3.2 INTERNET OF THINGS COMMUNICATION MODELS
It is very important that to have proper operations we need to focus on how Iot devices are connected and communicate. In March 2015, an architectural document was released by the Internet Architecture Board which focused on for communication models used by IoT devices.
Device to Device Communication:
In this model the devices are directly connected to each other and have communication between them directly instead of having any application server as an intermediate.
https://www.ijarcce.com/upload/2016/march-16/IJARCCE%20264.pdf
In order to have direct to direct communication, these devices use protocols like Bluetooth, ZigBee etc. These models use small packets of information which have low data rate requirements and very commonly used in home automation systems.
Device to Cloud Communication:
These kinds of model have internet cloud service through which IoT devices are connected and communicate with each other. The internet cloud service usually is an application server that helps in data exchange and controlling message traffic.
https://www.ijarcce.com/upload/2016/march-16/IJARCCE%20264.pdf
Nest labs learning thermostat and Samsung Smart TV have deployed this kind of model. In canse of nest learning thermostat, the data is transmitted to the cloud database through which the energy consumption at home can be analyzed and also this cloud enables the user to have remote access to their thermostat via smartphone.
Device to Gateway model:
These kinds of model use ALG service which is application layer gateway model, which acts as a channel to reach cloud service and provides security and data translation.
https://www.ijarcce.com/upload/2016/march-16/IJARCCE%20264.pdf
usually the smartphones act as the local gateway device which communicate with devices and relay data to the cloud. The common example of this model is SmartThings hub which is stand alone gateway device that uses Zigbee and Z-Wave transceivers which communicate with the devices.
And then further it connects to cloud service known as SmartThings cloud service which allows user to access the devices via smartphone app and an internet connection.
Back-End Data-Sharing Model:
In this the smart object data is exported and analyzed from a cloud service architecture integrating with data from other sources.
https://www.ijarcce.com/upload/2016/march-16/IJARCCE%20264.pdf
this provides the users the permit to uploaded sensor data to third parties. This model is basically an extension to the device to cloud model. For example, in a corporate is a user has to analyze and consolidate the energy consumption by all the IoT sensors on the premises, in case of single device to cloud model the data produced by each IoT sensor sits in stand-alone silo.
Having an effective back-end data sharing architecture will help to analyze and access the data in the cloud that is produced by whole spectrum of devices in the premises. Also, this architecture helps the users to move their data when they switch between IoT services.
3.3 ARCHITECTURE OF IoT
3.3.1 Three and Five-layer architecture:
http://ibnox.com/read/7eapo35/internet-of-things-architecture-and-platforms
3-layer architecture is the of the basic architectures of IoT that was introduced in the initial stages of research. It consists of three layers:
• Perception Layer: it basically senses and gathers the information about the environment with help of sensors. It acts as a physical layer. It is used for identification of smart objects in the environment.
• Network Layer: the function of this layer is to connect to networks devices and servers. The sensor data is transmitted and processed by network layer.
• Application Layer: It deals with the various applications that can be deployed with respect to Internet of things. Example smart homes, smart cities, smart health etc.
To have a better idea about IoT, 5-layered architecture was introduced which focuses on the finer aspects of IoT. The function of perception and application layer is same, but the functions of other 3 layers are:
• Transport Layer: the function is to transport the sensor data from perception layer to processing layer with help of networks such as wireless, 3G, Bluetooth etc.
• Processing Layer: the function of this layer is to analyze and process the data that it has received from transport layer. Various technologies that are employed by processing layer are cloud computing, big data processing modules etc.
• Business Layer: the main function of this layer is to manage the whole IoT system.
3.3.2 Cloud and Fog based Architecture
http://ibnox.com/read/7eapo35/internet-of-things-architecture-and-platforms
In fog architecture the data processing and analyzing is done by sensors and network gateways. Fog architecture is a layered architecture where security, storage, preprocessing and monitoring layers are placed between transport and physical layer.
• The function of monitoring layer is to monitor power, responses and services.
• The function of preprocessing layer is to filter, analyze and process the sensor data
• The function of storage layer is to provide storage functionalities like data replication, data distribution etc.
• The function of security layer is to perform encryption and decryption and also checks for privacy.
Before the data is sent to the cloud, monitoring and preprocessing of data is done at the edge of the network.
