Exact performance of the five-qubit code with coherent errors

Abstract

To well understand the behavior of quantum error correction codes (QECC) in noise processes, we need to obtain explicit coding maps for QECC. Due to extraordinary amount of computational labor that they entails, explicit coding maps are a little known. Indeed this is even true for one of the most commonly considered quantum codes-the five-qubit code, also known as the smallest perfect code that permits corrections of generic single-qubit errors. With direct but complicated computation, we obtain explicit process matrix of the coding maps with a unital error channel for the five-qubit code. The process matrix allows us to conduct exact analysis on the performance of the quantum code. We prove that the code can correct a generic error in the sense that under repeated concatenation of the coding map with itself, the code does not make any assumption about the error model other than it being weak and thus can remove the error(it can transform/take the error channel to the identity channel if the error is sufficiently small.). We focus on the examination of some coherent error models (non diagonal channels) studied in recent literatures. We numerically derive a lower bound on threshold of the convergence for the code. Furthermore, we analytically show how the code affects the average gate infidelity and diamond distance of the error channels. Explicit formulas of the two measurements for both pre-error channel and post-error channel are derived, and we then analyze the logical error rates of the aforesaid quantum code. Our findings tighten the upper bounds on diamond distance of the noise channel after error corrections obtained in literature.