Core Concepts
A sliding window decoder based on belief propagation with guided decimation is proposed to efficiently decode quantum low-density parity-check codes in the presence of circuit-level noise.
Abstract
The content introduces a sliding window decoder based on belief propagation (BP) with guided decimation for the purpose of decoding quantum low-density parity-check (QLDPC) codes in the presence of circuit-level noise.
Key highlights:
Windowed decoding is used to keep the decoding complexity reasonable when repeated rounds of syndrome extraction are required to decode.
Within each window, several rounds of BP with decimation of the variable node expected to be most likely to flip in each round are employed.
Ensemble decoding is used to keep both decimation options (guesses) open in a small number of chosen rounds, resulting in a guided decimation guessing (GDG) decoder.
Applied to bivariate bicycle codes, GDG achieves similar logical error rate as BP with an additional OSD post-processing stage (BP+OSD) and combination-sweep of order 10.
For a window size of three syndrome cycles, a multi-threaded CPU implementation of GDG achieves a worst-case decoding latency of 3ms per window for the [[144,12,12]] code.
Stats
The worst-case decoding latency of the GDG decoder is 3ms per window for the [[144,12,12]] code.