3.4 INTERNET OF THINGS (IoT) TAXONOMY
Coming to the architecture of IoT the first component is the perception layer which uses sensors to collect the data. There are various kinds of sensors used today and smartphone is one of them and has a lot of sensors embedded in it for example GPS, camera, light sensor, microphone. Etc. apart from these various sensors are being used to monitor various things like temperature, pressure, humidity, medical parameters of boy. Etc.
The next component is the data processing whose function is to filter and summarize data and make it available to the network. The next architectural component is communication where many devices communicate with each other using different protocol standards.
Protocols like radio frequency identification (RFID), and near field communication (NFC) are used for short range low power communication. Protocols like Bluetooth, ZigBee, Wi-Fi are used for medium range communication.
Communications in IoT requires different protocol mechanism that should run effectively on the devices to have an efficient communication with each other.
There are two kinds of software components for IoT:
1. Middleware: It basically creates an idea for the programmer where the hardware details can be hidden. This results in efficient interoperability. The various examples of middleware services for IoT are OpenIoT, MiddleWhere, Hydra, FiWare and Oracle Fusin Middleware.
2. Applications: it includes various IoT applications like home automation, health and fitness, smart vehicles and transport system, smart cities, smart environment and entertainment etc.
http://iotdirect.blogspot.ca/
3.4.1 Sensors and Actuators
Sensors are the basic requirements in the IoT applications because they are responsible for collecting the data from the environment. The various types of sensors used in IoT applications are:
1. Mobile Phone based Sensors: smartphone is a very user-friendly device that itself has a lot of sensors embedded in it. With increase in the popularity and need for smartphones many companies are using them for IoT purposes because of sensors embedded in them. To bring out the meaningful data applications are built on the smartphones that uses sensor data. The diverse types of sensors in a smartphone are:
Accelerometer sensor that helps to sense the motion and acceleration of a smartphone. Accelerometer can be used to sense physical activities of a user like walking, running with help of data patterns captured by the sensor.
For capturing audio an visual information microphones and cameras are used which can help to analyze and capture lots of user information like the user’s current environment.
For detecting the location, GPS is widely used in the smartphones. The detection of location is done using trilateration.
The light sensor in the smartphones can be used to control the lightning system.
2. Medical Sensors: In the healthcare system medical sensors are used to keep the track for persons health and fitness. Many devices are available in the market that are embedded with the sensors that are used for measuring various parameters like blood pressure, body temperature, heart rate and pulse etc.
There are many wearable devices present in the market like smart watches, fitness trackers, wristbands etc. smart watches basically have their connectivity with smartphones and smart sensors.
Another kind of advancement is the monitoring patches that are pasted on the skin which are used by patients for few days for monitoring their health parameters.
https://www.fitbit.com/en-ca/
3. Neural Sensors: EEG known as Electroencephalography is used for reading the brain signals. The electric field generated by neurons inside the brain can be measured from outside in terms of frequency. The various types of brain waves are alpha, beta, gamma, theta and delta. These waves can tell us regarding the brain activities. This information can be used in real time for training our brain to stay focused and manage stress etc,
http://www.livingly.com/Health+Diet+Fitness/articles/s8-Dm5kiR00/Gadget+Teaches+Meditate+Reading+Brain+Waves
4. Radio frequency Identification: It is used for identification purposes and consists of an RFID tag that consists of data and RFID reader is used to read the data. The two types of RFID technologies are used in the market, near and far. The RFID tags can be attached to a vehicle to determine whether its an authorized vehicle or not. It can be used for tracking the movement of objects.
5. Actuators: these devices are used for converting electrical energy into useful energy. Examples are speakers, lights, motors etc. In home automation systems actuators can be used for locking and unlocking of doors, switch on/off lights or any other electrical appliances etc. Actuators are of three types: electrical, hydraulic and pneumatic. The most common example in IoT is the digital finger used for switching on/off lights.
