22
The IoT Architectural Framework, Design Issues and Application Domains Gordana Gardas ˇevic ´ 1 Mladen Veletic ´ 1 Nebojs ˇa Maletic ´ 2 Dragan Vasiljevic ´ 1 Igor Radusinovic ´ 3 Slavica Tomovic ´ 3 Milutin Radonjic ´ 3 Published online: 25 October 2016 Ó Springer Science+Business Media New York 2016 Abstract The challenge raised by the introduction of Internet of Things (IoT) concept will permanently shape the networking and communications landscape and will therefore have a significant social impact. The ongoing IoT research activities are directed towards the definition and design of open architectures and standards, but there are still many issues requiring a global consensus before the final deployment. The paper presents and discusses the IoT architectural frameworks proposed under the ongoing standardization efforts, design issues in terms of IoT hardware and software components, as well as the IoT application domain representatives, such as smart cities, healthcare, agriculture, and nano- scale applications (addressed within the concept of Internet of Nano-Things). In order to obtain the performances related to recently proposed protocols for emerging Industrial Internet of Things applications, the preliminary results for Message Queuing Telemetry & Gordana Gardas ˇevic ´ [email protected] Mladen Veletic ´ [email protected] Nebojs ˇa Maletic ´ [email protected] Dragan Vasiljevic ´ [email protected] Igor Radusinovic ´ [email protected] Slavica Tomovic ´ [email protected] Milutin Radonjic ´ [email protected] 1 Faculty of Electrical Engineering, University of Banja Luka, Banja Luka, Bosnia and Herzegovina 2 IHP, Im Technologiepark 25, 15236 Frankfurt (Oder), Germany 3 Faculty of Electrical Engineering, University of Montenegro, Podgorica, Montenegro 123 Wireless Pers Commun (2017) 92:127–148 DOI 10.1007/s11277-016-3842-3

The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

  • Upload
    others

  • View
    5

  • Download
    0

Embed Size (px)

Citation preview

Page 1: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

The IoT Architectural Framework, Design Issuesand Application Domains

Gordana Gardasevic1 • Mladen Veletic1 • Nebojsa Maletic2 •

Dragan Vasiljevic1 • Igor Radusinovic3 • Slavica Tomovic3 •

Milutin Radonjic3

Published online: 25 October 2016� Springer Science+Business Media New York 2016

Abstract The challenge raised by the introduction of Internet of Things (IoT) concept will

permanently shape the networking and communications landscape and will therefore have

a significant social impact. The ongoing IoT research activities are directed towards the

definition and design of open architectures and standards, but there are still many issues

requiring a global consensus before the final deployment. The paper presents and discusses

the IoT architectural frameworks proposed under the ongoing standardization efforts,

design issues in terms of IoT hardware and software components, as well as the IoT

application domain representatives, such as smart cities, healthcare, agriculture, and nano-

scale applications (addressed within the concept of Internet of Nano-Things). In order to

obtain the performances related to recently proposed protocols for emerging Industrial

Internet of Things applications, the preliminary results for Message Queuing Telemetry

& Gordana [email protected]

Mladen [email protected]

Nebojsa [email protected]

Dragan [email protected]

Igor [email protected]

Slavica [email protected]

Milutin [email protected]

1 Faculty of Electrical Engineering, University of Banja Luka, Banja Luka, Bosnia and Herzegovina

2 IHP, Im Technologiepark 25, 15236 Frankfurt (Oder), Germany

3 Faculty of Electrical Engineering, University of Montenegro, Podgorica, Montenegro

123

Wireless Pers Commun (2017) 92:127–148DOI 10.1007/s11277-016-3842-3

Page 2: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

Transport and Time-Slotted Channel Hopping protocols are provided. The testing was

performed on OpenMote hardware platform and two IoT operating systems: Contiki and

OpenWSN.

Keywords Internet of Things (IoT) � Architectural framework � Applications � IndustrialIoT � MQTT � TSCH � OpenMote platform

1 Introduction

The Internet of Things (IoT) has been denoted as a new wave of information and com-

munication technology (ICT) advancements. The IoT is a multidisciplinary concept that

encompasses a wide range of various technologies, device capabilities, application

domains, different business and operational strategies, industry and market profiles, etc.

[1–3]. ITU-T Study Group 13 has defined the IoT as: ‘‘A global infrastructure for the

information society, enabling advanced services by interconnecting (physical and virtual)

things based on existing and evolving interoperable information and communication

technologies’’ [4]. This definition implies the significance of IoT vision and its impact on

future social, networking and communications landscape.

The number of IoT applications and scenarios is growing day by day and is influencing

our everyday life. They can be categorized into application domains, such as smart cities

and environments, smart grids, industrial automation, traffic management and logistics,

remote monitoring, healthcare and assisted living, agriculture and breeding, public safety,

and others. Recently, the concept of Internet of Nano-Things (IoNT) has been proposed in

order to provide the support for interconnecting the nano-scale devices to IP networks.

Therefore, we can think of IoT as a fuzzy and multi-tiered heterogeneous system based on

open architectural platforms and standards.

Smart objects or things are physical devices (i.e., sensors, actuators, RFID tags, mobile

phones, etc.) or virtual entities with capabilities to cooperate and interact within the IoT

infrastructure. The IETF gives the following definition of term ‘‘things’’: ‘‘In the vision of

IoT, ‘things’ are very diverse, such as computers, sensors, people, actuators, refrigerators,

TVs, vehicles, mobile phones, clothes, food, medicines, books, etc. These ‘things’ should

be identified at least by one unique way of identification for the capability of addressing

and communicating with each other and verifying their identities. In here, if the ‘thing’ is

identified, we call it the ‘object’’’ [5].

The aforementioned nature of IoT indicates that a single reference architecture cannot

be applied for majority of scenarios. It is of a particular importance to build the IoT

ecosystem in a way to support the standardized IoT reference models and architectures.

Although there are plenty of different scenarios and solutions, several common charac-

teristics can be extracted: modularity and interoperability of IoT components, open models

and architectures, flexible service compositions, integrated security solutions, semantic

data integration, etc.

There is an intensified effort regarding development of architectural frameworks and

solutions, such as the IEEE model, the ITU-T model, the NIST model for Smart Grid, the

ETSI model for SmartM2M communications, the Architectural Reference Model from the

EU IoT-A project, as well as other related work, such as the IETF, W3C, OASIS, IPSO

128 G. Gardasevic et al.

123

Page 3: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

Alliance, etc. [6]. The IoT related research has been selected as a topic of prime importance

within the scope of Horizon 2020 priorities and EU 2020 Digital Agenda.

The wireless embedded devices require a new generation of communication protocols in

order to be fully integrated into Future Internet infrastructure. These constrained devices

form the Low-Power and Lossy Networks (LLNs) that are connected by wireless, unre-

liable and error-prone links with restricted frame-sizes and low throughput [7]. Therefore,

the design of energy-efficient protocols and mechanisms remains one of the key goals in

creating the reference IoT protocol stack.

The IoT Operating Systems (OSs) are designed to run on devices with constrained

resources and limited functionalities. Therefore, the middleware, data management and

application programming interfaces (APIs) represent some of key design elements in IoT

architecture. The actual research in this field is related to support for real-time analysis,

policy-based access control, user-data management, context-based monitoring and secu-

rity, integration support, and others.

