Internet of Things: Architecture and Research Challenges Ankit Desai
Jekishan K. Parmar
Sanjay Chaudhary
PhD scholar, Institute of Engineering & Technology Ahmedabad University, Ahmedabad, Asst. Prof. Babaria Institute of Technology, Vadodara
Assistant Professor, Babaria Institute of Technology, Vadodara
Professor, Institute of Engineering & Technology Ahmedabad University Ahmedabad
[email protected]
[email protected]
[email protected]
Abstract The article first focuses on IoT Service Support and Economic Impact, and then explain IoT Applications and the IoT and M2M Ecosystem. In order to describe the IoT Architecture, details on the Application Layer, Management Service Layer, Gateway & Network Layer and Sensor Layer are explained. Finally, some important research and development areas are suggested along with IoT Technologies. IoT Scale IoT provides networking to connect people, things, applications, and data through the Internet to enable remote control, management, and interactive integrated services. IoT network scale, the number of mobile devices exceed the number of people on Earth. In addition, predictions are made that there will be 50 billion 'things' connected to the Internet by 2020. IoT Service Support Some advanced IoT services will need to collect, analyse, and process segments of raw sensor information, raw sensor data, and we need to turn this into operational control information. Some sensor data types may have massive sizes, because the number of sensor IoT devices are so large. Therefore a platform is required which can collect and store all of this massive amount of information. IoT databases will be needed, which can be done using Cloud Computing Support. IoT data analysis will be needed, which can be done using Big Data. The influence of IoT can be seen in people, processes, data, and things. If we see people wise, more things can be monitored and controlled, so people will become more capable. Process-wise, more users and machines can collaborate in real time, so more complex tasks can be accomplished in lesser amount of time because now we have more collaborative, more coordinated efforts that can be pulled together. Data wise, we can collect data more frequently and reliably. That would result in more accurate decision making. Things wise, things become more controllable. So therefore, mobile devices and things become more valuable. There is more that you can do with them. The overall economic impact, predictions have been made that IoT has the potential to increase global corporate profits by 21% by 2022. Table 1: IoT Market Asset Utilization
$2.5T
Employee Productivity
$2.5T
Supply Chain & Logistics
$2.7T
Customer Experience
$3.7T
Innovation
$3.7T
Table 2: IoT Technologies [1] $19 Trillion Market
M2M
$6.4T
M2P or P2M
$3.5T
P2P
$4.5T
45% 55%
Where is this all coming from? It is a combination of asset utilization, employee productivity, supply chain and logistics improvement, customer experience, and other type of combined innovations. The economic impact can be seen where machine to machine connections are increasingly becoming more and more important [2, 3]. However, still, person to person (P2P), as well as, person to machine (P2M), and machine to person (M2P), remains as the majority of IoT, is bigger economic value. The above table compares M2M to M2P or P2M, as well as P2P technology. IoT applications Security wise, surveillance applications, alarms, real-time object and people tracking and monitoring. Transportationwise, fleet management, road safety, emission control, toll payment, real-time traffic monitoring, and many more are intelligent transportation system applications. Healthcare-wise, e-health, personal security, body-sensor based customized healthcare systems. Utilities wise, measurement, provisioning, and billing of utilities for gas, water, electricity, and so much more. Manufacturing-wise, monitoring and automation of a product chain. Service and provisioning-wise, freight
supply, distribution monitoring, and vending machines can be controlled and provisioning support can be provided. Facility management wise, home, building, and campus automation can be achieved through IoT technology. Table 3: IoT M2M ecosystem Segment
Description
Companies / Organizations
BSP (Business Service Providers / Business Solution Providers) Services System integrators and IoT solution providers for enterprises
Network operators, communications Network
service
providers,
and
transport
infrastructure providers Table 3 depicts about IoT and M2M ecosystem. Moreover Table 4 depicts IoT based software and hardware categorization along with providers [6]. Table 4: IoT Software and Hardware Segment
Description
Companies / Organizations
Software manufacturers, middleware Software
and application infrastructure vendors, IoT OS providers
Manufacturers of GPS chips, Hardware
wireless sensor, warble devices, actuators, and embedded hardware devices
IoT Architecture First, there are four major layers. To start from the bottom, it is a sensor connectivity and network layer, layer one. On top of that is the gateway and network layer, layer two. Next, on top of that is the management service layer, layer three. Finally on top of it is the application layer, layer four. If you look at what is in here, for the service connectivity and network, there is the sensor network, sensors, and actuators, tags, which include RFID and barcodes, and other types of tags as well. At the gateway and network layer, we are talking about a wide area network, a mobile communication network, a Wi-Fi, Ethernet, gateway control and things like that. Then, going into the management service layer. Here, device modelling configuring and management is a major focus. Dataflow management, security control needs to be provided at the management service layer. Finally, we reach the overall application layer. This is where we have energy, environment, healthcare, transportation, supply chain, retail, people tracking, surveillance, and many, many more endless applications [10]. In order to understand this, each layer is elaborated here.