3.5 IoT APPLICATIONS
IoT has brought a lot of significant changes to the lives of people by connected them anytime, anywhere and at anyplace. It has played a significant role in various fields of life transforming into smart cities, smart healthcare, smart retail, smart transport system, smart vehicles. Etc. the key factors for building IoT are security, privacy, safety, integrity, trust, dependability, transparency, anonymity and ethics
http://www.internet-of-things-research.eu
IoT 3D Matrix
The applications of IoT are very diverse and we have them in our daily lives. The different application domains of IoT which are most popular are:
• Smart Cities: it is estimated that by 2020 the cities will have a networked architecture and they will b named as smart cities. By 2025, 60% of the population is expected to live in the urban cities and the urbanization will have a significant impact on the lives and mobility.as per estimation, there will be 30 mega cities globally by 2023 with 55% in countries like India, Russia, China and Latin America. Smart cities will enable various smart features like smart economy, smart buildings, smart mobility, smart energy, smart information and technology, smart planning, smart citizen and smart governance.
https://static.thenortheasttoday.com/wp-content/uploads/2016/09/smart-city-concept-internet-things-different-icon-elements-modern-design-future-technology-living-66876194.jpg
• Smart Health: the todays health market is focused on enabling such devices that are made up of diverse architectures and user friendly as well. IoT is playing a key role in smart health where patients can get help easily anywhere and anytime in the world. These techniques are very beneficial because it improves the quality of care and helps in cost reduction via eliminating the need of care takers.
The main aim is to improve the quality of lives of the people by providing right medical support.
http://www.pchalliance.org/00114-nhealthcare-smart-healthcare
• Smart Homes: the IoT has played a vital role in building smart homes with sensors and automation system that provides people monitoring and control of their home appliances and system whether they are outside or inside. Smart homes consist of devices that provide security, comfort, energy efficiency and these devices are usually controlled by smartphones.
These days smart homes have controls for windows, doors, security, lightning, temperature etc. requirements to setup a smart home includes sensors, actuators, controller, compatible appliances, network system and an interface.
Examples of smart home devices for 2017 are Amazon cloud cam, Philips hue white LED, ECOBEE thermostat, August’s Smart lock, Wi-Fi video door bell etc.
https://www.homedepot.ca/en/home/ideas-how-to/buying-guides/smart-home.html
• Smart Retail: the best example for the recent development in digital trends is shopping. Smart retail has made lives of consumers easy by making such applications that are user friendly and they can access information anytime and anywhere in the world.
The developments like online stores, physical stores, social media, mobile communications have impacted the retail sector in a terrific way. With help of IoT lot of retailers have moulded their way of retail to give better customer satisfaction.
Examples of IoT smart retail devices are Asset tracking via RFID, sensors and cameras for security and monitoring, instore Wi-Fi, digital signage, IoT based payment systems, smart lightning etc.
• Smart Water Systems: These days it has become really important to manage our water resources because of the increasing demand due to which many cities in the world are deploying meters on water supply lines to keep the record of water usage and wastage. The meters will keep a record of inflowing and outflowing water and will also check if there is any leakage or not.
• Smart transport: various applications are used in transport system to control and manage the traffic in the cities. Traffic conditions can be estimated when vehicles connect to each other through networks, RFID’s, GPS, cameras etc. with help of smartphone sensors like accelerometer and GPS traffic congestion can also be predicted. Intelligent parking management system is being used in the cities which uses sensors to detect whether the parking space is available or not.
Another new application is the smart traffic lights that have sensors and capture and analyze the information regarding traffic congestion in a intersection and will also forward the data to the neighbouring traffic light system.
3.6 IOT PROTOCOLS AND STANDARDS
The Protocols that are used in IoT focus on multiple sensors so that can communicate properly and aggregate information before getting to the internet. In the routing layer, protocols are defined to have proper routing for sensors. The session layer protocols are used for enabling the messaging among various elements of ecosystem. Etc
https://www.cse.wustl.edu/~jain/cse570-15/ftp/iot_prot.pdf
1. DATA LINK PROTOCOLS: It includes various physical and MAC layer protocols.
• IEEE 802.15.4: This protocol is widely used and is most common standard for MAC. It defines source address, destination address and how communication should be done between the nodes. Its extension was IEEE 802.15.4e which was introduced in 2008 to support low power consumption. Various features include:
Slotframe Structure: this structure deals with scheduling and informing nodes what to do. During the sleep mode, the node saves the power by turning off its radio and will collect all the information and messages that it will forward at next transmission opportunity. After transmitting the data, it waits for the acknowledgement. When the message has to be received, the node will turn on, receive the message and will send an acknowledgement and then it will turn off its radio and will deliver that data to the upper layer before going back to sleep.
Synchronization: It is very important that nodes communicate with each other properly and it is made sure by synchronization. Synchronization can be achieved in 2 ways: acknowledge based and frame based synchronization. Acknowledge based is the one where node sends the massage and receives the acknowledgement once the message is received and this ensure proper connectivity between them.
