User:Hsar500

In quantum information theory, a pair of orthonormal bases $$ \{|e_j\rangle \}$$ and $$ \{|f_k\rangle \}, 0\le j, k \le d-1$$ of the Hilbert space $$ \mathbb{C}^d $$ are said to be mutually unbiased if the square of the magnitude of the inner product between any basis states $$ |e_j\rangle $$ and $$ |f_k \rangle $$ equals the inverse of the dimension $$ d $$,


 * $$ |\langle e_j | f_k \rangle|^2 = \frac{1}{d}, $$ for all $$ 0\le j,k \le d-1 $$

These bases are unbiased: if a system is prepared in a state belonging to one of the bases, then the outcomes of the measurement with respect to the other basis will occur with equal probabilities.

Overview
The notion of mutually unbiased bases was first introduced by Schwinger in 1960, and the first person to consider applications of mutually unbiased bases was Ivanovic in the problem of quantum state determination.

Another area where mutually unbiased bases can be applied is quantum key distribution, more specifically in secure quantum key exchange. Mutually unbiased bases are used in many protocols since the outcome is random when a measurement is made in a basis unbiased to that in which the state was prepared. When two remote parties share two non-orthogonal quantum states, attempts by an eavesdropper to distinguish between these by measurements will affect the system and this can be detected. While many quantum cryptography protocols have relied on 1-qubit technologies, employing higher dimensional states, such as qutrits, allows for better security against eavesdropping. This motivates the study of mutually unbiased bases in higher-dimensional spaces.

Other uses of mutually unbiased bases include quantum state reconstruction, quantum error correction codes ,  and the so called "mean king's problem".

Mathematical Formulation
Two orthonormal bases $$ \{|e_j\rangle \} $$ and $$ \{|f_k\rangle \}, 0\le j, k \le d-1$$ of the Hilbert space $$ \mathbb{C}^d $$, satisfying

$$ |\langle e_j | f_k \rangle|^2 = \frac{1}{d}, $$ for all $$ 0\le j,k \le d-1 $$

are said to be mutually unbiased.

That is, the inner product between any basis vector $$ |e_j\rangle $$ and any basis vector $$ |f_k\rangle $$ is of magnitude $$ 1 / \sqrt{d} $$. The maximal number of mutually unbiased bases in the $$d$$-dimensional Hilbert space $$\mathbb{C}^d$$ is denoted by $$ \mathfrak{M}(d) $$.

Existence problem
It is an open question how many mutually unbiased bases, $$ \mathfrak{M}(d) $$, one can find in $$\mathbb{C}^d$$, for arbitrary $$d$$.

In general, if


 * $$ d = p_1^{n_1} p_2^{n_2}...p_k^{n_k} $$

is the prime number decomposition of $$ d $$, where


 * $$ p_1^{n_1} < p_2^{n_2}<...<p_k^{n_k} $$

then the maximal number of mutually unbiased bases which can be constructed satisfies


 * $$p_1^{n_1}+1 \le \mathfrak{M}(d) \le d+1 $$

It follows that if the dimension of a Hilbert space $$ d $$ is an integer power of a prime number, then it is possible to find $$ d+1$$ mutually unbiased bases. This can be seen in the previous equation, as the prime number decomposition of $$ d $$ simply is $$ d = p_1^{n_1} $$. Therefore,
 * $$ \mathfrak{M}(d) = d+1 $$

Though the maximal number of mutually unbiased bases is known when $$ d $$ is an integer power of a prime number, it is not known for arbitrary $$ d $$.

$$ d = 2 $$
The three bases provides the simplest example of mutually unbiased bases in $$\mathbb{C}^2$$. The above bases are composed of the eigenvectors of the Pauli spin matrices $$ \sigma_x, \sigma_z $$ and their product $$ \sigma_x \sigma_z$$.
 * $$ M_0 = \left\{ | 0 \rangle,| 1 \rangle \right\} $$
 * $$ M_1 = \left\{ \frac{| 0 \rangle+| 1 \rangle}{\sqrt{2}},\frac{| 0 \rangle-| 1 \rangle}{\sqrt{2}} \right\} $$
 * $$ M_2 = \left\{ \frac{| 0 \rangle+i | 1 \rangle}{\sqrt{2}},\frac{| 0 \rangle-i| 1 \rangle}{\sqrt{2}} \right\} $$

$$ d = 4 $$
For $$ d=4 $$, an example of $$ d+1 = 5 $$ mutually unbiased bases where each basis is denoted as $$ M_j, 0 \le j \le 4 $$ is given as follows