Fig 1. IoT Architecture The sensor layer provides sensor connectivity and networking. At the bottom, it starts off with the tags which includes RFID and barcodes. Then, on top of it, is sensors and actuators. This is a part that has solid state, catalytic, and also gyroscope, photoelectric, GPS, photochemistry, infrared, accelerometers, and similar things. On top of it is where, network connectivity comes in picture and that is like LAN, Wi-Fi and Ethernet. Wi-Fi for wireless, and Ethernet for wired local area networks. Then for personal area networks which are the smaller scale networks, which comes with wired and wireless side both. To focus on wireless, it includes Ultra Wi-Band (UWB), ZigBee, Bluetooth, 6LoWPAN, and there are other wired technologies [14, 15, 16]. The sensor layer is made up of sensors and smart devices, real-time information to be collected and processed. Sensors use low power and low data rate connectivity. This is where wireless sensor network formation need to be made such that, this sensor information is connected and can be delivered to a targeted location for further processing. Sensors are grouped according to their purpose and data types such as environmental sensors, military sensors, body sensors, home sensors, surveillance sensors, and other things. Also, sensor aggregators, and these are the gateway units, this needs to be provided through networking connectivity. At the local area network, there is Ethernet and Wi-Fi, at the Personal Area Network, there is ZigBee, Bluetooth, and 6LowPAN, and other protocols as well. At sensors which do not require connectivity to a LAN gateway, some of them may be directly connected to the Internet through a Wide Area Network [16, 17]. Now, gateway and network layer, which is on layer two. At this layer, the gateway needs to include micro-controllers, radio communication modules, signal processors and modulators, access points, embedded and operating systems, SIM modules, encryption, and units like that. On top of it is our gateway network which connects the gateways and the sensor information together. In this domain, wide area network and our local area network are located. In further details, the gateway and network layer are layer two. This must support massive volumes of IoT data produced by wireless sensors and smart devices. It requires a robust and reliable performance regarding private, public, or hybrid network modules. In addition, network models are designed to support the communication quality of service requirements for latency, error probability, scalability, bandwidth requirements, security while achieving high levels of energy efficiency meaning that they're low energy consuming. In addition, it is important to integrate different types of networks into a single IoT platform. IoT sensors are aggregated with various types of protocols and heterogeneous networks using different technologies. IoT networks need to be scalable to efficiently serve a wide range of services and applications over a large scale network where in this large scale network, some parts may have different protocols and different packet types, and different security requirements.
LAN Wi-Fi
Ethernet PAN
UWB
ZigBee
Bluetooth
Wired
6LoWPAN
Sensors / Actuators GPS
Infrared
Catalytic
Accelerometer
Photoelectric
Photochemistry
Gyroscope
Solid State
TAG RFID
Barcode (1D, 2D)
Fig 2. IoT Sensor Layer Now, management service layer. Operational support system (OSS), these includes device modelling, configuration, management, performance management, and security management, all of these rests in this layer. Then, there is the billing support system which includes billing reporting, service analytics platform, this is for statistical analytics, data mining, text mining, in-memory analytics, and predictive analytics. Then, management service for security, always needed access control, encryption, identify the accessed. In addition, Business rules management (BRM), rule definition, modelling, simulation and execution. Then there is the Business process management (BPM) which is in charge of workflow process modelling, simulation, and execution. In the management service layer, the management service layer is in charge of information analytics, security control, process modelling, and device management. The data management side needs to consider periodic and aperiodic characteristics. On the periodic side, for periodic IoT sensor data, this requires filtering because some data may not be needed, but because it is periodically going to be collecting information, there is going to be a lot of information, lot of sensor data that is not needed. Filter those out, choose the ones that is needed, and use and actuate, provide control management based upon these types of filter of the information that has something important included inside. Then comes aperiodic event triggered IoT sensor data. This may require immediate delivery and immediate response. For example, patient medical emergency sensor data, if you find something is wrong with your heart and a heart pacer is sending out a signal, well, that needs to be sent on the top priority. In addition, data management and data abstraction. On the data management side, this manages data information flow. In addition, information access, integration control all needs to provided, at this data management control unit. In addition, data abstraction, information extraction processing is needed. This needs to be used as a common business model because there will be so much information, that is needed to be able to provide an abstract view of the overall data that is in the system.
