Allowing five steps, one calculates that a lower bound for the growth factor of almost every element is given by
$${2}^{37/320}{3}^{1/16}{5}^{3/80}{7}^{1/40}{11}^{1/80}{13}^{1/160}\approx 1.35502.$$
Such calculations quickly get too long to do by hand, and the return on investment becomes minimal since the mean is growing like
${1.5}^{N}$
. A simple Mathematica routine can push the calculations significantly farther, however; for example if the basic step size is 15, then the growth factor is approximately
$1.43925$
. Combining this with ( 22 ) and Lemma 4.1 , we get
Theorem 4.2.
As
$N\to \infty $
, the expected value of the length of the simple closed geodesics associated to cycles of length
$N$
,
${E}_{N}\left(l\right)$
, is bounded by
$$(log1.43925)N\le {E}_{N}\left(l\right)\le (log1.5)N.$$
5 The Length of the Shortest Geodesic
In this section we will use the results from previous section to estimate the length of the shortest closed geodesic on
${S}^{C}(\Gamma ,\mathcal{O})$
. The length of the shortest closed geodesic is an important geometric invariant since it is twice the injectivity radius. Let
$syst\left({S}^{C}\right)(\Gamma ,\mathcal{O})$
be length of the shortest closed geodesic on
${S}^{C}(\Gamma ,\mathcal{O})$
, we will show that :
Theorem 5.1.
$$2.809\le syst\left({S}^{C}\right)(\Gamma ,\mathcal{O})\le 3.085$$
It is interesting to compare this result in [
?]
with the the behavior of the largest embedded ball on
$\left({S}^{C}\right)(\Gamma ,\mathcal{O})$
which gives a linear growth to the largest embedded ball.
We will start with the upper bound. As we have seen on a random graph the distribution of short cycles is independent on the size of the graph and it is Poisson distribution with mean
$\lambda =\frac{{2}^{k}}{2k}$
where
$k$
is the length of the cycle. Therefore the probability of not having a
$k$
cycle is
${e}^{-\frac{{2}^{k}}{2k}}$
and the probability of having at least one
$k$
-cycle on the graph is
$(1-{e}^{-\frac{{2}^{k}}{2k}})$
.
As we transition from the graph
$\Gamma $
to the compact surface
${S}^{C}(\Gamma ,\mathcal{O})$
, by Lemma 3.1 there are only 2 orientations that will make the curve that corresponds to a cycle null homotopic. Hence the probability that there is a
$k$
-cycle that gives rise to non trivial geodesic on
${S}^{C}(\Gamma ,\mathcal{O})$
is
$p\left(k\right)=\frac{{2}^{k-2}-1}{{2}^{k-2}}(1-{e}^{-\frac{{2}^{k}}{2k}})$
, and the probability that there is no
$k$
-cycle that produces a geodesic on
${S}^{C}(\Gamma ,\mathcal{O})$
will be
$(1-p(k\left)\right)$
. The expected value for the trace of a
$k$
cycle is
$tr\left(k\right)\frac{{3}^{k}+1}{{2}^{k}}$
and the expected value for the length of the geodesic is bounded from above by by
$2{cosh}^{-1}\left(\frac{tr\left(k\right)}{2}\right)$
.
$$E(syst({S}^{C}\left)\right(\Gamma ,\mathcal{O}\left)\right)\le {\sum}_{k=3}^{\infty}\left(p\right(k{)}^{k-1}{\prod}_{j=2}(1-p(j\left)\right)\left)2{cosh}^{-1}\right(\frac{{3}^{k}+1}{{2}^{k+1}})$$
$$={\sum}_{k=2}^{\infty}\left(\frac{{2}^{k-2}-1}{{2}^{k-2}}(1-{e}^{-\frac{{2}^{k}}{2k}}{)}^{k-1}{\prod}_{j=2}\left(1-\frac{{2}^{j-2}-1}{{2}^{j-2}}(1-{e}^{-\frac{{2}^{j}}{2j}})\right)\right)2{cosh}^{-1}\left(\frac{{3}^{k}+1}{{2}^{k+1}}\right)$$
It is easy to see that this series converges rapidly and we can get a numerical estimate that the value is
$\sim 3.085$
To get a lower bound we can replace
$logE\left(tr\right(M\left)\right)$
by
$E\left(2{cosh}^{-1}\right(\frac{tr\left(M\right)}{2}\left)\right)$
and use the rapid decay for the probability that a graph has large girth. We calculate
$${\sum}_{k=2}^{2}0\left(\frac{{2}^{k-2}-1}{{2}^{k-2}}(1-{e}^{-\frac{{2}^{k}}{2k}}{)}^{k-1}{\prod}_{j=2}\left(1-\frac{{2}^{j-2}-1}{{2}^{j-2}}(1-{e}^{-\frac{{2}^{j}}{2j}})\right)\right)E\left(2{cosh}^{-1}\right(\frac{M}{2}\left)\right)$$
and get the estimate of
$2.809$
.