All listed by you - means of remote control. They can be used to build, as a centralized management system (agent-manager, client-server), and distributed. The fact that the admin can connect to any machine from his workplace does not in itself negate the fact that each machine will have its own administrator account and each of the machines on the network will have to be configured separately. But at the same time, no one forbids connecting over the same SSH, for example, to a RADIUS server or a Cisco controller and changing the settings of all WiFi points at once, without going to each of them individually. Using the SNMP protocol, which you cited as an example, you can build a centralized management and monitoring system, but by itself it is not.
The essence of a centralized management system is that all subordinate network nodes are managed by one or a group of servers. Those. you make changes to the parameters on the server, and the agents independently (or after a kick from the server) pick them up and apply them. At the same time, it is practically not important which protocols will be used for this, It depends on the tasks solved by a specific control system. The control system is a complex of measures and means, but not as one specific protocol or technology.
If we already give an example of centralized management, then as an option ActiveDirectory and / or * nix-analogs with full control of all stations in the subordinate network, centralized configuration of the OS, etc., or the same Kaspersky with its agents and administration server. If you do not have strict requirements for specific solutions - look towards the well-known brands MS, HP, Cisco, Citrix, etc., they, as a rule, have fairly detailed descriptions of architectural solutions from which it is possible to make a squeeze for a diploma.
