Error Correction and Feedforward Timing in a Simulated Dual-Rail Photonic Architecture
Abstract In a linear-optical quantum computer, error correction must finish before the encoded photon leaves the optical delay line. We present a 33-mode dual-rail architecture for single-error detection and evaluate repeated-syndrome distance-3 correction using an extended 55-mode circuit-level simulation model. We co-design the timing budget and an FPGA decoder. On a Basys3 board, the decoder was validated at distances 3, 5, and 7 using 40,000 Stim-generated syndrome vectors per distance (120,000 in total), with no mismatches. Under a reference two-term circuit-level Pauli noise model, finite-size scaling gives a threshold estimate of 0.194 ± 0.001, with statistical uncertainty only; the fitted suppression exponents are consistent with ⌈d/2⌉. Post-place-and-route timing gives a conservative decoder cycle of 8.79 ns (113.8 MHz). With assumed detection and switching latencies of 50 ns and 10 ns, respectively, decoding uses about 9% of the 99.8 ns optical-buffer interval, identifying detector readout as the main target for latency reduction. The stated source-quality model also yields a photon-indistinguishability bound of V > 0.982; this is a model-dependent estimate, not an experimental measurement.
Authors
- Vibhu Srivastava (ORCID: https://orcid.org/0000-0002-0725-2994)
- Manu Bansal (ORCID: https://orcid.org/0000-0003-4654-7127)
- Anil Singh (ORCID: https://orcid.org/0000-0002-8087-1708)
- Sunny Sharma
- Angel Garg (ORCID: https://orcid.org/0009-0005-4742-7444)
- Aarti Singla
Institutions
- Thapar Institute of Engineering & Technology (IN)
- Indian Institute of Information Technology Allahabad (IN)
- Society of Health and Physical Educators (US)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-16
- DOI
- https://doi.org/10.21203/rs.3.rs-11538004/v1
- Primary Topic
- Quantum Computing Algorithms and Architecture
- Type
- preprint