 * $$ M_0 = \left\{(1,0,0,0),(0,1,0,0),(0,0,1,0),(0,0,0,1)\right\} $$
 * $$ M_1 = \left\{\frac{1}{2}(1,1,1,1),\frac{1}{2}(1,1,-1,-1),\frac{1}{2}(1,-1,-1,1),\frac{1}{2}(1,-1,1,-1)\right\} $$
 * $$ M_2 = \left\{\frac{1}{2}(1,-1,-i,-i),\frac{1}{2}(1,-1,i,i),\frac{1}{2}(1,1,i,-i),\frac{1}{2}(1,1,-i,i)\right\} $$
 * $$ M_3 = \left\{\frac{1}{2}(1,-i,-i, -1),\frac{1}{2}(1,-i,i,1),\frac{1}{2}(1,i,i,-1),\frac{1}{2}(1,i,-i,1)\right\} $$
 * $$ M_4 = \left\{\frac{1}{2}(1,-i,-1,-i),\frac{1}{2}(1,-i,1,i),\frac{1}{2}(1,i,-1,i),\frac{1}{2}(1,i,1,-i)\right\} $$

Weyl group method
Let $$ \hat{X} $$ and $$ \hat{Z} $$ be two unitary operators in the Hilbert space $$ \mathbb{C}^d $$ such that for some phase factor $$ q $$. If $$q$$ is a primitive root of unity, for example $$ q \equiv e^{\frac{2 \pi i}{d}} $$ then the eigenbases of $$ \hat{X} $$ and $$ \hat{Z} $$ are mutually unbiased.
 * $$ \hat{X}\hat{Z} = q\hat{Z}\hat{X} $$

By choosing the eigenbasis of $$ \hat{Z} $$ to be the standard basis, we can generate another basis unbiased to it using a Fourier matrix. The elements of the Fourier matrix are given by Other bases which are unbiased to both the standard basis and the basis generated by the Fourier matrix can be generated using Weyl groups. The dimension of the Hilbert space is important when generating sets of mutually unbiased bases using Weyl groups. When $$ d $$ is a prime number, then the usual $$ d+1 $$ mutually unbiased bases can be generated using Weyl groups. When $$ d $$ is not a prime number, then it is possible that the maximal number of mutually unbiased bases which can be generated using this method is $$ 3 $$.
 * $$F_{ab} = q^{ab}, 0 \le a,b \le N-1 $$

===Unitary operators method using Galois fields ===

When $$d=p$$ is prime we define the unitary operators $$ \hat{X} $$ and $$ \hat{Z} $$ by where $$ \{ |k \rangle | 0 \le j \le d-1 \} $$ is the standard basis and $$ \omega = e^{\frac{2\pi i}{d}} $$ is a root of unity.
 * $$\hat{X} = \sum_{k=0}^{N-1} |k+1 \rangle \langle k| $$
 * $$\hat{Z} = \sum_{k=0}^{N-1} \omega^k |k \rangle \langle k| $$

Then the eigenbases of the following $$ d+1 $$ operators are mutually unbiased :


 * $$ \hat{X}, \hat{Z}, \hat{X} \hat{Z}, \hat{X} \hat{Z}^2 ... \hat{X} \hat{Z}^{d-1} $$

When $$d=p^M$$ is the power of a prime we make use of the Galois field $$GF(d)$$ to construct a maximal set of $$ d+1 $$ mutually unbiased bases. We label the elements of the computational basis of $$ \mathbb{C}^d $$ using the Galois field: $$\{ |a \rangle | a \in GF(N) \}$$.

We define the operators $$ \hat{X_a} $$ and $$ \hat{Z_b} $$ in the following way
 * $$\hat{X_a} = \sum_{c \in GF(N)} |c + a \rangle \langle c| $$
 * $$\hat{Z_b} = \sum_{c \in GF(N)} \chi (bc)|c \rangle \langle c| $$

where $$\chi(\theta) = \exp \left [ \frac{2\pi i }{p} \left ( \theta+ \theta^p + \theta^{p^2}+ \cdots + \theta^{p^{M-1}} \right ) \right ]$$, and the addition and multiplication in the kets and $$\chi(\cdot)$$ is that of $$GF(d)$$.

Then we form $$d+1$$ sets of commuting unitary operators:


 * $$\{ \hat{Z_q} | q \in GF(N) \} $$ and $$ \{ \hat{X_q}\hat{Z_{qr}} | q \in GF(N) \} $$ for each $$ r \in GF(N) $$

The joint eigenbases of the operators in one set are mutually unbiased to that of any other set. We thus have $$ d+1 $$ mutually unbiased bases.

Hadamard matrix method
Given that one basis in a Hilbert space is the standard basis, then all bases which are unbiased with respect to this basis can be represented by the columns of a complex Hadamard matrix multiplied by a normalization factor. For $$ d=3 $$ these matrices would have the form 1 & 1 & 1 \\ e^{i \phi_{10}} & e^{i \phi_{11}} & e^{i \phi_{12}} \\ e^{i \phi_{20}} & e^{i \phi_{21}} & e^{i \phi_{22}} \end{bmatrix} $$
 * $$ U = \frac{1}{\sqrt{d}} \begin{bmatrix}

The problem of finding a set of mutually unbiased bases therefore corresponds to finding unequivalent Hadamard matrices which are mutually unbiased to each other. Two Hadamard matrices are considered equivalent if, through permutations of rows and columns, and multiplication of rows and columns by arbitrary phase factors, they can be transformed into each other.

