Equivalence of these quantum circuits

Click For Summary
SUMMARY

This discussion focuses on the equivalence of two quantum circuits involving qubits ##q_1## and ##q_2##, utilizing C-NOT and Hadamard gates. The C-NOT gate does not affect the control qubit, leading to the conclusion that ##q_1' = q_1 \oplus q_2## after the application of the gates. The participants explore the mathematical representation of the circuits, emphasizing the importance of the computational basis and matrix representations of the gates. The discussion concludes that both methods of demonstrating equivalence yield consistent results, confirming the circuits' functionality.

PREREQUISITES
  • Understanding of quantum circuits and qubits
  • Familiarity with C-NOT and Hadamard gates
  • Knowledge of computational basis in quantum mechanics
  • Ability to manipulate quantum state vectors and matrices
NEXT STEPS
  • Study the matrix representation of the C-NOT gate
  • Learn about the computational basis and its applications in quantum circuits
  • Explore the properties and applications of Hadamard gates in quantum computing
  • Investigate entanglement in quantum states and its implications
USEFUL FOR

Quantum computing enthusiasts, researchers in quantum mechanics, and students studying quantum circuits will benefit from this discussion.

lholmes135
Messages
12
Reaction score
0
TL;DR
C-NOT gate equivalent circuits
In the attached image, there are two quantum circuits that are equivalent. I am trying to understand how. Let's call the top qubit ##q_1## and the bottom one ##q_2##, and the outputs ##q_1'## and ##q_2'##. From what I understand, the C-NOT gate doesn't affect the control qubit. Because Hadamard gates are reversible, it looks to me like ##q_1'=q_1##. In the second circuit though, ##q_1'=q_1 \oplus q_2##. I also tried doing this through calculation. We'll make ##q_1=\begin{bmatrix}c_0\\c_1\end{bmatrix}## and ##q_2=\begin{bmatrix}d_0\\d_1\end{bmatrix}##. Looking at the left circuit, say the input state to the C-NOT gate is ##|\psi>## and the output state ##|\psi'>##.

\begin{equation*}
|\psi>=\frac{1}{2}\begin{bmatrix}(c_0+c_1)(d_0+d_1)\\(c_0+c_1)(d_0-d_1)\\(c_0-c_1)(d_0+d_1)\\(c_0-c_1)(d_0-d_1)\end{bmatrix}
\end{equation*}

C_NOT gate should flip third and fourth elements:

\begin{equation*}
|\psi'>=\frac{1}{2}\begin{bmatrix}(c_0+c_1)(d_0+d_1)\\(c_0+c_1)(d_0-d_1)\\(c_0-c_1)(d_0-d_1)\\(c_0-c_1)(d_0+d_1)\end{bmatrix}
\end{equation*}

The problem is that the output state is entangled, and I'm not sure how to calculate the two H gates on the right if I can't separate the states. This way turned out to be a dead end for me.
 

Attachments

  • Capture.PNG
    Capture.PNG
    1.1 KB · Views: 253
Physics news on Phys.org
You can find the matrix form of the two Hadamard gates acting on the two qubits. You can start by writing the 4x4 version of the Hadamard gate acting on a single qubit, ##H_1 \otimes I_2##.
 
This is what I was looking for, thanks.
 
I have a really basic understanding of quantum circuits but let me try to help

lholmes135 said:
In the attached image, there are two quantum circuits that are equivalent. I am trying to understand how.

There are two ways to show equivalence between circuits (that I am aware of).

##1)## By means of the computational basis ##\alpha## i.e.

\begin{equation*}
\alpha = \{ |00\rangle, |11\rangle, |10\rangle, |01\rangle\}
\end{equation*}

Where

\begin{equation*}
|0\rangle = \begin{pmatrix}
1 \\
0
\end{pmatrix}, \ \ \ \
|1\rangle = \begin{pmatrix}
0 \\
1
\end{pmatrix}
\end{equation*}

##2)## By means of the matrix representation of the involved gates with respect to the computational basis.

Let me go for ##1)##. I will let you show it via ##2)##.

