In this paper, a high fidelity scheme of quantum teleportation based on quantum neural network (QNN) is proposed. The QNN is composed of multi-bit control-not gates. The quantum teleportation of a qubit state via two-qubit entangled channels is investigated by solving the master equation in Lindblad operators with a noisy environment. To ensure the security of quantum teleportation, the indirect training of QNN is employed. Only 10% of teleported information is extracted for the training of QNN parameters. Then the outputs are corrected by the other QNN at Bob's side. We build a random series of numbers ranged in [0, π] as inputs and simulate the properties of our teleportation scheme. The results show that the fidelity of quantum teleportation system is significantly improved to approach 1 by the error-correction of QNN. It illustrates that the distortion can be eliminated perfectly and the high fidelity of quantum teleportation could be implemented.
Weak dissipation for high-fidelity qubit-state preparation and measurement