An example of a one parameter family of Hadamard matrices in a 4 dimensional Hilbert space is 1 & 1 & 1 & 1 \\ 1 & e^{i\phi} & -1 & -e^{i \phi} \\ 1 & -1 & 1 & -1 \\ 1 & -e^{i\phi} & -1 & e^{i\phi} \end{bmatrix} $$
 * $$ H_4(\phi) = \frac{1}{2} \begin{bmatrix}

The problem of finding a maximal set of MUBs when $$ d=6 $$
The smallest dimension that is not an integer power of a prime is $$ d=6 $$. This is also the smallest dimension for which the number of mutually unbiased bases is not known. The methods used to determine the number of mutually unbiased bases when $$ d $$ is an integer power of a prime number cannot be used in this case. Searches for a set of four mutually unbiased bases when $$ d = 6 $$, both by using Hadamard matrices and numerical methods have been unsuccessful. The general belief is that the maximum number of mutually unbiased bases for $$ d=6 $$ is $$\mathfrak{M}(6) = 3 $$.

==Entropic Uncertainty Relations and MUBs ==

There is an alternative characterisation of mutually unbiased bases that considers them in terms of uncertainty relations.

Entropic uncertainty relations are analogous to the Heisenberg uncertainty principle, and Maassen and Uffink found that for any two bases $$B_1 = \{ |a_{i}\rangle_{i=1}^d \} $$ and  $$B_2 = \{ | b_{j} \rangle _{j=1}^{d} \}$$:


 * $$ H_{B_1} + H_{B_2} \geq -2\log c.$$

where $$c = max | \langle a_j | b_k \rangle |$$ and $$ H_{B_1}$$ and $$H_{B_2}$$ is the respective entropy of the bases $$B_1$$ and $$B_2$$, when measuring a given state.

Entropic uncertainty relations are often preferable to the Heisenberg uncertainty principle, as they are not phrased in terms of the state to be measured, but in terms of $$ c $$.

In scenarios such as quantum key distribution, we aim for measurement bases such that full knowledge of a state with respect to one basis implies minimal knowledge of the state with respect to the other bases. This implies a high entropy of measurement outcomes, and thus we call these strong entropic uncertainty relations.

For two bases, the lower bound of the uncertainty relation is maximized when the measurement bases are mutually unbiased, since mutually unbiased bases are maximally incompatible: the outcome of a measurement made in a basis unbiased to that in which the state is prepared in is completely random. In fact, for a $$d$$-dimensional space, we have :


 * $$ H_{B_1} + H_{B_2} \geq \log (d)$$

for any pair of mutually unbiased bases $$B_1$$ and $$B_2$$. This bound is optimal : If we measure a state from one of the bases then the outcome has entropy 0 in that basis and an entropy of $$\log(d)$$ in the other.

If the dimension of the space is a prime power, we can construct $$d+1$$ MUBs, and then it has been found that


 * $$ \sum_{k=1}^{d+1} H_{B_k} \geq \frac{d+1}{2} \log(\frac{d+1}{2} )$$

which is stronger than the relation we would get from pairing up the sets and then using the Maassen and Uffink equation. Thus we have a characterisation of $$d+1$$ mutually unbiased bases as those for which the uncertainty relations are strongest.

Although the case for two bases, and for $$d+1$$ bases is well studied, very little is known about uncertainty relations for mutually unbiased bases in other circumstances. It is not obvious that strong uncertainty relations exist for all cases, but it has been shown that it is possible in principle.

Mutually unbiased bases in infinite dimension Hilbert spaces
While there has been investigation into mutually unbiased bases in infinite dimension Hilbert space, their existence remains an open question. It is conjectured that in a continuous Hilbert space, two orthonormal bases $$ |\psi_s^b \rangle $$ and $$ |\psi_{s'}^{b'} \rangle $$ are said to be mutually unbiased if For the generalized position and momentum eigenstates $$ | q \rangle, q\in \mathbb{R} $$ and $$ | p  \rangle,p\in \mathbb{R} $$, the value of $$ k $$ is
 * $$ |\langle \psi_s^b | \psi_{s'}^{b'} \rangle|^2 = k>0, s,s'\in \mathbb{R} $$


 * $$ |\langle q | p \rangle|^2 = \frac{1}{2 \pi \hbar} $$

The existence of mutually unbiased bases in a continuous Hilbert space remains open for debate, as further research in their existence is required before any conclusions can be reached.