• IEEE 802.11 AH: this standard is the low energy version of the previous one. It is widely used in laptops, computers, tablets, digital tv’s etc. various features include:
Synchronization Frame: A station should have complete medium information that allows it to capture the medium and stop packet exchange, then only the station can transmit. If the duration packet is correctly received, then the station will have information otherwise it has to wait a specific amount of time called probe delay.
Efficient Bidirectional packet exchange: In bidirectional packet exchange there is uplink and downlink communication between access point and sensor which saves the power. Once is communication is finished, the sensor will go back to sleep.
Short MAC Frame: the normal frame as per IEEE 802.1 is 30 bytes which is not suitable for IoT applications, IEEE 802.11ah introduces a short MAC frame which is 12 bytes.
• WirelessHART: It is based on time division multiple access and lies on the top of the IEEE 802.15.4 PHY. In order to offer better reliability, it uses advanced encryption to encrypt messages and calculate integrity. The architecture consists of Network manager, gateway to connect wireless devices to wired networks, security manager, routers, adapters etc.
https://www.cse.wustl.edu/~jain/cse570-15/ftp/iot_prot/
• Z-Wave: This protocol is mostly popular in smart home system because of its low power characteristic. It offers 30 m point to point communication and used for applications like light control, energy control, healthcare control etc. It is based on CSMA/CD for collision detection and ACK for good transmission.
• Zigbee Smart Energy: It is used in many IoT applications like smart homes, smart healthcare, remote controls etc. various topologies supported by Zigbee are star, peer to peer or cluster-free. There are 2 stack profiles that are defined by Zigbee standards: Zigbee and Zigbee Pro where both support full mesh networking.
• DASH7: It is a wireless communication protocol and is available in Industrial Scientific Medical band. It is used for long range outdoor coverage and supports IpV6 addressing. Its architecture is based on Master/Slave architecture. Various features defined are:
Filtering: CRC, a 4 bit subnet mask and link quality assessment are used for filtering of frames and processed further.
Addressing: Two types of addressing used by DASH7 are: unique identifier which is a EUI 64 bit ID and dynamic address identifier which is 16 bit address identified by network administrator.
Frame Format: The frame is of variable length consisting of maximum 255 bytes which had addressing, subnets, estimated power of transmission etc.
• LTE-A: It is used for M2M communications and applications in cellular network.
https://www.cse.wustl.edu/~jain/cse570-15/ftp/iot_prot.pdf
It is based on Orthogonal frequency multiple access where MAC layer divides frequency into multiple bands so that each one can be used separately. The various components of its architecture are Core network, radio access network, and mobile nodes. The function of core network is to monitor the mobiles devices and keep a track of their IP’s. the function is radio access network is to handle wireless connectivity and establish control and data planes.
• LoRaWAN: It offers low power, low cost, security, mobility and bi-directional communication. It is considered to be the new arising wireless technology which is designed for low power WAN networks.
2. NETWORK LAYER ROUTING PROTOCOLS: The network layer is divided into 2 sublayers: routing layer which is responsible for transfer of packets from source to destination and encapsulation layer which forms the packet.
• RPL: It is a distance vector protocol that has destination oriented directed acyclic graph which represents only one route from each leaf node to the root in which all the traffic will be routed to. It is defined as routing protocol for low power and lossy networks.
• CORPL: The extension to RPL is known as CORPL which is designed for cognitive networks. It chooses the forwarding path through multiple forwarders and chooses the next best hop to forward the packet after building a coordination with nodes.
• CARP: It stands for channel-Aware Routing Protocol and is designed for communications which are underwater. It has 2 cases: network utilization where HELLO packet is transmitted from sink to all other nodes in the network and data forwarding where in hop by hop fashion packet is routed from sensor to sink.
3. NETWORK LAYER ENCAPSULATION PROTOCOLS: One of the problems with the IoT applications is that it cannot accommodate IPV6 addresses because they are too long. Therefore, IETF is building mechanisms that can help to accommodate such long addresses. Various mechanisms are:
• 6LoWPAN: It is one of the most commonly used standard which accommodates long addresses in IEEE802.15.4 small packets which cannot exceed 128 bytes. This standard uses four types of headers: No 6LoWAN header (00), Dispatch header (01), Mesh header (02) and Fragmentation header (03). This standard gives specifications which support different topologies, low bandwidth, low power, mobility, long sleep time, low cost. Etc.
http://www.ti.com/wireless-connectivity/6lowpan/overview.html
Functions of four types of header are:
(00), frames are discarded if they don’t follow 6LOWAN specifications.