The Hadamard gate acts on a single qubit as follows

\begin{equation*}
H |0 \rangle = \frac{1}{\sqrt{2}}\left( |0 \rangle + |1 \rangle\right) := |+ \rangle
\end{equation*}

\begin{equation*}
H |1 \rangle = \frac{1}{\sqrt{2}}\left( |0 \rangle - |1 \rangle\right) := |- \rangle
\end{equation*}

Or more compacted

\begin{equation*}
H |x \rangle = \frac{1}{\sqrt{2}}\left( |0 \rangle + (-1)^x|1 \rangle\right), \ \ \ \ \text{where} \ x \in \{ 0,1\}
\end{equation*}

The CNOT gate involves two qubits: one is left unchanged while the other is bitwise-added to the first one i.e.

7382937273.png


OK, let us focus on the given circuit. Reading from left to right (where ## x,y \in \{ 0,1\}##) we get

\begin{align}
&\left( H \otimes H \right) \text{CNOT} \left( H \otimes H \right) |x\rangle \otimes |y\rangle \nonumber \\
&= \frac 1 2 \text{CNOT} \left( |0\rangle + (-1)^x |1\rangle \right) \otimes \left( |0\rangle + (-1)^y |1\rangle \right) \nonumber \\
&= \frac 1 2 \left( H \otimes H \right) \text{CNOT} \left( |00\rangle + (-1)^{x+y}|11\rangle + (-1)^{y}|01\rangle + (-1)^{x}|10\rangle \right) \nonumber \\
&= \frac 1 2 \left( H \otimes H \right) \left( |00\rangle + (-1)^{x+y}|10\rangle + (-1)^{y}|01\rangle + (-1)^{x}|11\rangle \right) \nonumber \\
&= \frac 1 2 \left( |++\rangle + (-1)^{x+y}|-+\rangle + (-1)^{y}|+-\rangle + (-1)^{x}|--\rangle \right) \nonumber \\
&= \frac 1 2 \Big[ |+\rangle \otimes \left( |+\rangle + (-1)^{y}|-\rangle \right) + |-\rangle \otimes \left( (-1)^{x+y}|+\rangle + (-1)^{x}|-\rangle \right)\Big] \nonumber \\
&= \frac 1 2 \Big[ |+\rangle \otimes \left( |+\rangle + (-1)^{y}|-\rangle \right) + (-1)^{x+y} |-\rangle \otimes \left( |+\rangle + (-1)^{-y}|-\rangle \right)\Big] \tag{1}\\
&= \frac 1 2 \Big[ |+\rangle \otimes \left( |+\rangle + (-1)^{y}|-\rangle \right) + (-1)^{x+y} |-\rangle \otimes \left( |+\rangle + (-1)^{y}|-\rangle \right)\Big] \tag{2}\\
&= \frac 1 2 \Big[\left( |+\rangle + (-1)^{x+y}|-\rangle \right) \otimes \left( |+\rangle + (-1)^{y}|-\rangle \right) \Big] \nonumber \\
&= \frac 1 4 \left( (1+ (-1)^{x+y})|0\rangle + (1- (-1)^{x+y})|1\rangle \right) \otimes \left( (1+ (-1)^{y})|0\rangle + (1- (-1)^{y})|1\rangle \right) \nonumber \\
&= |x \oplus y \rangle |y \rangle \nonumber
\end{align}

Where we notice that for ##(1) = (2)## to hold we need ##(-1)^{x} = (-1)^{2y + x}##. We have two cases; Case A ##x=1##: We note that ##y## is free to be either ##0## or ##1##. For any of those two values ##2y + x## will yield an odd number so ##(-1)^{x} = (-1)^{2y + x}## holds. Case B ##x=0##: We note that ##y## is again free to be either ##0## or ##1##. For any of those two values ##2y + x## will yield an even number so ##(-1)^{x} = (-1)^{2y + x}## again holds.

To show equivalence via ##2)## you will need to use the CNOT matrix representation as well as the ##4\times4## Hadamard matrix.
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 0 ·
Replies
0
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 3 ·
Replies
3
Views
1K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 27 ·
Replies
27
Views
3K