Algoritmos para a informação quântica: discriminação de estados quânticos e modelo híbrido

Abstract

In this thesis we consider a protocol to perform the optimal quantum state discrimination of N linearly independent non-orthogonal pure quantum states. This protocol obtains the best set of generalized measurements for the problem of the unambiguous state discrimination. Through the extension of the Hilbert space, it is possible to perform an unitary operation yielding a final configuration, which gives the best discrimination. We introduce a detailed mathematical procedure to realize this task by means of semidefinite programming and norm minimization. The former is used to fix which is the best detection conclusive probability amplitude for each quantum state of the ensemble. The latter determines the matrix which leads the input states to the final configuration. In a final step, we decompose the unitary transformation in a sequence of 1-qubit rotation matrices. Subsequently, we show applications of the protocol in the quantum cryptography and state filtering, and present a computational code in MATLAB. Our additional result deals with the computational hybrid models. In particular, we study the Semi-quantum Computer. We discuss the performance of the Search and Fourier Transform algorithms on a hybrid architecture constituted of classical and quantum processors working together. We show that the semi-quantum computer can reach an improvement over a pure classical architecture, no matter how few qubits are available and, therefore, it suggests an easier implementable technology.