Paper 3, Section II, C

Quantum Information and Computation | Part II, 2020

Consider the quantum oracle UfU_{f} for a function f:Bn→Bnf: B_{n} \rightarrow B_{n} which acts on the state ∣x⟩∣y⟩|x\rangle|y\rangle of 2n2 n qubits as follows:

Uf∣x⟩∣y⟩=∣x⟩∣y⊕f(x)⟩U_{f}|x\rangle|y\rangle=|x\rangle|y \oplus f(x)\rangle

The function ff is promised to have the following property: there exists a z∈Bnz \in B_{n} such that for any x,y∈Bnx, y \in B_{n},

[f(x)=f(y)] if and only if x⊕y∈{0n,z}[f(x)=f(y)] \text { if and only if } x \oplus y \in\left\{0^{n}, z\right\}

where 0n≡(0,0,…,0)∈Bn0^{n} \equiv(0,0, \ldots, 0) \in B_{n}.

(a) What is the nature of the function ff for the case in which z=0nz=0^{n}, and for the case in which z≠0nz \neq 0^{n} ?

(b) Suppose initially each of the 2n2 n qubits are in the state ∣0⟩|0\rangle. They are then subject to the following operations:

  1. Each of the first nn qubits forming an input register are acted on by Hadamard gates;

  2. The 2n2 n qubits are then acted on by the quantum oracle UfU_{f};

  3. Next, the qubits in the input register are individually acted on by Hadamard gates.

(i) List the states of the 2n2 n qubits after each of the above operations; the expression for the final state should involve the nn-bit "dot product" which is defined as follows:

a⋅b=(a1b1+a2b2+…+anbn) mod 2a \cdot b=\left(a_{1} b_{1}+a_{2} b_{2}+\ldots+a_{n} b_{n}\right) \bmod 2

where a,b∈Bna, b \in B_{n} with a=(a1,…,an)a=\left(a_{1}, \ldots, a_{n}\right) and b=(b1,…,bn)b=\left(b_{1}, \ldots, b_{n}\right).

(ii) Justify that if z=0nz=0^{n} then for any y∈Bny \in B_{n} and any φ(x,y)∈{−1\varphi(x, y) \in\{-1, +1}+1\}, the following identity holds:

∥∑x∈Bnφ(x,y)∣f(x)⟩∥2=∥∑x∈Bnφ(x,y)∣x⟩∥2\| \sum_{x \in B_{n}} \varphi(x, y)|f(x)\rangle\left\|^{2}=\right\| \sum_{x \in B_{n}} \varphi(x, y)|x\rangle \|^{2}

(iii) For the case z=0nz=0^{n}, what is the probability that a measurement of the input register, relative to the computational basis of Cn\mathbb{C}^{n} results in a string y∈Bny \in B_{n} ?

(iv) For the case z≠0nz \neq 0^{n}, show that the probability that the above-mentioned measurement of the input register results in a string y∈Bny \in B_{n}, is equal to the following:

zero for all strings y∈Bny \in B_{n} satisfying y⋅z=1y \cdot z=1, and 2−(n−1)2^{-(n-1)} for any fixed string y∈Bny \in B_{n} satisfying y⋅z=0y \cdot z=0.

[State any identity you may employ. You may use (x⊕z)⋅y=(x⋅y)⊕(z⋅y),∀x,y,z∈Bn(x \oplus z) \cdot y=(x \cdot y) \oplus(z \cdot y), \forall x, y, z \in B_{n}.]

Typos? Please submit corrections to this page on GitHub.