Infrastructure

class infrastructure.Infrastructure

Bases: PowerAware

Infrastructure graph of the simulated scenario.

The infrastructure is a weighted, directed multigraph where every node contains a Node and every edge between contains a Link.

Add a link to the infrastructure. Missing nodes will be added automatically.

add_node(node: Node)

Adds a node to the infrastructure.

Return all links in the infrastructure, optionally filtered by class.

measure_power() → PowerMeasurement

Returns the power that is currently used by the entity.

node(node_name: str) → Node

Return a specific node by name.

nodes(type_filter: Optional[Union[Type[_TNode], Tuple[Type[_TNode], ...]]] = None) → List[_TNode]

Return all nodes in the infrastructure, optionally filtered by class.

remove_node(node: Node)

Removes a node from the infrastructure.

Bases: PowerAware

A network link in the infrastructure graph.

This can represent any kind of network link, e.g.

  • direct cable connections

  • wireless connections such as WiFi, Bluetooth, LoRaWAN, 4G LTE, 5G, …

  • entire wide area network connections that incorporate different networking equipment you do not want to model explicitly.

Parameters:

  • src – Source node of the network link.

  • dst – Target node of the network link.

  • bandwidth – Bandwidth provided by the network link.

  • power_model – Power model which determines the power usage of the link.

  • latency – Latency of the network link which can be used to implement routing policies.

measure_power() → PowerMeasurement

Returns the power that is currently used by the entity.

class infrastructure.Node(name: str, cu: Optional[float] = None, power_model: Optional[PowerModelNode] = None, location: Optional[Location] = None)

Bases: PowerAware

A compute node in the infrastructure graph.

This can represent any kind of node, e.g. - simple sensors without processing capabilities - resource constrained nodes fog computing nodes - mobile nodes like cars or smartphones - entire data centers with virtually unlimited resources

Parameters:

  • name – Name of the node. This is used to refer to nodes when defining links.

  • cu – Maximum processing power the node provides in “compute units”, a imaginary unit for computational power to express differences between hardware platforms. If None, the node has unlimited processing power.

  • power_model – Power model which determines the power usage of the node.

  • location – The (x,y) coordinates of the node

measure_power() → PowerMeasurement

Returns the power that is currently used by the entity.

utilization() → float

Return the current utilization of the resource in the range [0, 1].