The rest of this paper is structured as follows. Section 2 describes the ongoing stan-

dardization efforts and general IoT reference model. The prerequisites for IoT deployment

in terms of hardware and software resources are given in Sect. 3. Section 4 discusses some

of IoT application domain representatives, such as smart cities, healthcare, agriculture, and

nano-scale applications. In Sect. 5, we provide the preliminary results in testing the

OpenMote platform for Industrial Internet of Things (IIoT) applications. Finally, con-

cluding remarks are provided in Sect. 6.

2 IoT Architectural Framework

The adoption of IoT general reference model is of crucial importance in order to promote

the global IoT deployment, and proliferation of IoT applications and services. A number of

different architectural solutions have been proposed in the past few years, including not

only the international standard organisations (ITU-T, 3GPP, ETSI, ISO, IETF, IEEE, W3C,

oneM2M, OASIS, etc.), but also academia and research institutions, companies, stake-

holders, and civil societies. The industrial vision is also an important element in IoT

deployment.

The core requirements related to IoT protocol stack include low-power communication

protocols and IP-based, secure and reliable communication stack. The standardisation

bodies are addressing the issues related to interoperability of protocol stacks, the trans-

parency of open standards, interfaces, and architectures for IoT. This initiative leads

towards the creation of IoT-specific protocol stack, since there exists a plethora of wireless

protocols, such as ZigBee, IPv6 over Low-Power Wireless Personal Area Network

(6LoWPAN), WirelessHART, IEEE 802.15.4e, ISA100.11a, etc.

The essential IoT technology implemented at the physical layer is Wireless Sensor

Networks (WSNs), but other low-power technologies, such as Power Line Communica-

tions (PLC), Bluetooth Low Energy (BLE), Radio-Frequency IDentification (RFID),

Digital Enhanced Cordless Telecommunications (DECT), Ultra-Wide Bandwidth (UWB),

and Near Field Communication (NFC) are also important for the implementation of var-

ious IoT applications within specific environments and domains, as illustrated in Fig. 1 [8].

Recently, the 3GPP started the work towards standardizing the Narrow-Band IoT (NB-

IoT). The objective is to specify a radio access for cellular IoT, based on a E-UTRA variant

that addresses improved indoor coverage, support for massive number of low throughput

The IoT Architectural Framework, Design Issues and... 129

123

Page 4: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

devices, low delay sensitivity, ultra low device cost, low device power consumption and

optimised network architecture [9].

The fundamental aspect of IoT infrastructure is the support for the global Internet

Protocol (IP) connectivity. Considering that the IoT is recognized as a Future Internet

architecture, the 6LoWPAN protocol stack has been developed to enable the full IP inte-

gration [10]. There are also other solutions, such as the ZigBee IP specification that adds

the network and security layers, as well as an application framework to the IEEE 802.15.4

standard. It supports the energy-efficient wireless mesh networking based on standard IP

protocols, such as 6LoWPAN, IPv6, RPL, UDP, TCP, Transport Layer Security (TLS), and

Protocol for Carrying Authentication for Network Access (PANA) [11].

The mechanisms at network layer are implemented by the IPv6 protocol. A represen-

tative at this layer is the Routing over Low-power and Lossy Networks (RPL) protocol,

standardized by the IETF ROLL group [12]. The RPL is a link-independent distance vector

routing protocol designed specifically for LLNs, where the connections between nodes

(routers) are lossy and with low packet delivery rates. Such networks usually do not have a

predefined topology and may deploy a large number of nodes. Correspondingly, the role of

RPL is to discover optimal links based on an established set of criteria. This protocol

supports various traffic models, such as point-to-point, point-to-multipoint, and multipoint-

to-point. The IP Security (IPSec) protocol is enabled at network layer in order to support

the secure communications. The IPv6 also provides an efficient and robust mobility

mechanism as a core of Mobile IPv6 (MIPv6) protocol.

At transport layer, two standard protocols may be used: Transmission Control Protocol

(TCP) and User Datagram Protocol (UDP). The UDP is preferably used for 6LoWPAN due

to its lightweight implementation. The flow control mechanism in TCP generates the

Fig. 1 The overview of IoT layered architecture [8]

130 G. Gardasevic et al.

123

Page 5: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

overhead that is considered too high for LLNs. But, recently developed IoT hardware

platforms have enough memory and processing resources to support the transmission over

TCP.

An example of application layer protocol is the Constrained Application Protocol

(CoAP) developed for LLNs [13]. The CoAP is built over UDP and therefore has a

significantly lower overhead and support for multicast transmission. As a comparison, the

HTTP relies on TCP that does not have the multicast support and is rather sensitive to

mobility. The CoAP is a request/response protocol that utilizes both synchronous and

asynchronous responses. The ZigBee Smart Energy Protocol version 2.0 (SEP 2.0) offers a

global standard for IP-based control, both wired and wireless, for energy management in

Home Area Networks and industrial automation applications [14]. SEP 2.0 is an industry

open standard, based on open technologies such as REST/HTTP and XML. The connec-

tivity protocol for IoT and Machine-to-Machine (M2M) applications is the Message

Queuing Telemetry Transport (MQTT). This lightweight protocol is designed to connect

the physical devices and networks with applications and middleware used in IT and Web

development [15]. Unlike the CoAP, the MQTT is running over TCP. Additionally, this

protocol provides three levels of Quality of Service (QoS) support for message delivery.

More details about this protocol are provided in Sect. 5.

As already mentioned, the IETF 6LoWPAN protocol family is a key enabler of IP

connectivity for low-power and constrained devices. One possible implementation of

6LoWPAN protocol stack is shown in Fig. 2. The existence of adaptation layer actually

represents the new element in comparison to standard Open Systems Interconnection (OSI)

protocol stack. This adaptation layer enables functionalities, such as header compression

and decompression, fragmentation and reassembly, address auto-configuration, as well as

the support for mesh based routing. Consequently, the transmission of IPv6 packet within

the IEEE 802.15.4 Medium Access Control (MAC) frame becomes possible. The

Fig. 2 The 6LoWPAN ProtocolStack

The IoT Architectural Framework, Design Issues and... 131

123

Page 6: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

operations at the physical layer are defined by the IEEE 802.15.4 standard with data rate of

250 kbps, operating in 2400–2483.5 MHz Industrial, Scientific, Medical (ISM) frequency

band. The design of MAC layer still remains an open issue, especially in non-centralized

networks or networks with numerous nodes. Recently, the IEEE 802.15.4e standard has

been proposed as a MAC amendment to the existing IEEE 802.15.4-2011 standard. The

specific mode, called Time Synchronized Channel Hopping (TSCH), significantly increases

robustness against external interference and persistent multi-path fading, while running on

IEEE 802.15.4 hardware [16].

Recent advances in emerging areas such as cloud computing, autonomic computing,

social networks, etc. have changed the scope of the IoT’s convergence. Data management

is another crucial aspect in IoT. When considering a world of objects interconnected and

constantly exchanging all types of information, the volume of the generated data and the

processes involved in the handling of those data become critical. Some of the most relevant

concepts, which enable us to understand the challenges and opportunities of data man-

agement, are: data collection and analysis, big data, semantic sensor networking, virtual

sensors, and complex event processing. The concept of Software Defined Network (SDN)

has also demonstrated its potential for various IoT scenarios and applications [17]. The

integration of SDN and IoT offers a new and promising approach in designing, deployment

and monitoring of network resources and services.

3 IoT Hardware and Software Platforms

An important requirement in designing the IoT devices is related to ‘‘minimum energy’’

concept. This requirement may also be spread to general design approach in creating IoT

hardware and software components.

The IoT devices range from very lightweight sensors powered by 8-bit microcontrollers

(MCUs) to devices equipped with more powerful, but energy-efficient 32-bit processors.