(01), it is used for multicasting and used for compressing IPV6 header.
(02), it is used for broadcasting.
(03), this is used to break the IPV6 header so that they can be accommodated in fragments of maximum 128 bytes.
• 6LO: This standard is used to facilitate the connectivity of IPV6 over restricted node networks. Two 6LO specifications have been defined:
a. IPV6 over G.9959: It defines a home network identifier which is 32 bits and is assigned by controller and also a host identifier which is 8 bit and allocated for each node.
b. IPV6 over Bluetooth low energy: The LLC and L2CAP layer provides segmentation and reassembly of large payloads into 27 byte packets.
4. SESSION LAYER PROTOCOLS: TCP and UDP have been used by various IoT applications and even IP applications for transport. The various session layer protocols defined are:
• MQTT: It stands for Message Queue Telemetry Transport and was introduced in 1999 by IBM and was standardized in 2013 by OASIS. On one side it provides embedded connectivity between applications and middleware and on the other side it provides networks and communication.
https://www.survivingwithandroid.com/2016/10/mqtt-protocol-tutorial.html
It is based on publish/subscribe architecture. The above figure shows the 3 components of architecture namely publisher, MQTT broker and subscriber. In IoT applications the publisher is the sensor which is lightweight and used to connect to the broker so that it can forward its data an go back to sleep whenever possible. Subscriber is the application that is connected to broker which informs them whenever a new sensory data is available. The MQTT broker will break the sensory data into similar group topics and send them to the subscribers according to their need.
• SMQTT: It is the extension of MQTT and uses encryption based on lightweight attribute. This kind of feature has an advantage of broadcast encryption which will encrypt the message and will be delivered to multiple nodes. It has four main stages:
a. Setup: in this phase the subscribers and publishers register themselves to the broker and in return they are assigned a master secret key.
b. Encryption: when the data is published by publishers it is encrypted at this stage.
c. Publish: when the data is received to broker it is then published and sent to subscribers.
d. Decryption: at this stage the data is decrypted at subscribers.
• AMPQ: this protocol stands for Advanced Message Queuing Protocol and was designed for finance industries. It is based on same architecture as MQTT and runs over TCP.
the main difference between architecture of MQTT and AMPQ is that the broker is divided into 2 categories exchange which receives the publisher messages and forwards to the queues based on the pre-defined categories. Queues are basically the topics that are broken according to subscribers needs.
• COAP: It stands for Constrained Application Protocol. The interface used between HTTP client and server is known as Representational state transfer (REST). COAP helps to provide a RESTful interface which uses low power sensors.
https://www.cse.wustl.edu/~jain/cse570-15/ftp/iot_prot/
to provide reliability, it is built over UDP instead of TCP. The two main sublayers of COAP are:
a. The messaging sublayer: it deals with duplication of messages and reliability
b. Request/response sublayer: it deals with communication.
CoAP has four messaging modes: confirmable and non-confirmable modes represent reliable and unreliable transmissions, piggyback is used for client and server communication while separate mode is used server comes across a message separate from acknowledgment. The request used by CoAP are GET, PUSH, PUT AND DELETE to retrieve, create, update and delete respectively.
• XMPP: It stands for extensible messaging and presence protocol which is used for chatting and message exchange applications. It is used in near real-time applications and is highly efficient over internet. However, its does not provide QOS and thus not suitable for M2M communication.
5. IOT MANAGEMENT PROTOCOLS: There are 2 management protocol standards that are used to provide communication between different data links and play a critical role in IoT applications.
https://www.cse.wustl.edu/~jain/cse570-15/ftp/iot_prot/
The 2 standards are: Interconnection of heterogeneous data links and Smart transducer interface. The main function of abstraction layer here is to hide the features of different protocols and exchanges messages known as control message data units.
The figure above shows the abstraction layer management entity used in IEEE 1905.1 devices which offers topology exchange, security associations, topology change notifications. Etc.
Smart transducer interface standard is uses plug and play identification with help of transducer electronic data sheets known is TED’s. These data sheets are present in transducers which contains information which is required by measurement system.