Suppose that ${a_n}$ and $b_n$ is decreasing sequence such that $a_0=A$, $lim_{n->\infty}a_n=0$ and $b_0=B$, $lim_{n->\infty}b_n=0$.

For fix n, we can construct n dimension linear equation system such that

\begin{eqnarray} a_n x_n+a_{n-1}x_{n-1}+\cdots+a_0x_0&=&b_0\\ a_{n}x_{n-1}+\cdots+a_1x_0&=&b_1\\ \vdots\\ a_{n}x_0&=&b_n \end{eqnarray}

then solutions are uniquely determined. If n increases to $\infty$, i want know whether the solutions are bounded. If so, how can it be proved? If not, which condition is added to satisfy the bounded solution?

Thanks in advances.

  • $\begingroup$ Your question does not really make sense as stated, because when $n \rightarrow \infty$, there is an infinite sum in each row. Therefore, you have to at least state in which sense your equations should hold. The only way I can see your system of equations make sense is that you construct an operator from $\ell^1$ to $\ell^\infty$. But I do not see why this operator should be invertible. $\endgroup$ Jul 13 '14 at 15:38
  • $\begingroup$ Or rather, it cannot be invertible, as the two spaces are not isomorphic. $\endgroup$ Jul 13 '14 at 15:44
  • $\begingroup$ @Kofi: I don't understand your objection. For each $n$, we have a perfectly well defined vector $x^{(n)} =(x^{(n)}_0,\ldots,x^{(n)}_n)$, and we can ask about the properties of this sequence as $n\to\infty$. For example, do we have a uniform bound (obviously we don't in general: at the very least, $b_n/a_n$ needs to be bounded)? $\endgroup$ Jul 13 '14 at 20:01

p.s. - This is really a comment but I don't have enough rep.

It seems, you want to bound $$ X_n = A_n^{-1} B_n $$. Let's say we simplify, that $A_n$ and $B_n$ are independent, and then start by assuming $B_n$ is bounded and then find conditions for bounded inverse of a triangular toeplitz matrix. (I did this because I thought that bounded inverses is a nice problem that many people would have worked on already. ) Then we can google "bounded inverse of triangular toeplitz matrix". I found Inversion of Toeplitz matrices II which gives a necessary and sufficient condition for bounded inverse. There is also Uniform bounds on the 1-norm of the inverse of lower triangular Toeplitz matrices. Let us know if that was helpful.


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.