These devices are based on different architectures, such as x86, MSP430, ARM7, Atmel

AVR, Cortex-M0, Cortex-M3, Cortex-M4. An IoT OS should run on various platforms,

including embedded devices as well as common PCs, and to support multiple drivers and

interfaces.

The examples of hardware platforms that support the IEEE 802.15.4 6LoWPAN con-

nectivity, are: TI CC2538 System-On-Chip (SoC) for 2.4-GHz IEEE 802.15.4-2006 and

ZigBee applications [18]; Libelium Waspmote based on ATmega1281 AVR RISC-based

MCU, with 17 different wireless interfaces including long range (3G/GPRS, LoRaWAN,

LoRa, Sigfox, 868MHz, 900MHz), medium range (802.15.4, ZigBee, WiFi) and short

range (Bluetooth 4.0, RFID, NFC), where more than 100 sensors are available to connect

to device [19]; Tmote Sky node based on TI MSP430F1611 MCU with ultra-low current

consumption and on-board humidity, temperature and light sensors, etc. [20]. The nodes

can be also divided into two groups: open-source and closed solutions. The example of

open-source node is Waspmote Pro from Libelium, the M3 open node based on a STM32

(ARM Cortex-M3) MCU. An example of a proprietary solution is, for example, MicaZ

platform [21]. The popular open-source prototyping platforms are Raspberry Pi [22],

Arduino [23], Intel Galileo [24], and BeagleBone [25]. Table 1 gives an overview of

various hardware platforms.

Software for IoT applications must support a variety of hardware platforms. Due to

memory and CPU constraints, the complexity of operations must be very low. Unlike

132 G. Gardasevic et al.

123

Page 7: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

Table

1Specificationsofvarioushardwareplatform

s

Device

MCU

CPU

(bits)

CPU

speed

Mem

ory

Connectivity

OS

Prog.language

Open

source

No.ofsensors

Rel.

Year

TelosB

TIMSP430

16

16MHz

10KB

RAM,

48KB

flash,1MB

external

flash

802.15.4,ZigBee,

6LoWPAN

TinyOS,Contiki

nesC,C

Yes

3(relativehumidity,

temperature,light)

2005

MICAz

ATmega128L

88MHz

4KBSRAM,

4KB

EEPROM,

128KB

flash

802.15.4

TinyOS

nesC

No

Sensingavailable

via

expansionboards

2004

OpenMote

ARM

Cortex-

M3SoC

32

32MHz

32KB

RAM,

512KB

flash

802.15.4,ZigBee,

6LoWPAN

Contiki,

FreeR

TOS,

OpenWSN,

RiOT

C,Python

Yes

3(relativehumidity/

temperature,

acceleration,light)

2013

Waspmote

ATmega1281

814MHz

8KBSRAM,

128KB

flash

802.15.4,ZigBee-

Pro,WiFi,3G/

GPRS,LoRa,

Bluetooth

4.0,

RFID

,NFC

No

APIlibraries

?compiler

No

2on-boardsensors:

temperature

and

accelerometer;more

available

toconnect

2013

Raspberry

Pi

3Broadcom

BCM2837

ARMv8

64/ 32

1.2

GHz

1GB

RAM

802.11nWLAN,

10/100Ethernet,

Bluetooth

4.1,BLE

Raspbian,Debian,

Fedora,ARCH,

Linux,ARM,

FreeB

SD,

NetBSD

Python,C,

C??,Java,

Scratch,

Ruby,anyfor

ARMv6

Partial

0,sensingavailable

via

external

sensors

2016

Arduino

Uno

ATmega328

16

16MHz

1KB

EEPROM,

2KB

SRAM,32

KBflash

Wi-Fi,GSM,

Bluetooth

Arduino,Linux

C,C??

Yes

0,sensingavailable

via

external

sensors

2010

The IoT Architectural Framework, Design Issues and... 133

123

Page 8: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

Table

1continued

Device

MCU

CPU

(bits)

CPU

speed

Mem

ory

Connectivity

OS

Prog.language

Open

source

No.ofsensors

Rel.

Year

IntelGalileo

IntelQuark

SoCX1000

32

400MHz

256MB,8

KB

EEPROM,

8MBflash

10/100Ethernet,

USB2.0

Arduino,Linux

distributionfor

Galileo

Withsupportto

GCCandICC

compilers

Yes

0,sensingavailable

via

external

sensors

2013

BeagleBone

Black

ARM

AM335x

32

1GHz

512KB

DRAM,4

GB

onboard

flash

10/100Ethernet,

USB2.0

Angstrom,Linux-

kernel

based

OSs

AnyforLinux

OS

Partial

0,sensingavailable

via

external

sensors

2012

134 G. Gardasevic et al.

123

Page 9: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

traditional OSs, the IoT OS has to cope with many limitations and restrictions, as well as

with heterogeneity of devices. The OS for embedded devices should have the lightweight

design, thus enabling the low code size, low complexity and low power consumption. The

mature and well tested embedded OSs for WSN applications are TinyOS [26], Contiki [27]

and Linux. Recently, the FreeRTOS [28], RIOT [29], and OpenWSN [30] have been

introduced. Commercial solutions, such as Sensinode’s NanoStack (TM) 2.0 [31] and the

Thingsquare platform [32], are also available on market. In May 2015, Google announced

the Brillo OS for IoT connectivity, and the Weave, a communications standard for smart

devices [33]. Brillo will be based on the lower levels of Android, thus supporting a wide

range of hardware platforms as well as Wi-Fi and Bluetooth connectivity.

TinyOS is an embedded, component-based, and open-source OS designed for low-power

wireless devices. It is written in Network Embedded Systems C (nesC), a variant of the C

programming language. The OS has a support for both 6LoWPAN and RPL, and also have

libraries for CoAP.

Contiki is an open-source, multitasking, event-driven OS designed for networked

embedded devices and smart objects. It consists of the kernel, libraries, the program loader,

and a communication stack with device drivers for a broad range of communication

hardware. Contiki uses a subset of the C programming language, and the configuration with

full IPv6 networking requires less than 10 KB of RAM and 30 KB of ROM.

A real-time OS FreeRTOS, from Real Time Engineers Ltd., supports 35 architectures

(MCUs). FreeRTOS provides methods for multiple threads or tasks and software timers. A

tickless mode is provided for low power applications. It has a very small memory footprint,

low overhead, and fast execution. It is written mostly in C programming language.

The RIOT OS represents the new generation of IoT OSs with real-time capabilities, thus

attempting to bridge the gap between IoT and WSN. The open source RIOT code requires

less than 5 KB of ROM and less than 1.5 KB of RAM. The RIOT uses a multi-threaded

programming model in combination with standard ANSI C code and a common POSIX-

like API for all supported hardware.

The OpenWSN project is a newly released open-source implementation of a fully

standards-based protocol stack for IoT capillary networks, rooted in the new IEEE

802.15.4e TSCH standard. A variety of hardware and software platforms are supported,

thus enabling the seamless integration within the IoT environment.

4 IoT Applications

In the coming years, the number of IoT applications and services will rapidly grow. There

are many application domains that will be impacted by the emerging IoT activities. In

order to address the relevant issues and requirements, we will discuss the following IoT

application domains: smart cities, healthcare, agriculture, and nano-scale applications.

Recently, the novel paradigm of IoNT has been introduced in order to incorporate nano-

networks into existing IoT scenarios.

