I am trying to undestand a proof and there is one part that's holding me back. By assumption we have that spectral radius $\rho(A) < 1$. Hence, following inequality should hold

$$\|A^k\| < C \mu^k,$$

where $C>0$ and $\mu \in (0,1)$. I tried to understand it through Gelfand's formula saying $$\rho(A) = \lim_{k\to \infty} \|A^k\|^{\frac{1}{k}}$$

for any matrix norm $\|.\|$, but I still do not undestand why $C>0$ is uniform for any $k$.

Any help please?

  • $\begingroup$ You seem to be convinced that we can find $0 < \mu < 1$ and $C$ which work for $k >K$, where $K$ is some large positive integer. If necessary, replace $C$ by ${\rm max}\{C, \frac{\|A^{k}\|}{\mu^{k}} : k \leq K \}.$ $\endgroup$ – Geoff Robinson Jun 27 '15 at 12:49

Since $\rho(A)<1$, there is an induced norm $N(.)$ s.t. $N(A)<1$. Let $\mu=N(A)$; thus $N(A^k)\leq \mu^k$. Since the norms are equivalent, there is a fixed $C$ s.t. $||.||\leq CN(.)$. Finally $||A^k||\leq CN(A^k)\leq C\mu^k$.

EDIT. If $E$ is a Banach space and $A\in L(E)$ is bounded, then one has the same result. According to Gelfand, there is $\epsilon>0,k_0$ s.t. for $k\geq k_0$, $||A^k||^{1/k}\leq (1-\epsilon)$, that is $||A^k||\leq(1-\epsilon)^k$. Putting $D=\sup_{1\leq k\leq k_0}\dfrac{||A^k||}{(1-\epsilon)^k}$ and $C=\sup(D,1)$, we are done.

  • $\begingroup$ Thank you. I see, that this works fine for finite dimensional spaces as we are using equivalence of norms. But what to do when I am in infinite dimensional space? $\endgroup$ – MathFan Sep 1 '15 at 19:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.