I've been working on this following problem:
Find a constant $c< 1$ such that $F_n \leq 2^{cn}$ for all $n \geq 0$.
I honestly have no idea where to begin on this. I've done plenty of proofs the one function is Big-O another, but I'm not sure where to start looking for c in this case. Guidance appreciated. Thanks!
Binet's formula seems like overkill, given the (relatively weak) desired conclusion.
A more naive strategy is to pretend that $F_{n} = 2^{cn}$ and see what the recursion relation $F_{n+2} = F_{n} + F_{n+1}$ suggests. Here, $$ 2^{c(n+2)} = 2^{cn} + 2^{c(n+1)}, \quad\text{or}\quad 2^{2c} = 1 + 2^{c}. $$ The latter is a quadratic in $2^{c}$, and by the quadratic formula has roots $$ 2^{c} = \frac{1 \pm \sqrt{5}}{2}. $$ It's therefore natural to attempt to show $F_{n} \leq \bigl[(1 + \sqrt{5})/2\bigr]^{n}$ for $n \geq 0$; this is a straightforward induction.