IoT. The internet of things: a ultra-interconnected world
Have you ever wondered a world where all the electronic devices that you can think of can be controlled, managed from everywhere at anytime. Well, this is the aim of the internet of things, create smart devices capable of being interactive through the use of internet.
We can define the internet as sophisticated set of applications built on top of a complex network of interconnections serving to millions of user on an uninterrupted manner. This ease on the way we communicate has created an new environment capable of creating devices that can in the future automate all the simple task of our life but we can not just think this phenomenon in our everyday life, this covers fields like commercial, industrial, educational and many more fields making IoT one of the trends that are going to have a huge impact on the way we live.
The smart devices (“things”) can be seen as sensors, human beings, electronic devices, actuators, and any other entity capable to be aware of his context and it has the capability to communicate with his owner and other devices without the restrictions of time and place.
The communication is the fundamental base for the development of the IoT, and to accomplish this requirement the applications has to support many devices and communication protocols ranging from tiny sensors to powerful servers that are used to analyze and report big amounts of data. This also requires the integration of mobile devices and network devices like routers, gates and also humans as controllers.
IoT definition and evolution
Kevin Ashton is accredited for using the term “internet of things” on a presentation about supply-chain management in 1999. In his presentation he setted the bases to think about how the smart devices, data collection, internet, computing and us interact with the environment and set the first ideas for the actual IoT developments. Since then many definitions has saw the light, some of them are IoE (internet of everything, used by cisco to refer to people, things, and places) or IIoT(industrial internet of things. Big data analysis, machine to machine communication and machine learning are the building block).
As identified at . IoT can be generalized into three categories
- Internet-oriented (middleware, on demand storage and computing tools)
- Things-oriented(sensors, hardware, embedded communication, actuators)
- Semantic oriented(knowledge, presentation of the data)
This three differentiations are necessary due to the amount of interdisciplines but the potential of the technology just can be unleashed on the application level where the three disciplines intersect.
In my opinion a good definition for the internet of thing is the given for gubbi et al, which is:
“Interconnection of sensing and actuating devices providing the ability to share information across platforms through a unified framework, developing a common operating picture for enabling innovative applications. This is achieved by seamless ubiquitous sensing, data analytics and information representation with Cloud computing as the unifying framework.”
On IoT normal isolated objects become objects that belongs to a network and with the use of the actual communication protocols like 5G and the number of devices growing every day the amount of data produced has acquired a new dimension where a new paradigm has been created, called big data. This has created a new necessity to store, process and present this data in a efficient manner.
Smartness in IoT
The smartness of a network distinguish IoT from a normal sensor network. A smart network device is characterized by the following:
- Standardization and openness of the communication standards used on the different layers of the protocol (unified framework)
- Object addressability (every object has his unique identification)and multifunctionality of the network that is the capability of a network to serve to different type of applications (home applications, industry application, etc).
The four stages of IoT
The four main stages of the IoT process are represented on the following image:
A typical workflow for a IoT project is showed in the image, the first step in this chain are the devices. This elements are usually electronic elements that captures information, for example a level sensor on a tank or a smartwatch that periodically takes measurements of your health indicators. This devices are managed for embedded systems, these are micro computers in charge of the management of the hardware and also provide the communication protocols with the receiver of the information so usually this devices uses communication RFID protocols like wifi or bluetooth. The next step on this line is the data aggregation and preprocessing which is in charge of adapting the data to the specific needs for the proper storage, the steps that are done here are usually to take a raw input and make it a tidy data shaped to be stored with the objective to make this information easier to handle for his analysis. Later the information is stored on a database on the most convenient format. This kind of storages are based on the cloud and are formed by dedicated servers on a datacenter with high processing and storage capabilities. The last step and one of the most important is the data analysis step because here the three before steps become relevant when with applications we can turn our simple data into relevant information to get insights or to take actions of different processes. The visualization of the information has become an important topic in IoT applications because this allows the interaction between the user and the environment.
Architecture of IoT
As mentioned before the aim of IoT is to picture an unified network of smart devices and humans capable of communicate each other, but this is not an easy task so we need an architecture that has to be reliable, robust, scalable, adaptable and serve as a link between the virtual world and the real one.
There are several types of architectures for an IoT environment mainly differentiated for his level of abstractions and each of one has is advantages, disadvantages, constraints and implementation details.
