Explaining Series: Fog Computing in the Internet of Things

Explaining Series: Fog Computing in the Internet of Things

Fog Computing -one of many new trendy terms that we see and read almost everywhere in this field.

What is it? -and how can IOTA enable the perfect fog-computing landscape, the IoT needs?

I give you a short explanation and good sources for a smooth heads-up.


Roundup:

This roundup is an experiment that aims for a better understanding of the greater picture. Some keywords before the actual article are meant as an information-index.

  • Internet of Things (IoT)= Term from the MIT, Kevin Ashton, 1999
  • Fog-Computing = Term from Cisco
  • Fog = Decentralized/Distributed
  • Cloud = Centralized
  • Realm = IoT + IIOT, B2B, M2M, IoE, Smart grids, Smarthome, Smart cities, interconnected world
  • Problem = Unused Sensor-Data, Need for a solution of a distributed network, Costs of cloud-computing, time
  • Application = Evolving Markets, Quality-as-a-service, Machine Communication, Scada
  • IoT Systems = Basically two groups: 1) Identification Group (sensors, data gathering) 2) Computational group (processing, data storage)
  • Limitations until now: Cloud computing (centralized, far away from consumers and devices)  doesn’t fit the requirements of the IoT (distributed, in need of close storage, computational resources, instant processing), Bandwidth
  • Connection Types: WiFi, Bluetooth, ZigBee, 2G/3G/4G/5G, Radio, Z-Wave, 6LowPan, Thread, Wifi, Cellular, NFC, Sigfox, Neul, LoraWan

The IoT

The vision of the Internet of Things is still in the making.

With the latest development in this interconnected world, new markets are emerging and a variety of requirements are born.

Wearables, smartphones, domestic devices like smart-home solutions for an intelligent household demand an interconnectivity solution that has yet to come.

It’s no secret that almost every company is also working on solutions to make it happen: a world, where data is a more valuable resource than oil. If not today, then in the near future.

This leads to a point, where technical barriers of today hinder progress for tomorrow.

The IoT, a distributed network around the world is more than the Internet.

A mesh-net that is connected with every possible connection type. Where devices work in local clusters, it’s obvious that centralized components, sometimes on a different continent, don’t fit in the greater picture.

Sensors, cameras, smart devices often use ad-hoc solutions to function in their specific field, such as monitoring systems like Scada, that send valuable data to a nearby control center in order to optimize industrial processes.

What if these monitoring systems are working time-sensitive, but the current solutions are slow and on top of that centralized and unsecured. The productivity could be better, employees may work in a more dangerous environment and as a result: the company could face problems.

Connected facilities incentivize industry-espionage and hacks.

Distributed denial of service attacks is a phenomenon of the last few years, where certain malicious parties are attacking infrastructural points in the web, to cripple communication of some systems and special services.

Sometimes as a decoy for a hack, sometimes for political or activist-reasons.

Not rarely, mentionable down times create financial losses or the blockage of regional infrastructure’s hits, next to the target, also other companies that are located in the surrounding area.

A problem of the Internet, not necessarily of the IoT.

Due to the distributed mesh-net characteristics, the IoT is envisioned as a network, that is self-sufficient, in which case it can connect devices of the identification group via many ways, not only one.

An attack on central points is per definition impossible because there is no center in the IoT.

That leads to a natural resistance against DDoS and other downtimes.

Legacy systems vs. new systems

An additional issue of cloud-computing in the IoT would be the costs. Legacy system use to ignore huge amounts of data because there is neither storage no need for them.

New systems in the IoT, with smart solutions, rely on this data, but sending them into the cloud would go beyond the scope of the IoT. Too much information is generated, and real-time analysis, as well as centralized cloud-computing solutions, are conflictive with each other as uploading these huge amounts takes time and money -especially if the cloud-storage is thousands of miles away.

Fog computing, however, creates a bridge-solution for the identification group and computation group: It is about forwarding the computational power to the edge of the network, where data is generated and the results are needed.

The benefits of using Fog computing instead of legacy cloud systems are tremendous.

Varghese, Wang, et. al [2017]. come to the conclusion that. “For an online game use-case, we found that the average response time for a user is improved by 20% when using the edge of the network in comparison to using a cloud-only model. It was also observed that the volume of traffic between the edge and the cloud server is reduced by over 90% for the use-case.

This is just one use case that can be mirrored on many other settings.

In consumer markets, Quality of Service and Quality of Experience are important factors.

Another example would be the transparent customer. When a transparent customer enters a big supermarket, his views and interests could be analyzed within seconds.

Cameras can detect his interest in certain devices or components, and advertisements on monitors along his path can be adjusted to his specific needs. With old legacy systems impossible due to the long processing times between these cameras, a cloud, and computational resources, with fog computing, however, the data can be processed way faster and deliver the necessary information back to the customer, along his way in the mall.

To draw a simplified picture of the fog-landscape:

The distributed mesh-net is growing in height z, if you will, whereas decentralized and centralized networks are growing on the x and y axis. Shorter ways from the data collectors to the computational resources are the result of fog-computing.

Concerns can be addressed with IOTA

Whether it’s the data-integrity, optimization or protection of the in-house Research & Development data, companies look for a lasting solution.

When data is stored centralized, hackers usually use social engineering, or phishing attacks to get access to the data.

As centrally stored data would be collected all in once with this method, Fog computing would make it possible to store sensitive information in small packets, distributed, with different passwords/keys/seeds to access them.

IOTA  can deliver a unique solution here. A data-stream, bound to countless seeds, in a distributed network, secured with sophisticated algorithms. Not even quantum computing would be a threat to the hashes.

As you may already know, IOTA is a distributed ledger technology, that enables fee free transactions.

For data-transfer with fog-computing, you wouldn’t even need tokens, the only condition would be to confirm two other transactions before sending one of your own.

A rule that enables true scalability for a billion device network on a global scale.

With Masked Authenticated Messaging, IOTA has an additional option to send and process sensitive data.

Now, a really big hurdle in the IoT is the availability of dozens of connections and different norms.

When devices could be connected in a similar way, the usability would increase. A plethora of standards that are built for the IoT can lead to a fragmentation of the network, as companies want to stick to their standards, to support their product-line or roadmap.

If IOTA would be the standard settlement and data layer, which is free to use, the Internet of Things could be a barrier-less environment with true scalability and data-integrity.

Due to the value of collected data, new markets would come up, that aim for selling this information in real-time.

People would possibly be able to sell their consumer data, each time they enter a shop, with true nano payments.

If data would be collected in the fog, BigChain DB a scalable distributed database for all kinds of data could deliver the necessary infrastructure for customers, institutes, and companies.

A seamless solution for the IoT.

Fog computing is, therefore, the next necessary milestone in the field of the Internet of Everything and a vital part of the vision of IOTA.

 

Video of Dominik Schieners Presentation at the Tech Open Air 2017

 

 

 


Sources:

https://www.cisco.com/c/dam/en_us/solutions/trends/iot/docs/computing-overview.pdf

http://www.springer.com/de/book/9783319576381

https://arxiv.org/pdf/1701.05451.pdf

 

Images:

<https://www.iot-now.com/2016/01/11/40800-connectivity-on-the-edge/>

 

Leave a Reply

Your email address will not be published. Required fields are marked *

Advertisment ad adsense adlogger