Simulation: Add endpoint simulation

[jonas-rem]

This commit adds the script simulate.py to build and run simulated Zephyr endpoints on a host PC. The endpoints can connect to a cloud or to a locally running flownexus instance.

• The endpoints can be started with a time delay to test how much load the cloud can handle.
• Build/run options.
• Logging enable/disable
• Number of endpoints can be specified (max 254 currently)

Preliminary performance measures:

• The server can currently handle 10 - 20 endpoint registrations per second on a Virtual Server (1 Core / 1 GB Ram)
• Having 254 Endpoints with a registration interval of 30 s connected did not caused significant CPU load on the server.
• Memory consumption for 254 registered endpoints is not noticeable.
• There is room for improvement by reducing logging, migrating the database to PostgreSQL and upgrading the server hardware capability.

Draft until the following points are implemented:

• [x] Add documentation
• [x] Use a container to execute the Zephyr clients and setup the zeth network nodes so that the simulation will reliably run on any platform.
• [x] Fix issue with DNS resolve
• [x] Fix connecting to local simulation

[jonas-rem]

@Kappuccino111 Just updated the simulation. It now handles the virtual network devices and the Zephyr endpoints in a Container for interoperability.

Check by executing the simulate.py script

python3 simulate.py --help
usage: simulate.py [-h] [-n NUM_CLIENTS] [-v] [-b] [-r] [-d DELAY] [-l]

Zephyr Build and Run Script

options:
  -h, --help      show this help message and exit
  -n NUM_CLIENTS  Number of client instances to start. (1 - 254) (default: 1)
  -v              Enable logging (default: False)
  -b              Build the client (default: False)
  -r              Run the client. (default: False)
  -d DELAY        Client start delay [ms] (default: 0)
  -l              Connect to locally running Leshan server (default: False)