Nowadays, research in the field of IoT is gaining ground and rapidly increasing. There are some challenges in the field of Internet of Things such as: 1)Standards, 2)Mobility support, 3)Naming, 4)Transport protocol, 5)Traffic characterization and QoS support, 6)Authentication, 7)Data integrity, 8)Privacy, and 9) Digital forgetting. Until today, many papers have written in this field but In your opinion, what is the most challenging problem that IoT is facing and why?
Thank you in advance.
Hi Mehdi
Regarding the IoT, I will first comment on your list of 9 points:
1) Standards is always a problem, because as new technology is activited, the Stardards Bodies are usually very far behind. For example, cloud really got going in 2007, yet the International Standards Organisation (ISO) did not produce their first cloud standard until 2014. In the US, the National Institution of Standards and Technology (NIST) were much quicker off the starting block, with their draft standard produced in 2009, and a more comprehensive one followed in 2011. There are currently over 30 standards bodies engaged in some form of cloud standards;
2) Mobility support - is a particular issue for those IoT components which are located far from electricity supplies in remote locations. If the battery runs out on site, they stop working and it is very expensive to have then replaced;
3) Naming becomes a potential issue due to the sheer scale of the number of components that could be brought into use, with estimates of over 50 Billion in use by 2020;
4) Transport protocol - clearly this volume will present technical challenges, not least because of the scale of data that will be generated by the ever increasing volume of components in use;
5) Traffic characterisation and QOS support - again, the ever increasing volume presents a huge challenge;
6) Authentication - presents a challenge due to a combination of the reduced size of memory and processing power, the possible latency introduced by communications channels, and the sheer volume involved. In addition, some IoT devices have such small capability, that they can only transmit data one way, i.e., up the line;
7) Data integrity - will certainly be a challenge due to the weaknesses inherent in communications protocols, such as some Bluetooth and WiFi links;
8) Privacy presents an immediate challenge, since there is unlikely to be sufficient resource capability to encrypt all data collected at source before transmitting up the line;
9) Digital forgetting is less likely to be a problem, since the vast majority of data collected will be transported on up the line to permanent storage in a data centre somewhere else.
The one you have not mentioned - security, I consider to be the most challenging problem. There are a great many reasons for this. Clearly the limited resource capability of IoT components is a big concern here. The fact that many are sold with zero security on board by default is a big concern. Making a component live with zero security is just asking for trouble. A penetration tester colleague wanted to run a honeytrap server to test for Mirai infections. To his astonishment, he collected his first infection within 60 seconds of going live. Talking of Mirai, while it was designed specifically to infect IoT devices, it has also been ported to conventional desktops, meaning that once an IoT device is infected with Mirai, the attacker can now infect the whole infrastructure of the organisation with great ease, a worrying development. Many components have too few resources to be capable of adding proper security. Hacking IoT components is usually far easier than hacking conventional corporate systems. The use of Bluetooth and WiFi does not hel either. Often, their remote location can allow attackers to perform a physical intervention to gain access to the network. The data collected will be stored somewhere in a data warehouse up the line. If the database software in use has not been properly hardened, there is nothing to stop attackers injecting malicious code via the IoT devices. Security is such an issue that the Open Web Application Security Project (OWASP) has set up a project to address this. I have added a link for this. You should also know that if there is no security, it will prove extremely difficult to implement most of your 9 concerns properly.
I trust this gives you a flavour of the problems in IoT.
Regards
Bob
https://www.owasp.org/index.php/OWASP_Internet_of_Things_Project
By far the biggest issue, IMO, is orphaned devices. As a vendor, when you IoT enable a product, continuous support for updating the device to maintain security is required. It is especially a problem for any durable good ( major appliances, vehicles, etc ) and one which everyone seems to be blissfully ignoring.
So consider some IoT range top or washing machine or fridge or vehicle. will you really be getting software updates for these embedded devices 5, 10, 15 years down the road? Heck, there are many cheap IoT devices which are insecure from day one, and will never be patched.
Think of "classic cars" - I am sure 100 years from now, a 200 years old Model A ford will still be able to function. It is an independant mechanical contraption. Will a 20 year old self driving car be able to operate?