4.1 Smart City Applications

There are many levels of improvement where the intelligent IoT solutions can increase the

quality of everyday life in dynamic urban environments [34, 35]. The concept of smart

city, which has already been applied in Barcelona, London, Amsterdam, Singapore, etc.,

The IoT Architectural Framework, Design Issues and... 135

123

Page 10: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

has proven its efficiency in terms of service management (e.g., energy, lighting, transport,

waste management), urban planning, traffic management, control of pollution and CO2

emissions, and so on. Many service sectors exist within this concept, like smart environ-

ment, smart governance, smart mobility, smart grid, smart utilities, smart traffic, smart

home/buildings, and others.

A novel paradigm referred as capillary network denotes the aggregated wireless con-

nectivity available in urban environment, and to provide seamless transmission from the

user terminal to access network [36].

WSN technologies enable the use of diverse sensors and actuators that must be

accessible individually, by uniquely assigned IP address. Here, devices mainly operate in

the 2.4 GHz band with data rate of 250 kbps, and sensor nodes are typically battery-

powered. The use of M2M communications for creating smart city applications allows a

plenty of user devices to exchange data seamlessly and transparently across the

infrastructure.

Several ongoing projects and campaigns are focusing on the identification of smart city

scenarios, the integration and interaction of various IoT data sources and systems, big data

representations and analytics, security and privacy issues, [37–40]. The primary goal is to

understand and model smart cities by taking into account different system topographies,

various data types, user and business preferences, etc.

One of the large-scale initiatives focused on IoT smart city applications is developed

under the FP7 SmartSantander project [41]. The infrastructure comprises of large number

IoT devices deployed in several urban scenarios, which are spread across several cities

(Guildford, Luebeck, Santander, and Belgrade). The project aims at enabling the large

scale experimentation, the analysis and evaluation of IoT concepts, under real-life

conditions.

The standardization activities in the smart city domain involves ISO/TC 268—Sus-

tainable development in communities; the ITU-T Study Group 5—Environment and cli-

mate change—Focus Group on Smart Sustainable Cities [42]; the IEEE 2030 series based

on a Smart Grid interoperability reference model (SGIRM) [43], etc.

4.2 Medical and Healthcare Applications

WSNs and Wireless Body Area Networks (WBANs) for medical and healthcare applica-

tions have received a significant attention from the research community. An efficient IoT

healthcare system aims at providing a remote monitoring of a patient health status in real

time, the prevention of critical patient conditions, life quality improvement of the elderly

through the smart environment, medical and drugs database administration, wellbeing

services, etc. [44, 45]. It is clear that the IoT networked devices will change radically the

current medical environments.

The IoT smart sensors for healthcare enable accurately measuring, monitoring and

analysing a variety of vital health status indicators, such as heart rate, ECG, blood pressure,

levels of glucose or oxygen saturation in the blood, etc. These parameters are then col-

lected and transferred to IP end-devices. Smart sensors are being assigned an unique IPv6

address for IoT connectivity.

The possibility to design and analyse a specific healthcare scenario within a simulation

platform is of a high importance [46]. The obtained results are giving the insight into the

implementation issues within the real environment. The important performance metrics

include parameters, such as robustness, accuracy and reliability, low power, durability,

136 G. Gardasevic et al.

123

Page 11: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

data validation, etc. Choosing the right platform and infrastructure technology for a par-

ticular healthcare scenario is a crucial and necessary requirement.

The standardization activities involve ITU-T: Focus Group on M2M Service Layer with

the priority on e-health, Requirements and network capabilities for e-health monitoring

services, Multimedia e-health data exchange services: data schema and supporting services

[47]; IEEE for wireless, ZigBee Alliance, ITU-T for M2M e-Health Service Layer, Con-

tinua Alliance for Use Case profiles and best practices, NIST, as well as diverse govern-

ment initiatives, etc.

Current research activities are directed towards the design of new and innovative MCU

components, which support the modular and interoperable approach, the high-speed

interfaces, specialized processes of control and sensing, security, patient safety, etc.

Another topic of interest is the generation of clinical data through IoT devices.

4.3 Agricultural Applications

In-field monitoring and control of agriculture and food production can be significantly

improved by adopting the IoT concept [48]. Such systems should ensure the quality of

products and may improve the management and monitoring of farms. Moreover, the IoT

system for agricultural applications should provide the in-depth analysis of application’s

requirement and smart utilization of available resources. The concept of IoT smart agri-

culture is also addressing the safety of agricultural products and the exchange of infor-

mation between producers and consumers.

Research in the areas of food and agriculture has been selected as one of priority areas

in ICT research and development. For example, the FP7 OpenIoT project covers a wide

range of areas spanning: middleware for WSNs, ontologies, semantic models and anno-

tations for representing internet-connected objects, cloud computing, etc. Additionally, the

Phenonet scenario for IoT agriculture has been developed. It enables plant breeders and

farmers to compare and evaluate the performance of different wheat varieties by using the

sensor network to gather environmental data.

The Waspmote Pro platform from Libelium is the example of product that supports

various applications in this domain, such as precision agriculture for predicting vineyard

conditions, monitoring greenhouse conditions, smart farming, smart factory for reducing

the maintenance cost, etc. [49].

4.4 Nano-Scale Applications

The concept of IoNT was proposed six years ago [50] as an intuitive extension of the IoT

paradigm. Nano-Things refer to interconnected miniature devices built either by down-

scaling processes of Mico-Electro-Mechanical Systems (MEMS), e.g., electron beam

lithography [51] or micro-contact printing [52], or from scratch using nano-materials, e.g.,

graphene [53, 54]. Single nano-devices are envisioned to be embedded from intra-body to

hard-to-access areas [55]. Once interconnected, they lead to the concept of nano-networks

[56], which imply the hierarchical architecture of the IoNT: nano-networks connected to

existing communication networks and the Internet.

Nano-networks are the key feature of the IoNT concept, whose realization in particular

has many challenges starting from the communication paradigm to be used. Nonetheless,

two are identified as candidates: molecular communication [57–59] and electromagnetic

communication. Molecular communication is already a well-functioning paradigm at

atomic and molecular level developed for us by the evolution. The central nervous system

The IoT Architectural Framework, Design Issues and... 137

123

Page 12: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

and cardiovascular system are typical representatives where the information between nano-

and micro-scale biological entities is carried via chain of chemical reactions and molecular

diffusion, calcium signalling or bacteria signalling. Molecular communication is used

among the macro-scale entities as well, e.g., animals use pheromones as messengers, which

sets this paradigm as a candidate not only for biomedical (intra-body) IoNT applications

but also for military, industrial and environmental. The challenges that remain are referred

to approaches in designing artificial nano-devices and nano-networks [56]. Electromag-

netic communication is a conventional paradigm which, however, still must be adopted for

the communication at nano-scale. Many issues are open and already identified starting

from quantum effects of nano-materials, terahertz frequency band of operation [50, 60],

channel models, information encoding, protocols for nano-networks and middleware that

connects nano-networks with conventional communication networks [55].

The IoNT offers advancements in various application fields owing to its compliance

with the IoT applications, but also enables completely new ones, e.g., those referred to the

intra-body biomedical applications, where the IoNT concept replaces the on-body wearable

devices with in-body nano-devices to advance monitoring, diagnosis and treatment through

sensing, imaging, detection and stimulation at nano-scale [61–63]. Moreover, the IoNT

offers advancements in general multimedia systems with an application in defense and

security [64], and environmental tracking and sensing [55].

