Minimum spanning tree

[Raghad1997]

@wisaaco Good morning, I have a question, is it possible to imply the minimum spanning tree algorithm? in the simulation?

for example, I have an orchestrator and nodes with their edges and I want the MST ?

Can we implement that ? in YAFS simulation

[wisaaco]

Hi, You don't need to implement the MST. It is already implemented in the networkx library. The G variable of the topology class is a networkx graph.

https://networkx.github.io/documentation/networkx-1.10/reference/algorithms.mst.html

[Raghad1997]

My other Question I have a physical topology so I want to build a logical topology ( based on the weight of links and node costs) from the physical can we apply this requirement in the simulation? if yes? how we can do that

[Raghad1997]

I want the logical topology to implement the MST to remove the cycle

@wisaaco

[wisaaco]

Hello,

In the topology, you can define as many attributes as you want. Networkx is a powerful library with multiples implemented graph algorithms. An example:

    t = Topology()
    #A grid
    t.G = nx.grid_2d_graph(10, 10, periodic=False)

    # for visualization purporses
    lat = {}
    lng = {}
    for k in t.G.nodes():
        lat[k[1] + k[0] * 10] = k[0]
        lng[k[1] + k[0] * 10] = k[1]

    li = {x: int(x[1] + x[0] * 10) for x in t.G.nodes}
    nx.relabel_nodes(t.G, li, False)  # Transform str-labels to int-labels

    # Adding diaganols
    for i in range(100):
        if i < 89 and not i in range(9, 100, 10):
            t.G.add_edge(i, i + 11)
        if i > 0 and i < 90:
            t.G.add_edge(i, i + 9)

   # Lat and lng attributes are used for visualization
    nx.set_node_attributes(t.G, values=lat, name="lat")
    nx.set_node_attributes(t.G, values=lng, name="lng")

    # Node mandatory attributes
    valuesOne = dict(itertools.izip(t.G.nodes(), np.ones(len(t.G.nodes()))))
    nx.set_node_attributes(t.G, name='IPT', values=valuesOne)
    # Edge mandatory attributes
    valuesOne = dict(itertools.izip(t.G.edges(), np.ones(len(t.G.edges()))))
    nx.set_edge_attributes(t.G, name='BW', values=valuesOne)
    nx.set_edge_attributes(t.G, name='PR', values=valuesOne)

    nx.write_graphml(t.G, "network_grid.graphml")

Other possible solution is to have another Graph. In your custom classes, you can have all variables than you need, so, you can create your `logical topology'.