Service oriented architecture is the traditional 4 layer method mentioned above, which are:
- Sensing layer: hardware sensing elements
- network layer: wireless o wired infrastructure for interconnectivity between the sensing layers
- service layer: services required for users and application
- interfaces layer: interaction methods between users and applications
for each of this layers the provider try to create services that can be used together to create the solution, like for example AWS microservices.
The advantages of this type of architecture is complex systems can be divided into subsystems, this easy the maintenances of the whole systems. Creates a more robust system, ensuring that fail on one component doesn’t affect the other subsystems. The system has the potential to be scalable and be interoperable. It has the ability to build diverse and complex services by composing functions at the systems. This type of architecture comes with overhead and complexity of this techniques that are some of the disadvantages. You can see this architecture as building block to arise a big lego building. The purpose for this architecture is to encapsulate a repeatable IoT task as a service.
In the other hand the API architecture tries to give an alternative to the SOA architecture. This type of architectures gives the flexibility to connect solutions more efficiently ans fast according to the big variety of offering of API for IoT world. For example the API uses JSON data-exchange that are a more light type of format and can be more convenient to use for limited resources hardware(sensors, actuators) instead of the large XML documents used on SOA. Also the API architecture allow to fast add functionality to your application, for example if you need security functionality for your application you can connect your application for a provider that will allow you to focus more in the functionality rather than the presentation. The big advantage of IoT API architecture is that it allow to have connectivity for developer and users creating collaborative data information gathering, sharing and updating.
The IoT world has application in so many areas of the society ranging from home appliances to e-government applications but they can be classified in three main area according to :
- Monitoring and actuating: this area is bassically a smart device like a sensor that connects to an API application and it can request measurements every period of time so the client can see the state of the object and take action according to that or it can have an automatic program that can take the decision to send a signal to an actuator to take a desired action. In this field we can see application like electrical distribution smart-grid where data collection is used to detect anomalies on the grid and also take actions to reduce cost and save energy. Also this kind of applications can be seen on homes where several devices connect to an application and it tells you the how much have you spent in water and energy and help you to save and manage misuse.
- Business process and data analytics: This sections can be divided into three sub categories: 1. Society level, where the main focus is to improve government services by reducing costs and increasing transparency. 2. Industry level, where the main focus is to improve profit, reducing costs and increasing efficiency in the monitoring process. 3.Organizational level: which benefits are the same that in organizational level. 4. Individual level: where the focus are continuous improvement, individual efficiency, productivity growth and healthcare monitoring.
- Information gathering and collaborative consumption: Siot, social network Iot is the main focus for this area and his objective is to link our daily objects with our social media and create ease to interact with other people.
This has been a big concern and a major obstacle to the full development of the technology. Due to the nature of the Iot network and communication way there is several ways the information and privacy of the final user can be vulnerated. This, because sensors and smart devices has several constraints of performances which is a big obstacle to create robust transfer protocols like the used on the client server protocol, this opens a gap to the attacker which can access to the information in a easier way on the device side, so the attacks has been changed from the robust well protected servers to the smart devices side. Also, physical smart devices are ease accessible to the people making them even more vulnerable to attacks. Another problem to increase security on the Iot technology is that, due to the distributed nature and the number of the devices a constant actualization and patching is more difficult. The actual RFID protocols for the sensor network is less robust than the actual internet protocols and this makes the net more vulnerable to attacks. according to  there is several security concerns on the wireless sensor network (WSN):
- Spoofing: replaying and misdirection of the traffic
- Hello flood attack: this attacks creates false messages that satures the traffic and stops the service
- Homing: the network is searched for clusters heads and key managers which have the ability to shut down the system
- Selecting forwarding: the traffic is deviated to the desired address of the attacker
these are just a few examples of the atacks that can be done on the network layer but there is more vulnerabilities and the other layers of the transport protocol, so for this technology to reach his full potential there is huge necessity for an a development of a new way to protect the system.
 Internet of Things -Principles and Paradigms. Rajkumar Buyya . et al.
 Internet of Things (IoT): A vision, architectural elements, and future directions — Jayavardhana Gubbi et al.
 Survey of Security and Privacy Issues of Internet of Things- Tuhin Borgohain et al.
 wikipedia — internet of things. https://en.wikipedia.org/wiki/Internet_of_things