5 Testing the OpenMote Platform for Industrial IoT Applications

The standardization of IIoT is still in its early stages. The main requirements relate to

operational efficiency, end-to-end automation, resource optimization, robustness and

reliability, while the key enablers are embedded sensors, cloud-based platforms, ubiquitous

connectivity, real-time analytics, etc. In this section, we will present preliminary results in

testing performances of hardware and software platforms for IIoT applications.

Recently, the OpenMote platform has drawn significant attention to research community

[65]. This platform belongs to the group of ‘‘resource-rich’’ devices as it has enough

memory and processing resources to support the TCP/IP protocol suite. The OpenMote is a

representative of new generation open-hardware platforms that is particularly adapted to

IIoT applications. This platform consists of three core elements: OpenMote-CC2538,

OpenBattery, and OpenBase, as shown in Fig. 3 (from the left-side to the right-side). The

OpenMote-CC2538 is based on TI CC2538 SoC functionalities, thus combining the 32-bit

Fig. 3 The OpenMote hardware platform [65]

138 G. Gardasevic et al.

123

Page 13: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

ARM Cortex-M3 MCU with the IEEE 802.15.4 transceiver. The OpenBattery enables

battery-powered supply for sensor board with three digital sensors: a temperature/humi-

dity sensor, an acceleration sensor, and a light sensor. The OpenBase is the interface board

for OpenMote and provides programming and debugging capabilities. The OpenUSB

provides the USB connection for easier programming and debugging capabilities. The

OpenMote platform has 32 KB of RAM and 512 KB of flash memory, and supports several

peripherals, such as GPIOs, ADC, I2C, SPI, UART, timer modules, and AES/RSA

Encryption Engine. This platform has also a full support for Contiki OS and FreeRTOS,

and a limited support for the RIOT. The OpenWSN OS is particularly tailored to run on

OpenMote. More details about this platform can be found in [66].

The OpenMote-based IoT testbed at the premises of University of Banjaluka, Faculty of

Electrical Engineering consists of 10 motes. The block-diagram in Fig. 4 illustrates the IP

connectivity between WSN and Internet. Motes are positioned in a way to provide the

multihop communications, based on RPL routing mechanisms. One mote is configured as a

root, thus acting as a bridge between the Internet and 6LoWPAN mesh network. Motes are

marked according to hexadecimal values of the last byte of their IPv6 address. In order to

demonstrate the functionalities of different OSs on this platform, we tested both Contiki

and OpenWSN. In order to test MQTT, we have used the latest Contiki 6LoWPAN

implementation for OpenMote platform [67]. For TSCH protocol testing, we have used the

OpenWSN implementation [68].

5.1 MQTT Protocol

Both CoAP and MQTT application protocols are designed to fulfil the requirements for

M2M applications in constrained environments. They provide the application’s interface

for remote monitoring of mote’s status (e.g., temperature, humidity, acceleration, light,

voltage, RSSI). The communication model of CoAP is similar to the client/server model of

HTTP, and relays on REST architecture. This requires that the client has a knowledge of

server preferences in advance to establish a connection. In order to avoid intensive

resources required by HTTP, CoAP is preferably used in edge-based devices. As already

mentioned, CoAP runs over UDP with support to multicast addressing. Similar to HTTP,

Fig. 4 Block-diagram of OpenMote IoT testbed

The IoT Architectural Framework, Design Issues and... 139

123

Page 14: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

CoAP provides secure communications based on Datagram Transport Layer Security

(DTLS) [69].

Unlike the CoAP, the MQTT runs over TCP/IP. This protocol has been recognized as a

particularly suitable for IIoT constrained devices and unreliable mesh networks. Due to its

lightweight realisation, the MQTT has been used in a variety of sensor networks scenarios:

home automation, telemetry, remote monitoring, alerting systems, healthcare IoT scenar-

ios, etc. In contrast to the CoAP, the MQTT uses a publish/subscribe architecture, as

illustrated in Fig. 5. The MQTT interface differs from the REST architecture in the sense

that there is a broker in between the source and the user. This type of interface requires the

devices to connect and publish data to a topic on an intermediary broker. MQTT uses SSL/

TLS protocols for secure communications [70].

For MQTT, there are a number of available brokers, both commercial and open source.

We have used the Mosquitto, an open source message broker that implements the MQTT

protocol versions 3.1 and 3.1.1 [71]. The MQTT client can be used to publish sensor

readings to an MQTT broker (Mosquitto in this case), as well as to subscribe to a specific

topic of interest and receive commands from an MQTT broker. This is particularly suit-

able for M2M messaging and IoT applications. We have tested the Contiki’s MQTT

protocol implementation on OpenMote hardware.

Figure 6 shows the status of real mote deployed in our testbed that can be monitored

remotely: sequence number, uptime, default route, RSSI, on-chip temperature, and voltage

level. After a connection between client and server is established, the first packet sent from

the client to the server is a CONNECT packet. The PINGREQ packet is sent from a client

to the server in order to indicate that the network connection is active, as well as to request

from server to confirm that it is active. In the latter case, the server will respond with the

PINGRESP packet. The delivery of application messages is performed according to the

QoS levels. The QoS level 0 indicates that the message is delivered based on the network

capabilities (best effort delivery), that is the message arrives at the receiver either once or

not at all. When using the QoS level 1, it is guaranteed that a message will be delivered at

least once to the receiver (it can be also delivered more than once). The QoS level 2

provides the highest QoS when neither loss nor duplication of messages are acceptable.

Fig. 5 Comparison of MQTT and CoAP

140 G. Gardasevic et al.

123

Page 15: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

The selection of particular application protocol depends on IoT scenario. In case of

complex system, it is possible to use several protocols. Publish/subscribe protocols (‘‘one-

to-many’’ or ‘‘many-to-many’’ messaging distribution), such as MQTT, provide better

performances when the IoT infrastructure is not deterministic, or when the deployment

scenario is frequently changing. On the other side, client/server protocols, such as CoAP,

are more interoperable and secure since they are based on point-to-point connections. As a

consequence, the protocol scalability is lower due to the fact that point-to-point connec-

tions are more resource-demanding and more complex to manage. In contrast, pub-

lish/subscribe protocols are more scalable since the decoupling of sources and users (sinks)

allows each to be added and removed independently. In comparison to CoAP, providing

the security for these protocols is more complex task. The interoperability issues may also

appear due to the lose coupling of the source and user. The ongoing research activities

include protocol optimisation and reliable message transmission [72]. In order to overcome

some of MQTT shortcomings, the MQTT-SN protocol is proposed [73]. It specifies a UDP

mapping of MQTT and adds broker support for indexing topic names.

5.2 Testing the OpenWSN OS on OpenMote platform

The another protocol, also recognised as one of a crucial for reliable IIoT applications, is

IEEE 802.15.4e Time-Slotted Channel Hopping (TSCH). It is the MAC amendment to the

existing IEEE 802.15.4-2006. TSCH combines time slotted access, multi-channel com-

munication and channel hopping, which significantly increases the robustness against

external interference and multi-path fading. For more details about TSCH and its Open-

Mote implementation, please refer to [68].

In order to test the TSCH performances in the presence of dynamic environment and

interference from WiFi access point and signal-generator that is generated the interfering

noise, we have created the experimental platform (see Fig. 4.). Motes are positioned in a

way to form a multihop network, where the node 4F is placed nearby the WiFi access

point. The distance between the two farthest motes (4F and E3) is around 15 m.

