We often have questions related to edge, fog, mist and cloud computing. Is it just a question of marketing never-ending creativity? Is there a technical reality beyond those terms? Is one approach better than another one?
Cloud computing is a very centralized paradigm where most data analysis/processes are realized on centralized servers that are located all over the world . It is the (in)famous cloud where you can compute and store data without the need to know where are your data and processes.
The main benefit is that you can scale easily your processes and your data storage at a low cost.
Internet (and its very robust and efficient TCP/IP protocols) is the only remaining global network (but a few years ago, there were other types of prehistoric network such as X25 in France for example) that “links” applications, databases, servers, smartphones and even smart objects with growing interest for 6LowPAN.
What’s wrong with Cloud Computing?
But this approach does not fit well with smart cities and smart industries (often called “industry 4.0”) and a more decentralized approach called edge computing or fog computing (actually, they describe the same thing) is more adapted. Why?
- Latency and Response time is often a critical part, especially when you deal with human life or emergency procedure.
- Bandwidth Cost and Capacity is very often underestimated. If you want to use N smart devices requiring each one to communicate M bytes of data then you can quickly reach huge bandwidth requirements reaching Mbit/s or even Gbit/s at a gateway level. This is totally unrealistic in terms of cost and physical capacity.
- Security and Privacy are raising concerns among citizens and industrial organisations that cannot accept anymore to have data and processes located anywhere in unknown worldwide servers. In addition, a very centralized area is very attractive and often easy target for hackers. A more local and decentralized approach makes more senses.
- Power consumption (environmental issue) is a very well hidden and sensitive secret from big pure cloud players. Once upon a time, Google and others were ready to disclose how many servers were running but it’s not the case anymore. What is sure is that Cloud computing is energy-hungry and that it is a concern for a low-carbon economy.
- Data obesity – In a traditional cloud approach, huge amount of untreated data are pumped blindly into the cloud that it is supposed to have magical algorithms written by data scientists. This vision is really not the best efficient and it is much more wise to pre-treat data at a local level and to limit the cloud processes at the strict minimum.
- Offline usages versus only-online usages – Last but not the least, pure cloud services do not allow offline usages. It is a major shortcoming since smart cities and industry 4.0 applications require a dual offline/online paradigm.
To overcome all those cloud limitations, even though edge/fog computing bring positive inputs, another approach is still required.
MIST Computing: the perfect technical architecture paradigm?
Mist computing might be THE solution bridging the gap between central cloud and edge decentralization. A mist device is an edge/fog device with super-powers similar to the ones found in cloud servers. For example, our mist devices allow:
- Local analytics and decision-making data using big data DB and processes
- Learning and adapting machines
- Redundant, Resilient and Highly Robust data and apps
- Root of trust similar to HSM modules found a cloud server area
- Data access control mechanism to enforce privacy consent at a local level and security & privacy by design / by default
Obviously, Mist and Cloud computing are complementary to each other as offline and online usages are complementary to each other. Nevertheless, in a perfect distributed architecture, the very interesting idea is that a mist application does not require the cloud and vice-versa!