The missing piece in decentralizing agriculture—and many other industries

Blockchains on mobile, IoT devices: Can fog computing make it happen?

According to researchers from the IEEE, it’s plausible.

Decentralize everything

Recently, there have been two main terms being thrown around that are slowly seeping into the blockchain industry: edge computing and fog computing. These two refer to a system that would supposedly supersede what we currently know as cloud computing. And the shortest way to explain what the upgrade will entail is that edge and fog computing will decentralize cloud computing.

How “centralized” is cloud computing exactly? You see, at the current state of things, servers in a cloud work in such way that there is a central server or computer provided by a certain service—Google or Amazon, for example.

“As centralized as this all sounds, the truly amazing thing about cloud computing is that a seriously large percentage of all companies in the world now rely on the infrastructure, hosting, machine learning, and compute power of a very select few cloud providers: Amazon, Microsoft, Google, and IBM,” Paul Miller wrote for the Verge.

Instead of centralized servers, fog computing uses a distributed network of storage and computing power providers. Not only is it far more difficult to take down compared to a central server, it’s also more secure: hackers looking to instigate an exploit to acquire specific data will definitely have a much harder time doing so—they wouldn’t even know which server to target. Having said this, the larger the network of computers becomes, the stronger its security gets.

From the cloud to the fog

Edge computing is a way to bring the processing center closer to the source of data, or the “edge,” significantly cutting down costs and processing time by tapping on a network of computers who are offering their storage and processing power to the network’s clients in exchange for pay. Edge computing doesn’t necessarily need to be blockchain-based, but in several ways, the two technologies overlap.

In essence, they’re like blockchain miners, except anyone can use their processing power for any process at any given time—it could be mining, scientific calculations, video streaming, or anything else. Unlike blockchains, edge computing services are not limited to a specific use case.

The quickest differentiator I’ve seen between edge and fog is from Cisco: “Fog computing is a standard that defines how edge computing should work, and it facilitates the operation of compute, storage and networking services between end devices and cloud computing data centers. Additionally, many use fog as a jumping-off point for edge computing.”

Fog computing is another emerging technology that can make blockchains even more powerful than they already are. Fog computing, which Cisco describes as a way to “extend the Cloud to where the things are,” is actually already creeping into the blockchain space. At this point, some may already have an inkling about how this relates to blockchain and processing. But apart from the obvious initial assumptions, one of its bigger implications brings blockchain applications to an even higher ubiquity level: mobile and Internet of Things (IoT) devices.

Mobile edge computing (MEC): blockchain applications for mobile and IoT devices

While mobile cryptocurrency wallets are already common, other mobile blockchain applications remain on hold due to technical challenges. So can fog computing really make blockchain applications feasible on little mobile and IoT devices? According to researchers from the Institute of Electrical and Electronics Engineers (IEEE), it’s a viable solution.

“Although blockchain has been widely adopted in many applications, e.g., finance, healthcare, and logistics, its application in mobile services is still limited. This is due to the fact that blockchain users need to solve preset proof-of-work puzzles to add new data, i.e., a block, to the blockchain. Solving the proof-of-work, however, consumes substantial resources in terms of CPU time and energy, which is not suitable for resource-limited mobile devices,” Zehui, Yang, Dusit, Ping, Zhu (2018) wrote in their paper.

“To facilitate blockchain applications in future mobile Internet of Things systems, multiple access mobile edge computing appears to be an auspicious solution to solve the proof-of-work puzzles for mobile users.”

SONM ‘fog supercomputer’ vs Cloud computing

One company that is shooting to bring the promises of fog computing is SONM, which plans to create a supercomputer using the collective computing power of all the computers within their fog network. I asked SONM co-founder, Aleksei Antonov, how much the projected impact of this is.

Comparing the costs of fog computing to Amazon’s cloud computing, Antonov says the difference is quite enormous.

“While a consumer of computing resources can rent a NVIDIA Tesla V100 (P3.2xlarge) module for $660 a month at a hash rate of 94 MH/s, the rental of a comparable volume of computing power via the SONM marketplace will cost around $137 a month. The reason is simple: the suppliers of computing resources on the SONM platform are PC owners with NVIDIA GeForce GTX 1080 GPUs. The savings for consumers of SONM computing resources can reach 80% in comparison with services provided by cloud platforms,” Antonov wrote in an email.

“For those renting computing resources, using SONM means making savings. Let’s look at the effectiveness of SONM for mining. The rental of equipment with a hash rate of 94 MH/s will cost $0.918 an hour on a AWS cloud platform. Renting that very same capacity on the SONM platform will cost $0.19 an hour. This means the cost of mining when renting SONM fog computing is 4.8 times cheaper than the cost of using the services of cloud platforms,” he explains.

Millions of Teraflops Ahead

From what I understand on SONM’s website, an institution or individual can tap into the network to gain an unprecedented amount of computing power — which would supposedly enable them to conduct certain activities that they couldn’t do before (unless they spent vast amounts of money). I assume this is dependent on how big the network of hardware providers is. But do they have any estimates on how much power this will be?

“According to our calculations, the aggregated capacity of devices used by GPU miners around the world is 30 million TFlops (teraflops). This is the volume of computing resources that GPU owners can provide for rental via the SONM marketplace. For comparison, the productive capacity of Cray Inc’s most powerful super computer is 259,900 TFlops.”

Imperfect technology under development

While fog computing has its promises, like all new technologies, there are several vulnerabilities ahead, some of which even professionals may not foresee until they are rigorously tested.

“The chief risk is the threat to confidentiality and data authenticity posed by a fog computing node. The supplier of computing resources has access to all the information about running tasks. They can receive code and confidential data tasks. What’s more, the supplier of fog computing can even emulate tasks and give the consumer false computing results,” Antonov admits.

To mitigate any potential risks, the company is tapping on the online community through a developer challenge. Such bounties are common to ensure white hat hackers weed out any possible risks, lest malicious entities get to them first.


Cecille de Jesus

New to blockchain? Check out CoinGeek’s Blockchain for Beginners section, the ultimate resource guide to learn more about blockchain technology.