WAN
LAN
Gateway Network 3G
LTE
LTE - A
Micro-Controller
Wi-Fi
Radio Communication Module
Signal Processor & Modular
Gateway
Embedded/OS
Ethernet
Access Point
SIM Module
Encryption
Fig 3: Gateway and Network Layer Now, the application layer. In the application layer, first, we describe the horizontal market, fleet management, asset management, supply chain, people tracking, and surveillance. The sectors that use this overall domain of the application are environmental, energy, transportation, healthcare, retail, and military. In the application layer, various applications from industry sectors can use IoT for service enhancement. Applications can be classified based on the type of network availability, the coverage size, the heterogeneity. Also, business model as well as real-time or non-real-time requirements. At enterprise level of IoT, the scale of a community is much larger. Moreover, there are different characteristics that needs to be consider once reaching at the enterprise domain of application services for IoT. Now, the utility level, and here it is much larger, a national or regional scale of IoT service support. Now, then there is a mobile devices, which are usually spread across other domains, and this is because they have mobility. A lot of the devices will be battery operated and they will be portable IoT devices, and therefore they are going to move around, or the car, or the train, or some other type of transportation mechanism. In that case, mobility support is very important [7].
OSS (Operational Support System): Device Modelling / Configuration / Management, Performance Management, Security Management
Service Analytics Platform: Statistical Analytics, Data Mining, Text Mining, In-Memory Analytics, Predictive Analysis
BSS (Billing Support System): Billing Reporting
Security: Access Controls, Encryption, Identity Access
BRM (Business Rules Management): Rule Definition / Modeling / Simulation / Execution
BPM (Business Process Management): Workflow Process Modeling / Simulation / Execution
Fig 4. IoT Management Service Layer
Sector Environmental
Energy
Transportation
Healthcare
Retail
Military
Horizontal Market Fleet Management
Person Tracking
Supply Chain
Asset Management
Surveillance
Fig 5. IoT Applications Table 5: IoT Applications
Network Size
Smart
Smart
Smart
Smart
Smart
Smart
Smart
Smart
Home
office
Retail
City
Agriculture
Energy
Transpor-
Military
& Fual
tation
Large
Large
Large
Small
Small
Small
Medium
Medium / Large
Network
WPAN,
WPAN,
RFID,
RFID,
WLAN,
WLAN,
WLAN,
RFID, NFC,
Connectivity
WLAN,
WLAN,
NFC,
NFC,
Satellite
3G, 4G,
3G, 4G,
WPAN,
3G, 4G,
3G, 4G,
WPAN,
WLAN,
commu.,
Microwa
Satellite
WLAN, 3G,
Internet
Internet
WLAN,
3G, 4G,
Internet
ve links,
Commu.
4G, Satellite
3G, 4G,
Internet
Satellite
Internet Bandwidth
Small
Small
Small
Commu.
Commu. Large
Medium
Medium
Requirement
WLAN: Wi-Fi, WAVE, IEEE 802.11 a/b/g/p/n/ac/ad, etc.
WPAN: Bluetooth, ZigBee, 6LoWPAN, IEEE 802.15.4, UWB, etc.
