Search papers, labs, and topics across Lattice.
This paper introduces Remote Graph State Preparation (RGSP), a novel framework that directly prepares arbitrary entangled graph states from a single high-dimensional photonic qudit, bypassing the need for server-side entangling gates required by prior separable state RSP protocols. RGSP encodes multiple qubits and their graph connectivity into the photon's phase profile, achieving topology-independent fidelity under a reasonable noise model and demonstrating significant enhancement against fiber phase drift with a highest-weight-first reordering strategy. This approach enables resource-efficient, topology-invariant blind quantum computation, drastically reducing qubit overhead for universal computation, exemplified by a six-qubit
Key contribution not extracted.
Measurement-based quantum computation (MBQC) relies on entangled graph states, yet existing remote state preparation (RSP) protocols prepare only separable states, requiring subsequent entangling gates on the remote server. Here, we introduce Remote Graph State Preparation (RGSP), a framework that prepares arbitrary graph states directly from a single high-dimensional photonic qudit. By encoding multiple qubits and their graph connectivity into the photon's structured phase profile, RGSP can reduce or, for small computations, completely eliminate the need for server-side entangling operations among qubits. We show that under a reasonable noise model the resulting state fidelity is independent of the graph topology. Under cumulative fiber phase drift, we demonstrate that this fidelity is significantly enhanced by a ``highest-weight-first'' mode reordering strategy. Finally, we show that RGSP enables reduced-SWAP universal blind quantum computing on arbitrary topologies, reducing qubit overhead e.g., for a six-qubit Quantum Fourier Transform from $420$ RSP qubits with a standard brickwork topology to $48$ RSP qubits. These results establish RGSP as a resource-efficient, topology-invariant primitive for quantum-secured cloud computing.