Quantum state discrimination is an important task for quantum information processing. It is known that in quantum mechanics, one could only perfectly discriminate orthogonal states. When the states are non-orthogonal, several strategies can be adopted to discriminate them. One such strategy is the minimum error strategy where we minimize the probability of error resulting from imperfect discriminations. However, exact solution is only known for two states discrimination case. In this project, we are expecting a search based algorithm to help find the optimal quantum circuit for state discrimination.
- Local search based: simple hill climbing We use step by step approach of approximation to reach the top of the hill in success rate.
-
The quantum states in forms of (a,b) in (a * |0> + b * |1>) qp_1: (1, 0)
qp_2: (0.7071067811865476, 0.7071067811865476)
qp_3: (0, 1)
qp_4: (0.7071067811865476, 0.7071067811865476j)
qp_5: (0.5, 0.8660254037844386)
qp_6: (0.8660254037844386, 0.5)
qp_7: (0.5773502691896257, 0.816496580927726)
qp_8: (0.816496580927726, 0.5773502691896257) -
Discrimination success rates
indexing: (1)q-(2) qp_(3) qp_(4) --> (1): number of qubits, (2): experiment index, (3),(4): state index2q-1 qp_1 qp_2 (Theoretical maximum is about 85%)
average is: 0.8289742857142854
stddev is: 0.016673642648913692q-2 qp_1 qp_3
average is: 0.9775971428571428
stddev is: 0.0189978852743656472q-3 qp_7 qp_8
average is: 0.6590907142857144
stddev is: 0.0081055842114040912q-4 qp_5 qp_6
average is: 0.7422142857142857
stddev is: 0.0097143907557346033q-1 qp_1 qp_2 qp_3 (Theoretical maximum is about 67%)
average is: 0.6167666666666666
stddev is: 0.037158833255215233q-2 qp_1 qp_2 qp_4
average is: 0.5366933333333334
stddev is: 0.01920634617342233q-3 qp_1 qp_5 qp_6
average is: 0.5426266666666667
stddev is: 0.046376988546763864