enables stakeholders to create new IoT systems and to
rapidly harness available information using Systems-of-Systems (SoS) capabilities for
connected smart objects by providing standardised open APIs for the interoperability
between smart objects of different platforms. Two Open API standards are mentioned
Open Messaging Interface and Open Data Format. Different from other projects, the
SymbIoTe15 provides a middleware which focuses on the federation of IoT platforms.
Syntactic interoperability is addressed by a high-level API which acts like an adapter
to provide a uniform access to resources of all platforms. Semantic interoperability is
addressed by semantic mapping between the platform-specific information models,
where platform-specific extension of one platform is translated into the platform-spe-
cific exaction of the other platform. Similar to SymbIoTe, the Big-IoT16 project focuses
on the federation between IoT platforms, developing a generic, unified Web API for
smart object platforms focusing on syntactic and semantic interoperability enabling ap-
plication developers to interact with different IoT platforms. Vital17 provides syntactic
interoperability using SOA and enables RESTful web services for communication in-
terchange mechanism, and semantic interoperability is achieved by using a common-
data model using Linked Data standards such as RDF (for modelling and accessing
metadata and data), JSON-LD, and ontologies. Vital also aims to integrate different
IoT platform, but it doesn’t address cross-domain mechanisms and is limited to smart
city domain. Unlike BigIoT, Vital stores the data coming from IoT systems. The
VICINITY18 platform supports semantic interoperability (building on LinkSmart/Hy-
dra [12]) and the use of existing ontologies (e.g. from Ready4SmartCities 19 ,
oneM2M20) to provide “interoperability as a service”. The openIoT project focuses on
an open source middleware for creating real-time IoT services on demand. However, it
does not address cross-platform and cross-domain mechanisms. The Inter-IoT21 aims
to provide an interoperable and open IoT framework for the integration of heterogenous
IoT platforms with the consideration of cross-domain interoperability. Unlike the other
existing projects, this project considers interoperability at all the mentioned levels. The
FIESTA-IoT22 project is considering the semantic interoperability of testbeds regard-
less of the application domain.

3.2 Interoperability analysis results

From the analysis of the approaches taken by different projects shown in Table 1, it is
clear that most of the projects address two to five interoperability levels and their focus

14 www.biotope-project.eu
15 http://iot-epi.eu/project/symbiote
16 http://big-iot.eu
17 http://vital-iot-eu
18 http://vicinity2020.eu/vicinity
19 http://www.ready4smartcities.eu
20 www.onem2m.org
21 www.inter-iot-projects.eu
22 http://fiesta-iot.eu
is providing interoperability solutions to connect existing IoT commercial and open
source platforms. It is also clear that there are several efforts towards solving the in-
teroperability issue within the application and data and semantic layer. This is because
interoperability at the application level is still not mature since the existing solutions
lack information models and have a strong relationship with the underlying communi-
cation architecture (RPC or RESTful design). In addition, many of the projects propos-
ing semantic-based components are not interoperable with each other. For instance, the
existing projects don’t use the same data model to structure the data produced by smart
objects or the same reasoning approach to deduce new knowledge from data produced
by smart devices. Moreover, current implementations focus on specific IoT application
domains neglecting cross-domain interoperability.
To allow the development of applications on top of IoT platforms, the IoT platforms
should provide the developers an APIs to their functionality. Further, to enable an effi-
cient development of cross-IoT platform applications, these APIs should be uniform
across the platforms to the extent possible. Today’s IoT platforms almost all provide a
public REST API to access the services. The APIs are usually based on RESTful prin-
ciples; however, most platforms use custom REST APIs and data models which com-
plicates the mashing up of data across multiple platforms. From our results, using stand-
ards such as HyperCat23 should be adopted to address such issues.

4 Conclusion

In this paper, we have answered two questions: what are the different categories for
an interoperable IoT ecosystem and how interoperability has been addressed in the lit-
erature. At the device level gateways and smartphone solutions are the main method to
address the connectivity issues. In the networking level, IPv6 and other standard tech-
nologies such as SDN, NFV and Fog are promising. From the Syntactic and Semantic
perspectives, web technologies (open APIs, RESTful web services, JSON-like diction-
ary, and mashups) and semantic web technologies provide a high degree of interopera-
bility. Finally, interoperability at the higher levels (cross-platform and cross-domain)
can be achieved by the collaboration and agreement between IoT platform owners on
many essential issues such as exposing the resources, interfaces, services, and data
models. The main results of our research are that we believe that there is not likely a
common set of standards that will be universally accepted which will allow IoT devices
and platforms to work together. However, by applying some of the presented techniques
interoperability can be improved.

References
1. D. Lund and M. Morales, “Worldwide and Regional Internet of Things ( IoT ) 2014 – 2020
Forecast : A Virtuous Circle of Proven Value and Demand,” Int. Data Corp. (IDC),
Tech.Rep, 2014.

23 www.hypercat.io
 2. D. Evans, “The Internet of Things: How the Next Evolution of the Internet is Changing
Everything,” CISCO white Pap., vol. 1, no. 2011, pp. 1–11, 2011.
3. J. Manyika, M. Chui, P. Bisson, J. Woetzel, R. Dobbs, J. Bughin, and D. Aharon, “The
Internet of Things: Mapping the value beyond the hype. McKinsey Global Institute,”
McKinsey Glob. Inst., p. 3, 2015.
4. L. Da Xu, W. He, and S. Li, “Internet of things in industries: A survey,” IEEE Trans. Ind.
informatics, vol. 10, no. 4, pp. 2233–2243, 2014.
5. S. Agrawal and M. L. Das, “Internet of Things- A paradigm shift of future Internet
applications,” in Engineering (NUiCONE), 2011 Nirma University International
Conference on Engineering, 2011, pp. 1–7.
6. J. Radatz, A. Geraci, and F. Katki, “IEEE standard glossary of software engineering
terminology,” IEEE Std, vol. 610121990, no. 121990, p. 3, 1990.
7. T. Salman and R. Jain, “Networking Protocols for Internet of Things,” pp. 1–28, 2013.
8. A. Workshops and M. Serrano, “SOFT-IoT : Self-Organizing FOG of Things,” 2016.
9. Y. Jararweh, M. Al-Ayyoub, A. Darabseh, E. Benkhelifa, M. Vouk, and A. Rindos, “SDIoT:
a software defined based internet of things framework,” J. Ambient Intell. Humaniz.
Comput., vol. 6, no. 4, pp. 453–461, 2015.
10. A. Alamri, W. S. Ansari, M. M. Hassan, M. S. Hossain, A. Alelaiwi, and M. A. Hossain, “A
survey on sensor-cloud: architecture, applications, and approaches,” Int. J. Distrib. Sens.
Networks, vol. 9, no. 2, p. 917923, 2013.
11. J. Soldatos, N. Kefalakis, M. Serrano, and M. Hauswirth, “Design principles for utility-
driven services and cloud-based computing modelling for the Internet of Things,” Int. J.
Web Grid Serv. 6, vol. 10, no. 2–3, pp. 139–167, 2014.
12. M. Eisenhauer, P. Rosengren, and P. Antolin, “Hydra: A development platform for
integrating wireless devices and sensors into ambient intelligence systems,” in The Internet
of Things, Springer, 2010, pp. 367–373.

View publication stats