So there needs to be a stated product lifespan and support lifecycle after which the product should be recycled - and protections for consumers to ensure companies abide by those promises. This will ensure they engineer the products properly TODAY to prevent them from being a mess TOMORROW.
1. There is no standard set for IOT
2. Security issue because most of the IOT hardware's are cant update security updates.
Hi Mehdi
Regarding the IoT, I will first comment on your list of 9 points:
1) Standards is always a problem, because as new technology is activited, the Stardards Bodies are usually very far behind. For example, cloud really got going in 2007, yet the International Standards Organisation (ISO) did not produce their first cloud standard until 2014. In the US, the National Institution of Standards and Technology (NIST) were much quicker off the starting block, with their draft standard produced in 2009, and a more comprehensive one followed in 2011. There are currently over 30 standards bodies engaged in some form of cloud standards;
2) Mobility support - is a particular issue for those IoT components which are located far from electricity supplies in remote locations. If the battery runs out on site, they stop working and it is very expensive to have then replaced;
3) Naming becomes a potential issue due to the sheer scale of the number of components that could be brought into use, with estimates of over 50 Billion in use by 2020;
4) Transport protocol - clearly this volume will present technical challenges, not least because of the scale of data that will be generated by the ever increasing volume of components in use;
5) Traffic characterisation and QOS support - again, the ever increasing volume presents a huge challenge;
6) Authentication - presents a challenge due to a combination of the reduced size of memory and processing power, the possible latency introduced by communications channels, and the sheer volume involved. In addition, some IoT devices have such small capability, that they can only transmit data one way, i.e., up the line;
7) Data integrity - will certainly be a challenge due to the weaknesses inherent in communications protocols, such as some Bluetooth and WiFi links;
8) Privacy presents an immediate challenge, since there is unlikely to be sufficient resource capability to encrypt all data collected at source before transmitting up the line;
9) Digital forgetting is less likely to be a problem, since the vast majority of data collected will be transported on up the line to permanent storage in a data centre somewhere else.
The one you have not mentioned - security, I consider to be the most challenging problem. There are a great many reasons for this. Clearly the limited resource capability of IoT components is a big concern here. The fact that many are sold with zero security on board by default is a big concern. Making a component live with zero security is just asking for trouble. A penetration tester colleague wanted to run a honeytrap server to test for Mirai infections. To his astonishment, he collected his first infection within 60 seconds of going live. Talking of Mirai, while it was designed specifically to infect IoT devices, it has also been ported to conventional desktops, meaning that once an IoT device is infected with Mirai, the attacker can now infect the whole infrastructure of the organisation with great ease, a worrying development. Many components have too few resources to be capable of adding proper security. Hacking IoT components is usually far easier than hacking conventional corporate systems. The use of Bluetooth and WiFi does not hel either. Often, their remote location can allow attackers to perform a physical intervention to gain access to the network. The data collected will be stored somewhere in a data warehouse up the line. If the database software in use has not been properly hardened, there is nothing to stop attackers injecting malicious code via the IoT devices. Security is such an issue that the Open Web Application Security Project (OWASP) has set up a project to address this. I have added a link for this. You should also know that if there is no security, it will prove extremely difficult to implement most of your 9 concerns properly.
I trust this gives you a flavour of the problems in IoT.
Regards
Bob
https://www.owasp.org/index.php/OWASP_Internet_of_Things_Project
In my (biased :p) opinion one of the main issue faced by the IoT is the lack of transparency. We have published a number of paper on related subject.
https://scholar.harvard.edu/files/tfjmp/files/2016iotjnl.pdf
http://scholar.harvard.edu/files/tfjmp/files/2016bipreprint.pdf
https://scholar.harvard.edu/files/tfjmp/files/main.pdf
Security aspects are within the main problems and challenges for IoT, and one example is that parents would never feel comfortable putting sensors or cameras or any type of electronic devices in their babies rooms without knowing to a high degree that these systems are secured and no one can hack them to mess with their children health or life.
I agree security is the most challenging problem that IoT is facing today. Blockchains for IOT help solve the issue. See our research paper.
Conference Paper Zenith Certifier: A Framework to Authenticate Academic Verif...
Security (6,7,8), Energy management (2), Traffic characterizatiion and modifications of TCP mechanisms to adapt it to IoT (4,5)
- Badges
- Science topic