Medium –
Medium –
Large
Large
Table 5. depicts the application layer and looks at it in terms of smart environment application domains. Moreover, Table 6. shows services for same smart applications of IoT [4, 5, 8, 9, 15]. Table 6. IoT smart applications and its services Service Domain
Services
Smart Home
Entertainment, Internet Access
Smart Office
Secure File Exchange, Internet Access, VPN, B2B
Smart Retail
Customer Privacy, Business Transactions, Business Security, Business Security, B2B, Sales & Logistics Management
Smart City
City Management, Resource Management, Police Network, Fire Department Network, Transportation Management, Disaster Management
Smart Agriculture
Area Monitoring, Condition Sensing, Fire Alarm, Trespassing
Smart Energy &
Pipeline Monitoring, Tank Monitoring, Power Line Monitoring, Trespassing & Damage
Fuel
Management
Smart
Road Condition Monitoring, Traffic Status Monitoring, Traffic Light Control, Navigation
Transportation
Support, Smart Car support, Traffic Information Support, Intelligent Transport System (ITS)
Smart Military
Command & Control, Communications, Sensor Network, Situational Awareness, Security Information, Military Networking
IoT Technologies To understand IoT technologies, first we describe Radio Frequency Identification (RFID), RFID chip holds information about the object, the thing. The RFID chip is attached and transfers data to the reader. Now, the antenna on the RFID module is used to receive energy that is used to operate the RFID device and transmit information back to the reader device [9].
Fig 6. Working of RFID
RFID enables efficient management, tracking and monitoring processes and therefore it is important in logistics and supply chain applications. RFID research and development is focused on streams of data support, chip design, energy usage optimization, automatic meter reading, home automation applications, and vehicle and transportation applications. Wireless sensor networks are also very important to IoT. Efficient, low cost, low power devices for use in remote sensing applications is very important. Low power integrated circuits and wireless communications are essential. A large number of intelligent sensors collect raw data and create valuable data by processing, analysing, and spreading data. Now, challenges are related to limited processing capability and limited storage, and sensor data sharing for multiple device cooperation is a very important challenge. For advanced IoT services, IoT networks may need to collect, analyse and process segments of raw data and turn it into operational control information. Therefore, advanced IoT services will need support of cloud computing. Numerous IoT connections will be made to various devices and sensors. Many IoT devices will not have a PC or smartphone level sufficient data processing capabilities. Or they will not have interoperability functions. Therefore IoT applications will need support from a reliable, fast and agile computing platform. IoT devices can overcome lack of software, lack of firmware, lack of memory, lack of hardware, lack of data processing capability through cloud computing. Now, cloud service models such as Software as a Service (SaaS), Platform as a service (PaaS), and Infrastructure as a Service (IaaS), are the services models that can support IoT technology. In other words, whatever software the IoT device does not have, then SaaS can provide it. Whatever database system that the IoT network does not have can be supported by PaaS. Whatever virtual machine or whatever processing that needs to be done, IaaS can provide that for the IoT network and IoT devices. Many IoT devices have small memory, and limited processing and limited communication functionality. Also, they are battery operated. Therefore, IoT requires integration of multitechnology networks to a common IP network platform. Therefore, IPv4 and IPv6 protocols, support addressing, management and scalability requirements. IoT will have significant influence on future Internet architectures. IoT services must guarantee the security, privacy and integrity of information and user confidentiality. Therefore some of the key features are thing authentication and authorization, user authentication and authorization. Now, what is this about? The IoT network is there, now what things, what objects, are going to be allowed data to be collected from. In addition, when a control signal is sent, what things are going to be controlled? This needs to be authenticated and authorized. In addition, what users will be allowed to access to IoT network to look at the data that is sensed and collected, and also control the objects, the things? The users need to be authenticated and authorized. In addition, thing to thing access control as in machine to machine access control. In addition, for security, IoT public key management and IoT private key management is very important. In addition, IoT low overhead protocol and IoT low complexity processing is also very important. In addition, mobility support is also important. Mobility support increases the applicability of IoT to new areas. Now, mobile platform based IoT enables an enormous range of future applications, such as location based services (LBS), social
networking, and environment monitoring and interaction. In addition, energy and resource management. Now, energy issues are related to optimization of energy harvesting, conservation, and usage and are essential to the development of IoT. It is important to consider resource constrictions, such as wakeup delays, power consumption, and limited battery and also packet size. Then the identification technology is another important area. IoT devices produce their own contents, and the contents are shared by any authorized user. Identification and authentication technologies need to be converged and interoperated at a global scale. Such that global users can use IoT devices far away. Management of unique identity for thing and handling of multiple identifiers for people and locations is very important. Conclusion Internet of Things is one of the most emerging area of research and development. IoT has received a huge attention due to support of technology and it is ability to penetrate into the existing system very effectively for the purpose of improvement in performance. This article mainly showcases the market of IoT, the applications and its domains and architecture of IoT. IoT Technologies are also presented. Future, research and development areas are also suggested. References [1] J. Bradley, J. Barbier, and D. Handler, "Embracing the Internet of Everything To Capture Your Share of $14.4 Trillion", Cisco, White Paper, 2013. [2] J. Bradley, C, Reberger, A. Dixit, and V. Gupta, "Internet of Everything: A $4.6 Trillion Public-Sector Opportunity", Cisco, White Paper, 2013. [3] D. Evans, “The Internet of Everything," Cisco IBSG, White Paper, 2012. [4] S. Mitchell, N. Villa, M. Stewart-Weeks, and A. Lange, “The Internet of Everything for Cities," Cisco, White Paper, 2013. [5] Hersent, O., Boswarthick, D. and Elloumi, O. (2011) IEEE 802.15.4, in The Internet of Things: Key Applications and Protocols, John Wiley & Sons, Ltd, Chichester, UK. [6] "M2M
Sector
Map,"
Beecham
Research,
Sep.