Fig. 6 Testing the MQTT protocol for sensor’s readings

The IoT Architectural Framework, Design Issues and... 141

123

Page 16: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

The host computer is running Ubuntu 14.04 LTS with OpenWSN and OpenVisualizer

(OV) installed. The OV creates IPv6 TUN (network TUNnel) interface that is a virtual

IPv6 interface tunnelled through the serial port. This interface enables the IPv6 commu-

nication between the host and motes’ network. OV also provides the graphical web

interface for network management purpose and various traffic and mote parameters

statistics. The mote that is directly connected to host is declared as DODAG root (C3).

Destination Oriented Acyclic Graph (DODAG) represents the hierarchical organization of

nodes in the network, based on RPL protocol specifications.

The transmit power used during experiment was 0 dBm. All motes in a TSCH network

are synchronized, where the time is divided into time slots. A slotframe consists of these

timeslots and it continuously repeats over time. Each mote follows a communication

schedule, where this schedule represents a matrix of cells. The number of cells depends on

a specific IoT application and the trade-off between latency, reliability and power con-

sumption is required. For measurement purpose, we selected the following TSCH

parameters: slotframe size ¼ 9 time slots, number of active time slots = 5, time slot length

¼ 15ms. In addition, 16 channels are available for frequency hopping. The measurements

were carried out in a way that motes are sending packets to the DODAG root. The number

of nodes that are sending the packets at the same time are 3, 5 and 7, respectively. We

performed the extensive measurements campaign. For each measurement, 1000 packets

Table 2 3 nodes; interpacket time ¼ 3000ms

Hop, node Sent packets Received packets Duplicate packets Regular packets PLR (%)

3, 4F 1000 1071 95 976 2.4

2, 31 1000 995 15 980 2

1, 6F 1000 1067 72 995 0.5

Table 3 5 nodes; interpacket time ¼ 3000ms

Hop, node Sent packets Received packets Duplicate packets Regular packets PLR (%)

3, 4F 1000 853 57 796 20.4

2, 31 1000 1001 13 988 1.2

1, 6F 1000 1100 107 993 0.7

1, B4 1000 1028 51 977 2.3

2, 54 1000 1033 62 971 2.9

Table 4 7 nodes; interpacket time ¼ 3000ms

Hop, node Sent packets Received packets Duplicate packets Regular packets PLR (%)

3, 4F 1000 894 62 832 16.8

2, 31 1000 414 8 406 59.4

1, 6F 1000 1080 92 988 1.2

1, B4 1000 1046 53 993 0.7

2, 54 1000 1066 74 992 0.8

1, E3 1000 699 0 699 30.1

2, 5A 1000 776 58 718 28.2

142 G. Gardasevic et al.

123

Page 17: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

were generated. The payload size is 20 bytes. We have used different inter-packet times:

1000, 1500 and 3000 ms. Due to space limitation, we present here the results in terms of

Packet Loss Rate (PLR) for 5 different test scenarios, Tables 2, 3, 4, 5 and 6.

The results show that, even in the presence of interference and dynamic environment

where people are moving, PLR is still at acceptable level. PLR depends also on number of

motes that are transmitting at the same time, and is the highest in the case where all 7

motes are sending packets to root (Table 4), and for motes closest to WiFi access point

(E3) and interferer (31). The another important parameter is the interpacket time and it

should be carefully chosen in order to avoid the queue overflow. The duplicate packets

appear as a consequence of interference, where the DODAG root can not process all

packets at scheduled times. Thus, the careful balancing of interpacket times and TSCH

parameters is required.

6 Conclusions

IoT applications are numerous, versatile, and closely related to our everyday life and

activities. There are many challenges and issues related to specific IoT scenario, envi-

ronment, context, technology, or application. Future research directions in IoT applications

and scenarios should also consider the social dimension and integration with social net-

works. This is important also in terms of increased number of IoT devices and connectivity

models. The programming paradigms should be focused on creating the self-aware

applications and systems with the ability of self-adaptation. Due to a variety of hetero-

geneous IoT solutions, protocols and data characterisation models, the interoperability of

standards and architectural frameworks is of prime importance.

This paper gives an overview of ongoing standardization efforts, design issues in terms

of the IoT hardware and software components, and addresses the requirements of relevant

IoT application domains for future IoT scenarios. We present the results obtained by

testing performances of OpenMote hardware platform for industrial IoT applications,

Table 5 3 nodes; interpacket time ¼ 1500ms

Hop, node Sent packets Received packets Duplicate packets Regular packets PLR (%)

3, 4F 1000 882 48 834 16.6

2, 31 1000 972 16 956 4.4

1, 6F 1000 1040 75 965 3.5

Table 6 5 nodes; interpacket time ¼ 1500ms

Hop, node Sent packets Received packets Duplicate packets Regular packets PLR (%)

3, 4F 1000 955 50 905 9.5

2, 31 1000 741 7 734 26.6

1, 6F 1000 1034 62 972 2.8

1, B4 1000 1003 48 955 4.5

2, 54 1000 984 50 934 6.6

The IoT Architectural Framework, Design Issues and... 143

123

Page 18: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

where the preliminary results are obtained by testing two protocols for emerging Industrial

IoT applications: MQTT and TSCH. In order to demonstrate the functionalities of different

OSs on this platform, we tested both Contiki and OpenWSN. The MQTT represents the

lightweight messaging protocol for IIoT and M2M communications, optimized for con-

strained networks and resources. Regarding the TSCH protocol, the experimental results

show that this protocol can provide good immunity to interference. In terms of a future

research, we will further investigate the performances of TSCH protocol in different

scenarios and for other metrics such as latency, throughput and energy-efficiency. More-

over, we will explore the possible implementations of QoS mechanisms at different layers

of protocol stack.

Acknowledgments This work is partially supported by the Ministry of Education and Science of Bosniaand Herzegovina and the Ministry of Science of Montenegro, under the bilateral Project Architecture, designand performances of DCQ switch, and by the Ministry of Science of Montenegro under Grant 01-451/2012(FOREMONT). The research is also partially supported by NORBOTECH (NORway-BOsnia TECHnologyTransfer) project under reference 2011/1370, within the Programme in Higher Education, Research andDevelopment; financed by the Norwegian Ministry of Foreign Affairs.

References

1. Botterman, M. (2009). Internet of Things: An early reality of the Future Internet. Workshop Report,European Commission Information Society and Media, May 2009.

2. Atzori, L., Iera, A., & Morabito, G. (2010). The internet of things: A survey. Computer Networks,54(15), 2787–2805.

3. Stankovic, J. A. (2014). Research directions for the Internet of Things. IEEE Internet of Things Journal,1(1), 3–9.

4. ITU-T. Internet of things global standards initiative. http://www.itu.int/en/ITU-T/studygroups/2013-2016/13/Pages/default.aspx.

5. IETF. (2010). The internet of things—Concept and problem statement. http://tools.ietf.org/id/draft-lee-iot-problem-statement-00.txt.

6. IEEE Interent of Things, Towards a Definition of the Internet of Things (IoT), Revision 1, May 27(2015).

7. Vasseur, J. (2011). Terminology in low power and lossy networks. IETF Internet Draft, September2011.

8. Vermesan, O., & Friess, P. (2013). Internet of things: Converging technologies for smart environmentsand integrated ecosystems. The River Publishers Series in Communications, June 2013.

