Convergence of Square of Linear Combination of Sequences whose Squares Converge

From ProofWiki
Jump to navigation Jump to search

Theorem

Let $\sequence {x_i}$ and $\sequence {y_i}$ be real sequences such that the series $\ds \sum_{i \mathop \ge 0} {x_i}^2$ and $\ds \sum_{i \mathop \ge 0} {y_i}^2$ are convergent.

Let $\lambda, \mu \in \R$ be real numbers.


Then $\ds \sum_{i \mathop \ge 0} \paren {\lambda x_i + \mu y_i}^2$ is convergent.


Proof

Let $n \in \N$.

Then:

$\ds \sum_{i \mathop = 1}^n \paren {\lambda x_i + \mu y_i}^2 = \lambda^2 \sum_{i \mathop = 1}^n {x_i}^2 + \mu^2 \sum_{i \mathop = 1}^n {y_i}^2 + 2 \lambda \mu \sum_{i \mathop = 1}^n x_i y_i$

By Cauchy's Inequality:

$\ds \sum_{i \mathop = 1}^n x_i y_i \le \paren {\sum_{i \mathop = 1}^n {x_i}^2}^{\frac 1 2} \paren {\sum_{i \mathop = 1}^n {y_i}^2}^{\frac 1 2}$

Hence:

\(\ds \sum_{i \mathop = 1}^n \paren {\lambda x_i + \mu y_i}^2\) \(\le\) \(\ds \lambda^2 \sum_{i \mathop = 1}^n {x_i}^2 + \mu^2 \sum_{i \mathop = 1}^n {y_i}^2 + 2 \size {\lambda \mu} \sum_{i \mathop = 1}^n x_i y_i\)
\(\ds \) \(\le\) \(\ds \lambda^2 \sum_{i \mathop = 1}^n {x_i}^2 + \mu^2 \sum_{i \mathop = 1}^n {y_i}^2 + 2 \size {\lambda \mu} \paren {\sum_{i \mathop = 1}^n {x_i}^2}^{\frac 1 2} \paren {\sum_{i \mathop = 1}^n {y_i}^2}^{\frac 1 2}\)
\(\ds \) \(=\) \(\ds \paren {\size \lambda \paren {\sum_{i \mathop = 1}^n {x_i}^2}^{\frac 1 2} + \size \mu \paren {\sum_{i \mathop = 1}^n {y_i}^2}^{\frac 1 2} }^2\)
\(\ds \) \(\le\) \(\ds \paren {\size \lambda \paren {\sum_{i \mathop \ge 0} {x_i}^2}^{\frac 1 2} + \size \mu \paren {\sum_{i \mathop \ge 0} {y_i}^2}^{\frac 1 2} }^2\)

Thus the sequence of partial sums $\ds \sum_{i \mathop = 1}^n \paren {\lambda x_i + \mu y_i}^2$ is bounded above.

We also have that $\ds \sum_{i \mathop = 1}^n \paren {\lambda x_i + \mu y_i}^2$ is also increasing.

So by the Monotone Convergence Theorem, $\ds \sum_{i \mathop = 1}^n \paren {\lambda x_i + \mu y_i}^2$ is convergent.

$\blacksquare$


Sources