The Internet of Things

7 min readNov 26, 2018

A future of ubiquitous computing — the physical world integrating with an array of connected devices [4]

Defining the Internet of Things

The Internet of Things (IoT) is a term to describe the collection of computing devices that seamlessly integrate into the world around us, almost to the point of being invisible.

Such devices contain sensors to harvest data from the world these objects exist inside. The special thing about all the data harvested is its ability to provide some utility for an end user. Utility is generated from models that can can extrapolate information from the data and enhance our interaction with the world around us. Basically, the data from these devices is used to provide our life with useful information.

Smart refrigerators, thermostats, doorbells, and watches are just a few examples of devices being connected to the internet. Gone are the days where a device didn’t have any sort of connectivity to the internet. The internet enables these devices to communicate with us so we can figure out interesting things to do with them. Wireless connectivity exposes these devices to the global area of the internet, effectively stripping away their previous insular nature. Refrigerators can be programmed to tell when you’re running low on items and thermostats can automatically adjust the temperature based on your presence inside the home, to name a few relevant examples.

Mark Weiser, Chief Scientist at Xerox Parc during the 1980s, described it with the term “ubiquitous computing” and compared the proliferation of computing devices to the compilation of words on tangible technology. The meaning these words convey on paper, whiteboards, signs, and billboards is understood without requiring to us think for more than a few seconds. That same philosophy was kept in mind when imagining a future that blanketed computing devices into the world around us.

IoT’s explosion goes hand in hand with the development of RFID technology. Computing devices, as is, are not that interesting in the future of IoT. It is their fusion with ID tags that makes them powerful in the vision of pervasive computing [8]. Kevin Ashton put it best during his time at Auto-ID Labs, sensors and RFID technology enable computers to interact with the environment without the intervention of humans [9].

The role of RFID is significant in the fact that objects can be identified based on a unique tag that responds to radio waves. These tags can be either passive or active, and the difference lies in battery usage. Passive tags can store a code and most importantly, the location of the tagged object [10]. Users have the ability to track devices and gather information from them. As an example, businesses can use these tags for monitoring inventory in real time. The growth of IoT is facilitated, in part, by the deployment of RFID technology.

The combination of sensors and RFID tags are a core part of the IoT movement. Smart devices in this movement use sensors to collect the data and RFID tags to communicate that data. The RFID readers that communicate with the tags are connected to the internet. These readers are the middlemen between devices and the internet [11].

On the consumer side of this growing space, the Internet of Things is being facilitated by the development of “smart” devices. There are various ecosystems currently available on the market ranging from players like Google, Amazon, Apple, and Samsung. These ecosystems generally consist of a hub and compatible devices[4].

Home automation and the interplay of smart devices were becoming popular well before the major players entered the space. IFTTT, If This Then That, is a platform that allows users to build their own Applets, that do something in response to some kind of cause. Triggers are connected to results, so a trigger may be data from an infrared sensor that hasn’t registered any movement in a house, and the ultimate result may be to turn the lights off. An example Applet involves using the Nest Thermostat to detect if a temperature is below a threshold and activate a bed warmer so you can be comfortable when you get home [5].

The future of IoT lies in the proliferation of such devices, cloud platforms that host the harvested data, and the ubiquity of data’s availability [6]. Instead of individuals seeking information first, relevant information will be brought to us based on context. Reality will be augmented with data from IoT devices and force the paradigm to shift away from traditional computing. The paradigm of computing will shift away from user driven searches to context driven computing.

Security Issues

The collection of data and its communication are powerful tools for the end user but those functionalities do not adhere to any kind of global IoT security standard.

Vendors are free to create their own platforms for hosting devices and providing utilitarian functionality but struggle to keep them free of security holes. Since IoT is blowing up, companies are striving to pivot their focus to products that are connected to the internet but neglect the security ends of their products [3].

In a 2015 research paper published in the International Journal of Computer Applications, the Internet of Things can be divided into a perception layer, network layer, middle-ware layer, and an application layer [12]. Before diving into the security issues in all these layers, let’s take a moment to understand each layer’s purpose. The perception layer concerns itself with the sensors and RFID tags, basically a device’s ability to “perceive” its environment. The Network layer concerns itself with the transportation of data from the sensors to a hosting service. Lastly, the middleware and application layers host the data and perform computational operations on it.

The first avenue of security issues comes from the integrated sensors and RFID tags. The tags are susceptible to eavesdropping, unauthorized access, and jamming. Most RFID systems do not implement a proper authentication mechanism which can allow unauthorized users to access information stored in the tags. Hackers are able to create replicas of the RFID tagged objects to fool RF readers into seeing false data [12].

A number of possible solutions involve encrypting the data before transport, anonymizing key aspects of the data, and detecting intrusions.

Moving away from the layer involving sensors and tags, the transmission of the data from objects to a location storing the data. Data that should remain private can be accessed by unauthorized agents. This is significant because it impacts the reliability of data gathered from IoT devices. Wireless channels are susceptible to individuals sniffing them for private data.

Implementing some security in the routing of the data from a tagged object to a database storing that information is critical. Ensuring that the data at the beginning of the journey is the same at the end is key because it prevents that data from being manipulated in the middle of transport. Additional security measures involve examining the data as it is being transported. This means that some form of error checking should exist along the way to the destination [12].

The last avenue of security stems from the services that host all the harvested data. These services are susceptible to attacks that take down the entirety of hosting services. Much of that can be achieved through Denial of Service attacks and those that disable access to the entire system. In addition, let’s not forget the human element involved. Ensuring that companies implement policies preventing wholesale access to data will mitigate instances where unauthorized individuals obtain access to the information [12].

Solutions specific to this zone of the IoT layer are similar to the previous layers. Since companies are offering services that host all the data, it is important that they have policies on hand for handling intrusions, unauthorized access, and data redundancy. The assumption is that all the data harvested from sensors is transported to hosting services and perhaps that might not be the right approach. In order to prevent private data from being intercepted by malicious agents, it may be prudent to allow data filtration, only allowing certain data to be stored on the hosting service [12].

Wrapping Up

We’ve covered three major aspects of IoT: devices, services, and security. Devices are the bedrock from which services are able to store harvested data. Ultimately, the joint operation of sensors and RFID technologies enable devices to exist outside the scope of incredibly private networks. This allows individuals to perform tasks like turning off lights away from home or purchasing a connected thermostat that uses machine learning models to predict indoor temperatures.

Yet, with all the hype and positivity of a world where computing devices are embedded into the environment, there are issues with security. The internet is prone to malicious agents and networks are susceptible to snooping. Great vigilance is required from the services hosting data from computing devices. A more connected world requires a privacy centric focus, one which will require concrete implementations of privacy measures.

Mark Weiser’s world of pervasive computing and Kevin Ashton’s world of networked objects are the future. But, to build this future in the correct and safest way will require standards to maintain security and the usability of connected devices.