Global Methods for Combinatorial Isoperimetric Problems

Let C be a simple closed curve in the plane, of length l enclosing a region, S, of area A. The isoperimetric problem of Greek geometry was to find the maximum value of A for a fixed l, the obvious candidate for a solution being a circle of radius r = l/2 ? and area A = ?r 2 = l 2 /4 ?. Since under dilation of the plane by a factor ? > 0, every such planar set is mapped to a set with boundary of length ?l and area ? 2 A, as a function of l the maximum value of A will be A ( l) = ?l 2 for some constant ? ? 1 /4 ?. Thus A ( l) must be monotone increasing. Minimizing l for fixed A is an equivalent form of the isoperimetric problem whose solution,
, is also monotone increasing. The region S may be assumed convex since, if not, its convex closure would only decrease l and increase A.
Let L be any line in the plane and we assume for convenience that it passes through the centroid of S. Take it to be the x-axis for a coordinatization of the plane and let the y-axis also pass through the centroid...