Example usage:

> import network_generator
> nwkgen = network_generator.NetworkGenerator(50)

The parameter to NetworkGenerator is the size (in kms) of the deployment area.

> nwkgen.generate_demographic((25,25), 5)

This adds a population of UEs distributed following a 2D Gaussian distribution centered at (25,25) kms, with a standard deviation of 5kms.

> infra = nwkgen.generate_5g_infra(1)

This generates the 5G infrastructure to support the population of UEs created previously. The argument provided is a random seed.

> g = infra.generate_nx_graph()

g contains two types of nodes:

• computational nodes provide the following attributes:
  • 'type' is 'pop' (point of presence)
  • 'cores': number of processing cores
  • 'storage': TB of storage
  • 'gpus': number of gpus for cloud gaming (or other apps)
  • 'traffic': input rate into the node.
• access points with the following attributes:
  • 'type' is 'sap'
  • 'tas': a list of Tracking Areas aggregated into this AP
  • 'traffic': (input rate, output rate).

The edges of the graph represent communication links with the following attributes:

• 'capacity' in Mbps of the link
• 'latency' of the link (just the transport, does not consider queues at the two endpoints of the link).

The figure below shows a sample of an infrastructure graph. The red nodes are PoPs, the blue are SAPs.