ABC Cloud

1. Summary

The cloud is the core of ABC ecosystem. It consists of two main parts: server cluster and high speed telecommunication such as 6G. Before proceeding, readers are strongly recommended to read articles below.

• Edge computing acts on data at the source[Text | Video]

• 6G - What lies beyond 5G network technology?[Youtube]

• Towards 6G wireless communication networks: vision, enabling technologies, and new paradigm shifts†[PDF Link].

• Positioning in 5G and 6G Networks—A Survey[PDF Link]

• 6G White Paper on Localization and Sensing[PDF Link]

2. 6G & Cloud Edge Computing

The most important technology is 6G, and China is a prominent nations which lead 6G technologies.

• The cloud is where most computing processes are done, and devices are mainly for user interface.

• Almost all kinds of devices, machines and facilities are linux machines at the same time.

• All linux machines are connected to the cloud via 6G network.

For example, assuming Alice is a citizen of an ABC city, she has a linux smartphone, "Linux in Edge", and a linux distro instantiated in the cloud, "Linux in Cloud". So every citizen or machine in the city has at least two linuxes, one in his/her/its edge and the other in the cloud.

"Linux in Edge" is a gateway for Alice to enter the web via her "Linux in Cloud", and all her activities in the web is recorded by her "Linux in Cloud" onto a gitlab server in the cloud. All the same is every citizen or machine in the city.

As said, most computing processes of Alice are performed by "Linux in Cloud". The main role of "Linux in Edge" is to display data from and to send sensing data to her "Linux in Cloud".

The problem is how to distribute the data payload on the network to avoid the expected bottleneck in the cloud.

1. ABC Distributed Cloud

Our solution is to spread the cloud across many small cloud cells. For example, let's assume an ABC city with 1,000 6G cells, 200,000 citizens and 300,000 electronics including personal computing devices, self-driving cars and so on. We call those citizens and devices as edges So, there are 500,000 linux machines for 500,000 edges, and 1,000 cells, in the city.

• Assign a server to each of 1,000 cells.

• Assign 200 citizens and 300 electronics to each server. So, each server has at least 500 users.

• Instantiate linux distros at each server as many as its users. So, each server hast at least 500 linux distros each of which represents a user either citizen or machine.

• At each server, install most applications necessary for operations of machines and works of citizens. For example, a cloud-based word processing application installed at a server can be shared by 200 citizens.

• Install 6G networking modules, DNS modules and a Gitlab server etc. on each of the 500 servers.

Pseudo Code

2. Edges

Edges are linux machines with limited computing power. They include computing devices for human and those for machines.

Their main role is to display data from the cloud and to send sensor data to the cloud. For example, Alice's smartwatch may detect her abrupt increase of body temperature and heartbeats, and send the data to the cloud. The cloud may start countermeasures for this detection such as sending medical staffs to her.

Some examples of edges are;

• Personal computing devices such as smartwatches, smartphones, tablets, smart rings, and desktops etc.

• Machines such as refrigerators, washing machines, self-driving cars etc.

• Facilities such as houses, factories, shops and buildings etc.

• Legal entities such as corporations and organizations etc.

Pseudo Code

3. GitLab & Openhash

Continuing the previous example, let's organize the assumed 1,000 servers of the city in a hierarchical manner.

• Each server is both a linux machine and a 6G cell to server approximately 500+ users at the same time.

• Every citizen, facility, legal entity and machine etc. belong to ONE and ONLY ONE server, a 6G cell, in the cloud. For example, Alice belongs to one of the 1,000 servers in the cloud. That server is called DOMICILE of Alice. So, every entity in the city has his/her/its domicile server.

• A server has one gitlab server which openhashes its 500+ users' linux distros, "Linux in Cloud". That is, a server in the cloud has 500+ linux distros and 1+ gitlab server(s).

• A user may access to his/her/its linux distro in the cloud via a linux device such as a smartphone.

• Each linux distro in a server openhashes its log file which reflects the activities of its user, and periodically merges that log file into a branch in the gitlab server. For example, Alice's all activities are recorded in the log file of her linux distro in the server via the Openhash mechanism, and her linux distro merges her log file into a branch of the gitlab server in the domicile server which she belongs to.

So, there are 1,000+ servers in the city, and each server has 500+ linux distros and 1+ gitlab server which openhashes the log files of the 500+ linux distros each of which again openhashes the log file of its user such as Alice.

Again, there is ONE city server which openhashes the 1,000+ servers. Then, how many and where openhash processes are done in the city?

• Firstly, the linux distro of a citizen or other entity openhashes its log file.

• Secondly, a domicile server openhashes its 500+ log files of the linux distros it has.

• Thirdly, the city server openhashes the 1,000+ domicile servers in the city.

Pseudo Code

4. Machine Learning

Linux machines for personal usages such as smartphones and PCs have limited computing power. So, most computing processes are done in the ABC Distributed Cloud. But, even in distributed clouds, this computing architecture may substantially increase data communications between the cloud and edges.

• Increase computing powers of edges will increase the cost of manufacturing those edges.

• Minimizing computing powers of edges will increase data traffics between edges and the cloud, and it also increase the cost of network operation.

We can use ML to find out optimal balance between the computing powers of edges and the data load of the network. Moreover, ML can be utilized in determining what kinds of computing processes had better be done in edges.