2011.
[Online]
Available
from:
http://www.beechamresearch.com/download.aspx?id=18 [Accessed Oct 10, 2015] [7] F. Behmann and K. Wu, Collaborative Internet of Things (C-IoT). John Wiley & Sons, 2015. [8] J. Gubbia, R. Buyyab, S. Marusica, and M. Palaniswamia, "Internet of Things (IoT): A vision, architectural elements, and future directions," Future Generation Computer Systems, vol. 29, no. 7, pp. 1645-1660, Sep. 2013. [9] L. Atzori, A. lera, and G. Morabito, "The Internet of Things: A survey", Computer Networks, vol. 54, no. 15, pp. 2787-2805, Oct 2010. [10] S. Li, L. D. Xu, and S. Zhao, "The Internet of Things: a Survey," Information Systems Frontiers, vol. 17, no, 2, pp. 243-259, Apr. 2015. [11] A. J. Jara, L. Ladid, and A. Skarmeta, "The Internet of Everything through IPv6: An Analysis of Challenges, Solutions and Opportunities", Journal of Wireless Mobile Networks, Ubiquitous Computing, and Dependable Applications, vol. 4, no. 3, pp. 97- 118,2013. [12] O. Vermesan and P. Friess, Internet of Things - Global Technological and Societal Trends From Smart Environments and Spaces to Green ICT. River Publishers, 2011. [13] O. Vermesan, P. Friess, P. Guillemin, S. Gusmeroli, H. Sundmaeker, A. Bassi, I. S. Jubert, M. Mazura, M. Harrison, M. Eisenhauer, and P. Doody, "Internet of Things Strategic Research Roadmap," European Research Cluster on the Internet of Things, Sep. 2011. [14] IEEE Std. 802.15.4-2006, Part 15.4: Wireless Medium Access Control (MAC) and Physical Layer (PHY) Specifications for Low-Rate Wireless Personal Area Networks (LR-WPANs), IEEE, Sep. 2006. [15] N. Kushalnagar, G. Montenegro, and C. Schumacher, "IPv6 over Low-Power Wireless Personal Area Networks (6LoWPANs): Overview, Assumptions, Problem Statement, and Goals," IETF RFC 4919, Aug. 2007. [16] G. Montenegro, N. Kushalnagar, J. Hui, and D. Culler, "Transmission of IPv6 Packets over IEEE 802.15.4 Networks," IETF RFC 4944, Sep. 2007.
[17] ZigBee Alliance, "ZigBee specification: ZigBee document 053474r13 Version 1.1." Dec. 2006.
About Authors: Ankit Desai is an Assistant Professor at Babaria Institute of Technology and Ph. D. Scholar at Institute of Engineering and Technology, Ahmedabad University. His areas of research interest includes Big Data Analytics, Data Mining Classification and Distributed Systems. Jekishan K. Parmar is an Assistant Professor at Babaria Institute of Technology. His areas of research interest includes Wireless Sensor Networks (WSN) with specialization in Underwater WSN, next generation networks and Mobile Computing. Sanjay Chaudhary is a Professor and Research Head at Institute of Engineering and Technology, Ahmedabad University. His areas of research interest includes Distributed Computing, Cloud Computing, Data Analytics, ICT Applications in Agriculture and Rural Development.