Dissertação de Mestrado - Ivan Oglobin

Efficient simulation of errors in Fusion Based Quantum Memory

Abstract:

This thesis studies fault tolerant quantum memory in a photonic Fusion Based Quantum Computing (FBQC) architecture. Using the stabilizer formalism, we specify small entangled resource states and show how a layered six ring fusion network implements the identity gate as a memory primitive. We embed a rotated surface code to define logical qubits and operators, linking code geometry to the fusion topology through correlation surfaces and surgery style constructions that preserve the encoded state across layers.

We model imperfections as noisy channels applied after each fusion/measurement, focusing on outcome flip errors, and perform purely classical decoding via Pauli frame updates. To assess performance, we simulate Clifford only circuits with the Pauli frame algorithm, efficiently sampling logical error rates across network sizes and memory depths. The resulting finite size scaling exhibits a crossing region, indicating a threshold for the studied model between 0.7% and 1% fusion error probability.

The results provide theoretical and numerical evidence that a photonic FBQC memory can achieve fault tolerance under realistic noise assumptions. We outline next steps toward experimental relevance and scalability: incorporating more hardware faithful noise processes, exploring code variants and stronger decoders, and extending beyond the identity operation to assemble a practical logical gate set.

Banca Examinadora

Prof. Boyan Slavchev Sirakov - (Orientador) - PUC-Rio
Prof. Sergey Tikhomirov (Co-Orientador) - PUC-Rio
Prof. Roberto Imbuzeiro Moraes Felinto de Oliveira - IMPA
Prof. Thiago Barbosa dos Santos Guerreiro - PUC-Rio
Prof. Sergey Sysoev - Saint Petersburg State University)