Edge Coloring-I
Edge Coloring-I
1
Edge Coloring for 𝑲𝟓 & 𝑲𝟔 :
Edge-coloring 𝐾5 & 𝐾6 is not quite so obvious as those from above.
Since no two adjacent edges can be given the same color, we know every edge out of a
vertex must be given different colors.
Since every vertex in K5 has degree 4, we know at least 4 colors will be required.
Start by using 4 colors out of one of the vertices of the K5. As shown in the next graph on
left, we started at a.
Moving to the edges incident to b, we attempt to use our pool of 4 previously used colors;
however, one of these is unavailable since it has already been used on the ab edge.
A possible coloring of the edges incident to b is given in the next graph on the right
2
3
The example below investigates a suboptimal edge-coloring using a greedy algorithm and explains
a better procedure for finding an optimal (or nearly optimal) edge-coloring.
4
5
Beyond the analogous definitions and procedures between edge coloring and vertex-coloring,
there is a very direct relationship between the two by the use of a line graph
6
The result above shows we can find an edge-coloring of any graph by simply vertex-coloring its line
graph.
However, other results on line graphs provide some interest, namely if G is eulerian then
L(G) is hamiltonian!
Applications of edge-coloring abound, in particular scheduling independent tasks onto
machines and communicating data through a fiber optics network.
7
The example below relates edge-coloring to Section 1.1 and how to schedule games between
teams in a round-robin tournament.
Although the example above is fairly easy to solve as the graph model is a complete graph, the
same procedure can be extended to a larger number of teams where each team only plays a
8
subset of all teams in the league. In particular, edge-coloring can be used to determine team
schedules in the National Football League!