9. NarrowBand IoT. http://www.3gpp.org/news-events/3gpp-news/1733-niot.10. IETF 6LoWPAN Working Group. http://tools.ietf.org/wg/6lowpan.11. ZigBee IP and 920IP. http://www.zigbee.org/zigbee-for-developers/network-specifications/zigbeeip/.12. Winter, T. et al. (2012). RPL: IPv6 routing protocol for low-power and lossy networks. In RFC 6550

(Proposed Standard), Internet Engineering Task Force, March 2012.13. IETF, The Constrained Application Protocol (CoAP), in RFC 7252 (Proposed Standard), June 2014.

https://tools.ietf.org/html/rfc7252.14. ZigBee Smart Energy. http://www.zigbee.org/zigbee-for-developers/applicationstandards/zigbeesmart

energy/.15. MQTT Protocol. http://docs.oasis-open.org/mqtt/mqtt/v3.1.1/os/mqtt-v3.1.1-os.16. IETF, Using IEEE 802.15.4e Time-Slotted Channel Hopping (TSCH) in the Internet of Things (IoT):

Problem Statement. https://tools.ietf.org/html/rfc7554.17. Buratti, C., Stajkic, A., Gardasevic, G., Milardo, S., Abrignani, M. D., Mijovic, S., et al. (2016). Testing

protocols for the internet of things on the EuWIn platform. IEEE Internet of Things Journal, 3(1),124–133.

18. Texas Instruments, CC2538 Powerful Wireless Microcontroller System-On-Chip for 2.4-GHz IEEE802.15.4, 6LoWPAN, and ZigBee Applications. http://www.ti.com/lit/ds/symlink/cc2538.

19. Libelium Waspmote. http://www.libelium.com/development/waspmote.20. Tmote Sky platform. http://www.advanticsys.com/shop/mtmcm5000msp-p-14.html.

144 G. Gardasevic et al.

123

Page 19: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

21. MICAz platform. http://www.cmt-gmbh.de/MICAz.22. Raspberry Pi platform. http://www.adafruit.com/pdfs.23. Arduino platform. https://www.arduino.cc/.24. Intel Galileo platform. http://www.intel.com/content.25. BeagleBone platform. http://beagleboard.org/bone.26. TinyOS Operating System. http://www.tinyos.net/.27. Contiki Operating System. http://www.contiki-os.org/.28. FreeRTOS Operating System. http://www.freertos.org/.29. RIOT Operating System. http://www.riot-os.org/.30. Watteyne, T., Vilajosana, X., Kerkez, B., Chraim, F., Weekly, K., Wang, Q., et al. (2012). OpenWSN:

A standards-based low-power wireless development environment. Transactions on EmergingTelecommunications Technologies, 23, 480–493. doi:10.1002/ett.2558.

31. Sensinode NanoStack 2.0. http://sensinode.sivuviidakko.fi.32. Thingsquare platform. http://www.thingsquare.com/.33. Google Brillo OS. https://developers.google.com/brillo/.34. Schaffers, H., Komninos, N., Pallot, M., Trousse, B., Nilsson, M., & Oliveira, A. (2011). Smart cities

and the future internet: Towards cooperation frameworks for open innovation. The Future Internet.Lecturer Notes in Computer Science, 6656, 431–446.

35. Zanella, A., Bui, N., Castellani, A., Vangelista, L., & Zorzi, M. (2014). Internet of things for smartcities. IEEE Internet of Things Journal, 1(1), 22–32. doi:10.1109/JIOT.2014.2306328.

36. Aug-Blum, I., Boussetta, K. Rivano, H. Stanica, R., & Valois, F. (2012). Capillary networks: A novelnetworking paradigm for urban environments. In Proceedings of the first workshop on Urban net-working (UrbaNe ’12), ACM, New York, NY, USA, 25-30. doi:10.1145/2413236.2413243.

37. Open Cities, EU FP7 project. http://www.opencities.net/.38. VITAL, EU FP7 project. http://www.vital-iot.eu/vision.39. RERUM, EU FP7 project. https://ict-rerum.eu/.40. CityPulse, EU FP7 project. http://www.ict-citypulse.eu/page/.41. Smart Santander, EU FP7 project. Future internet research and experimentation. http://www.

smartsantander.eu/.42. ITU-T Study Group 5. Environment and climate change. http://www.itu.int/en/ITU-T/studygroups/

2013-2016/05/Pages/default.aspx.43. IEEE 2030 Smart Grid Interoperability Working Group. https://standards.ieee.org/findstds/standard/

2030-2011.html.44. Riazul Islam, S. M., Daehan, K., Humaun, K., Hossain, M., & Kyung-Sup, K. (2015). The internet of

things for health care: A comprehensive survey. IEEE Access, 3, 678–708. doi:10.1109/ACCESS.2015.2437951.

45. Fang, H., Dan, X., & Shaowu, S. (2013). On the Application of the internet of things in the field ofmedical and health care. In Green Computing and Communications (GreenCom), 2013 IEEE andInternet of Things (iThings/CPSCom), IEEE International Conference on and IEEE Cyber, Physicaland Social Computing, (pp. 2053–2058), 20-23 Aug. 2013. doi:10.1109/GreenCom-iThings-CPSCom.2013.384.

46. Ugrenovic, D., & Gardasevic, G. (2015). Performance analysis of IoT wireless sensor networks forhealthcare applications. In The 2nd International Conference on Electrical, Electronic and ComputingEngineering IcETRAN 2015, Silver Lake, Serbia, June 8–11, 2015.

47. ITU-T Focus Group on Smart Sustainable Cities, An overview of smart sustainable cities and the role ofinformation and communication technologies. www.itu.int/en/ITU-T/focusgroups/ssc/.../TR-Overview-SSC.docx.

48. OpenIoT, EU FP7 project. http://www.openiot.eu/.49. Libelium Agriculture. http://www.libelium.com/case-studies/.50. Akyildiz, I. F., & Jornet, J. M. (2010). The internet of nano things. IEEE Wireless Communications,

17(6), 58–63.51. Tseng, A., Chen, K., Chen, C., & Ma, K. (2003). Electron beam lithography in nanoscale fabrication:

Recent development. IEEE Transactions on Electronics Packaging Manufacturing, 26(2), 141–149.52. Lee, H. H., Menard, E., Tassi, J. A. R. N. G., & Blanchet, G. B. (2005). Large area microcontact

printing presses for plastic electronics. Materials Research Society Bulletin, 846, 731–736.53. Geim, A. K., & Novoselov, K. S. (2007). The rise of graphene. Nature Materials, 6(3), 183–191.54. Geim, A. K. (2009). Graphene: Status and prospects. Science, 324(5934), 1530–1534.55. Balasubramaniam, S., & Kangasharju, J. (2013). Realizing the internet of nano things: Challenges,

solutions, and applications. Computer, 46(2), 62–68.

The IoT Architectural Framework, Design Issues and... 145

123

Page 20: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

56. Akyildiz, I. F., Brunetti, F., & Blazquez, C. (2008). Nanonetworks: A new communication paradigm.Computer Networks, 52, 2260–2279.

57. Nakano, T., Eckford, A., & Haraguchi, T. (2013). Molecular communication. Cambridge: CambridgeUniversity Press.

58. Atakan, B. (2014). Molecular communication and nanonetworks: From nature to practical systems.Berlin: Springer.

59. Pierobon, M., & Akyildiz, I. F. (2013). Fundamentals of diffusion-based molecular communication innanonetworks. Founds and Trends in Networking, 8(1–2), 1–147.

60. Jornet, J. M., & Akyildiz, I. F. (2014). Femtosecond-long pulse-based modulation for terahertz bandcommunication in nanonetworks. IEEE Transactions on Communications, 62(5), 1742–1754.

