245 lines
5.7 KiB
TeX
245 lines
5.7 KiB
TeX
\section{Quantum Gates}
|
|
|
|
|
|
In the previous section, we stated that a quantum gate is a linear map. \par
|
|
Let's complete that definition.
|
|
|
|
\definition{}
|
|
A quantum gate is a \textit{orthonormal matrix}, which means any gate $G$
|
|
satisfies $GG^\text{T} = I$. \par
|
|
This implies the following: \par
|
|
|
|
\begin{itemize}
|
|
\item $G$ is square \par
|
|
\note{
|
|
If we think of $G$ as a map, this means that $G$ has as many inputs as it has outputs. \\
|
|
This is to be expected: we stated earlier that quantum gates do not destroy or create qubits.
|
|
}
|
|
|
|
\item $G$ preserves lengths; i.e $|x| = |Gx|$. \par
|
|
\note{This ensures that $G\ket{\psi}$ is always a valid state.}
|
|
\end{itemize}
|
|
|
|
(You will prove all these properties in any introductory linear algebra course. \\
|
|
This isn't a lesson on linear algebra, so you may take them as given today.)
|
|
|
|
\definition{}
|
|
Let $\mathbb{U} \subset \mathbb{R}^2$ be the set of points in the unit circle. \par
|
|
We can restate the above definition as follows: \par
|
|
A quantum gate is an invertible map from $\mathbb{U}^n$ to $\mathbb{U}^n$.
|
|
|
|
|
|
\definition{}<qgateislinear>
|
|
Let $G$ be a quantum gate. \par
|
|
Since quantum gates are, by definition, \textit{linear} maps,
|
|
the following holds: \par
|
|
|
|
\begin{equation*}
|
|
G\bigl(a_0 \ket{0} + a_1\ket{1}\bigr) = a_0G\ket{0} + a_1G\ket{1}
|
|
\end{equation*}
|
|
|
|
\problem{}
|
|
Consider the \textit{controlled not} (or \textit{cnot}) gate, defined by the following table: \par
|
|
\begin{itemize}
|
|
\item $\text{X}_\text{c}\ket{00} = \ket{00}$
|
|
\item $\text{X}_\text{c}\ket{01} = \ket{11}$
|
|
\item $\text{X}_\text{c}\ket{10} = \ket{10}$
|
|
\item $\text{X}_\text{c}\ket{11} = \ket{01}$
|
|
\end{itemize}
|
|
In other words, the cnot gate inverts its first bit if its second bit is $\ket{1}$. \par
|
|
Find the matrix that applies the cnot gate.
|
|
|
|
\begin{solution}
|
|
\begin{equation*}
|
|
\text{CNOT} = \left[\begin{smallmatrix}
|
|
1 & 0 & 0 & 0 \\
|
|
0 & 1 & 0 & 0 \\
|
|
0 & 0 & 0 & 1 \\
|
|
0 & 0 & 1 & 0 \\
|
|
\end{smallmatrix}\right]
|
|
\end{equation*}
|
|
|
|
\vspace{4mm}
|
|
|
|
If $\ket{a}$ is $\ket{0}$, $\ket{a} \otimes \ket{b}$ is
|
|
$
|
|
\left[
|
|
\begin{smallmatrix}
|
|
\left[
|
|
\begin{smallmatrix}
|
|
b_1 \\ b_2
|
|
\end{smallmatrix}
|
|
\right]
|
|
\\ 0 \\ 0
|
|
\end{smallmatrix}
|
|
\right]
|
|
$, and the \say{not} portion of the matrix is ignored.
|
|
|
|
|
|
\vspace{4mm}
|
|
|
|
If $\ket{a}$ is $\ket{1}$, $\ket{a} \otimes \ket{b}$ is
|
|
$
|
|
\left[
|
|
\begin{smallmatrix}
|
|
0 \\ 0 \\
|
|
\left[
|
|
\begin{smallmatrix}
|
|
b_1 \\ b_2
|
|
\end{smallmatrix}
|
|
\right]
|
|
\end{smallmatrix}
|
|
\right]
|
|
$, and the \say{identity} portion of the matrix is ignored.
|
|
|
|
|
|
The state of $\ket{a}$ is always preserved, since it's determined by the position of
|
|
$\left[\begin{smallmatrix}b_1 \\ b_2\end{smallmatrix}\right]$ in the tensor product.
|
|
If $\left[\begin{smallmatrix}b_1 \\ b_2\end{smallmatrix}\right]$ is on top, $\ket{a}$ is $\ket{0}$,
|
|
and if $\left[\begin{smallmatrix}b_1 \\ b_2\end{smallmatrix}\right]$ is on the bottom, $\ket{a}$ is $\ket{1}$.
|
|
\end{solution}
|
|
|
|
|
|
|
|
|
|
\vfill
|
|
\pagebreak
|
|
|
|
|
|
|
|
\problem{}<applycnot>
|
|
Evaluate the following:
|
|
\begin{equation*}
|
|
\text{X}_\text{C}
|
|
\Bigl(
|
|
\frac{1}{2}\ket{00} +
|
|
\frac{1}{2}\ket{01} -
|
|
\frac{1}{2}\ket{10} -
|
|
\frac{1}{2}\ket{11}
|
|
\Bigr)
|
|
\end{equation*}
|
|
|
|
|
|
\vfill
|
|
|
|
\problem{}
|
|
If we measure the result of \ref{applycnot}, what are the probabilities of getting each state?
