This algorithm finds the routes,by cost
precedence.Let's assume that every cost is a positive number,and
assume the same in the cost function ** c **as in 5.4
paragraph.

The idea for this algorithm is based on
the fact that every ** m.r** containing more than one
side is the expansion of another

__METHODOLOGY__** **

Let's call D(1),D(2) the routes (found by the Dijkstra Algorithm)
for which { *(cost of D**(1)**
) <= cost of D**(2)**<=...**}.*We
will show that every ** m.r.** contains

On contrary to our previous
symbolism,here we will name V(0) the starting point and V(1),...,V(n-1) the rest of
the peaks of ** G**. We assign 2 numbers in every peak
with V(i)<>V(0) (as in the Prim Algorithm.) as follows.

When another terminal ** m.r.**
for V(m) is found, we examine whether the
direct connection of V(i) to V(m) gives us less cost,that is smaller e(i).In such a case we must set: p(i)=m.For
initialization we consider p(i)=0,that
means direct connection of V(i) to V(0),(obviously
e(i)=infinite if the e(0i) route does not exist).

HTML PAGE DIRECTOR :Papaioannou Panagiotis