61. Akyildiz, I. F., Pierobon, M., Balasubramaniam, S., & Koucheryavy, Y. (2015). Internet of bio-nan-othings. IEEE Communications Magazine, 53(3), 32–40.

62. Nakano, T. et al. (2008). Microplatform for intercellular communication. In 3rd IEEE InternationalConference on Nano/Micro Engineered and Molecular Systems

63. Mesiti, F., & Balasingham, I. (2012). Nanomachine-to-neuron communication interfaces for neuronalstimulation at nanoscale. IEEE Journal on Selected Areas in Communications, 31(12), 695–704.

64. Jornet, J. M., & Akyildiz, I. F. (2012). The internet of multimedia nano-things. Nano CommunicationNetworks, 3, 242–251.

65. Vilajosana, X., Tuset, P., Watteyne, T., & Pister, K. (2015). OpenMote: Open-source prototypingplatform for the industrial IoT. In International Conference on Ad Hoc Networks (AdHocNets), Sep2015, San Remo, Italy (pp. 211–222).

66. OpenMote platform. http://www.openmote.com/.67. Contiki OpenMote. https://github.com/OpenMote/contiki.68. OpenWSN project. http://www.openwsn.org/.69. Capossele, A., Cervo, V., De Cicco, G., & Petrioli, C. (2015). Security as a CoAP resource: An

optimized DTLS implementation for the IoT. In 2015 IEEE international conference on communica-tions (ICC), London (pp. 549–554). doi:10.1109/ICC.2015.7248379.

70. Lesjak, C., et al. (2015). Securing smart maintenance services: Hardware-security and TLS for MQTT.In 2015 IEEE 13th international conference on industrial informatics (INDIN), Cambridge (pp.1243–1250). doi:10.1109/INDIN.2015.7281913.

71. Mosquitto project. http://www.mosquitto.org/.72. Hwang, H. C., Park, J., Shon, J. G. (2016). Design and implementation of a reliable message trans-

mission system based on MQTT protocol in IoT. Wireless Personal Communications, 1–13. doi:10.1007/s11277-016-3398-2.

73. MQTT for Sensor Networks (MQTT-SN) Protocol Specifications (Ver. 1.2). http://mqtt.org/.

Gordana Gardasevic received the B.Sc., M.Sc., and Ph.D. degrees inElectrical Engineering (with specialization in Telecommunications)from the Faculty of Electrical Engineering, Banja Luka, Bosnia andHerzegovina, in 1995, 2001, and 2008, respectively. She was aResearch Fellow at School of Electrical and Computer Engineering,National Technical University of Athens (NTUA), Greece(2006–2008). She was a Postdoctoral Fellow at the University ofBologna, (2013–2014). Currently, Prof. Gardasevic is the AssociateProfessor and Chief of Department of Telecommunications at Facultyof Electrical Engineering, Banja Luka. Prof. Gardasevic participated inmany national and international projects (PHARE, COST, TEMPUS,FP6, FP7, etc.). She is currently the B&H representative in COSTaction ‘‘Inclusive Radio Communication Networks for 5G andBeyond-IRACON’’, (2016–2020). Prof. Gardasevic has publishedmore than 50 scientific papers in international journals and conferenceproceedings. Key research interests of Prof. Gardasevic include next

generation networks architectures and applications, cross-layer protocol design, embedded wireless systems,Internet of Things architectures and applications.

146 G. Gardasevic et al.

123

Page 21: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

Mladen Veletic received the B.Sc. and M.Sc. degrees in electronicsand telecommunications from the Faculty of Electrical Engineering,University of Banja Luka (UNIBL), Banja Luka, Bosnia & Herze-govina in 2010 and 2012, respectively. Since January 2013, he is aPh.D student at the Department of Electronics and Telecommunica-tions, Norwegian University of Science and Technology (NTNU),Trondheim, Norway, and the Faculty of Electrical Engineering,UNIBL, as a scholar within HERD/ICT NORBAS project under ref-erence 2011/1383 supported by the Norwegian Ministry of ForeignAffairs. His research interests include wireless communications,positioning in cellular networks, molecular and nano-neural commu-nications. Mladen Veletic was awarded a Gold Plaque from theUNIBL for his achievements throughout the undergraduate education.

Nebojsa Maletic born in 1986, received B.Sc. and M.Sc. degrees fromthe School of Electrical Engineering, University of Belgrade in 2008and 2010, respectively. Presently, he is a research scientist at IHP(www.ihp-microelectronics.com). His research topics are in wirelesscommunication systems, including millimetre-wave communicationand beamforming systems, relaying systems and hardware impair-ments effects.

Dragan Vasiljevic received the bachelor degree in Electrical Engi-neering from the Faculty of Electrical Engineering, University ofBanja Luka, Bosnia and Herzegovina in 2002. Since 2002, he isemployee of Telecommunications of Republic of Srpska. Currently, heis master student in telecommunications at Faculty of ElectricalEngineering, University of Banja Luka. He has passed exams andsubmitted the master thesis for evaluations. His research interests arewithin Internet of Things applications, Wireless Sensor Network pro-tocols and routing, and embedded systems.

The IoT Architectural Framework, Design Issues and... 147

123

Page 22: The IoT Architectural Framework, Design Issues and ...ksuweb.kennesaw.edu/~she4/2018Spring/cs4491/ReadingList/IoT_Architecture.pdfThe IoT Architectural Framework, Design Issues and

Igor Radusinovic received the Telecommunications Engineeringdegree by the University of Montenegro in 1994. He received theM.Sc. and Ph.D. degree by the University of Belgrade, Serbia, in 1997and 2003 respectively. He is a Full professor in Telecommunicationsnetworks and Switching systems at the Faculty of Electrical Engi-neering, University of Montenegro. From 2009 to 2011 Prof. Radusi-novic was Deputy Minister for Science, Research and TechnologicalDevelopment in the Ministry of Education and Science of theGovernment of Montenegro. He was Montenegrin representative to theBoard of Governors of the Joint Research Centre of the EuropeanCommission (JRC EC) and European Research Area Committee(ERAC). Prof. Radusinovic was Chairmen of Montenegrin Council forScience and Research. He is a Member of the Managing Board ofUniversity of Montenegro and President of the Managing Board ofMontenegro Post. He has published as an author or co-author morethan 120 papers in international and national scientific journals and

international and regional conferences. He participated in a number of international (FP7, COST action) andnational research teams and projects, as well as bilateral research cooperation. Prof. Radusinovic hasconsiderable industry and operating experiences working on many strategic studies, regulatory issues,development strategies and technical solutions in area of ICT. Research interest of Prof. Radusinovic isfocused, but not limited, on: packet switching systems, telecommunications network theory, congestioncontrol algorithms, software defined networking and 5G.

Slavica Tomovic received B.Sc. degree in Electronics, Telecommu-nications and Computer Science from Faculty of Electrical Engineer-ing in Podgorica, University of Montenegro. From the same faculty shereceived M.Sc. degree in Telecommunication (2015). Currently,Slavica is Ph.D. student on Faculty of Electrical Engineering,University of Montenegro. She is also teaching/research assistant at thesame faculty. Her main research interests are in the areas of software-defined networking, quality of service management and architectures,Internet of Things and 5G wireless network design.

Milutin Radonjic received the B.Sc., M.Sc., and Ph.D. degrees inElectrical Engineering from University of Montenegro in 1991, 1997and 2011, respectively. He is an Assistant Professor at University ofMontenegro. Key research and expertise areas of Prof. Radonjic arepacket switching systems, computer networks, embedded systems anddigital system design.

148 G. Gardasevic et al.

123