|
|
|
|
\vfill
|
|
|
|
\generic{Remark:}
|
|
As we just saw, a quantum gate is fully defined by the place it maps our basis states $\ket{0}$ and $\ket{1}$ \par
|
|
(or, $\ket{00...0}$ through $\ket{11...1}$ for multi-qubit gates). This directly follows from \ref{qgateislinear}.
|
|
|
|
\pagebreak
|
|
|
|
%\problem{}
|
|
%Now, modify the CNOT gate so that it inverts $\ket{a}$ whenever it is applied.
|
|
%
|
|
%\begin{solution}
|
|
% \begin{equation*}
|
|
% \text{CNOT}_{\text{mod}} = \begin{bmatrix}
|
|
% 0 & 1 & 0 & 0 \\
|
|
% 1 & 0 & 0 & 0 \\
|
|
% 0 & 0 & 1 & 0 \\
|
|
% 0 & 0 & 0 & 1
|
|
% \end{bmatrix}
|
|
% \end{equation*}
|
|
%\end{solution}
|
|
|
|
%\problem{}
|
|
%Finally, modify the original CNOT gate so that the roles of its bits are reversed: \par
|
|
%$\text{CNOT}_{\text{flip}} \ket{ab}$ should invert $\ket{a}$ iff $\ket{b}$ is $\ket{1}$.
|
|
%
|
|
%
|
|
%\begin{solution}
|
|
% \begin{equation*}
|
|
% \text{CNOT}_{\text{flip}} = \begin{bmatrix}
|
|
% 1 & 0 & 0 & 0 \\
|
|
% 0 & 0 & 0 & 1 \\
|
|
% 0 & 0 & 1 & 0 \\
|
|
% 0 & 1 & 0 & 0 \\
|
|
% \end{bmatrix}
|
|
% \end{equation*}
|
|
%\end{solution}
|
|
%
|
|
%\vfill
|
|
|
|
|
|
|
|
\definition{}
|
|
The \textit{Hadamard Gate} is given by the following matrix: \par
|
|
\begin{equation*}
|
|
H = \frac{1}{\sqrt{2}}\begin{bmatrix}
|
|
1 & 1 \\
|
|
1 & -1
|
|
\end{bmatrix}
|
|
\end{equation*}
|
|
\note{Note that we divide by $\sqrt{2}$, since $H$ must be orthonormal.}
|
|
|
|
\begin{ORMCbox}{Review: Matrix Multiplication}{black!10!white}{black!65!white}
|
|
Matrix multiplication works as follows:
|
|
|
|
\begin{equation*}
|
|
AB =
|
|
\begin{bmatrix}
|
|
1 & 2 \\
|
|
3 & 4 \\
|
|
\end{bmatrix}
|
|
\begin{bmatrix}
|
|
a_0 & b_0 \\
|
|
a_1 & b_1 \\
|
|
\end{bmatrix}
|
|
=
|
|
\begin{bmatrix}
|
|
1a_0 + 2a_1 & 1b_0 + 2b_1 \\
|
|
3a_0 + 4a_1 & 3b_0 + 4b_1 \\
|
|
\end{bmatrix}
|
|
\end{equation*}
|
|
|
|
|
|
Note that this is very similar to multiplying each column of $B$ by $A$. \par
|
|
The product $AB$ is simply $Ac$ for every column $c$ in $B$:
|
|
|
|
\begin{equation*}
|
|
Ac_0 =
|
|
\begin{bmatrix}
|
|
1 & 2 \\
|
|
3 & 4 \\
|
|
\end{bmatrix}
|
|
\begin{bmatrix}
|
|
a_0 \\ a_1
|
|
\end{bmatrix}
|
|
=
|
|
\begin{bmatrix}
|
|
1a_0 + 2a_1 \\
|
|
3a_0 + 4a_1
|
|
\end{bmatrix}
|
|
\end{equation*}
|
|
|
|
This is exactly the first column of the matrix product.
|
|
\end{ORMCbox}
|
|
|
|
|
|
\problem{}
|
|
What is $HH$? \par
|
|
Using this result, find $H^{-1}$.
|
|
|
|
\begin{solution}
|
|
$HH = I$, so $H^{-1} = H$
|
|
\end{solution}
|
|
|
|
\vfill
|
|
|
|
\problem{}
|
|
What are $H\ket{0}$ and $H\ket{1}$? \par
|
|
Are these states entangled?
|
|
|
|
\begin{solution}
|
|
$H\ket{0} = \frac{1}{\sqrt{2}}\bigl(\ket{0} + \ket{1}\bigr)$ and $H\ket{1} = \frac{1}{\sqrt{2}}\bigl(\ket{0} - \ket{1}\bigr)$ \par
|
|
Both of these are entangled states.
|
|
\end{solution}
|
|
|
|
\vfill
|
|
\pagebreak
|