arXiv:2610.01941
Universality Sacrifices Reliability in Classical-Quantum Channel Coding
Kaito Watanabe, Masahito Hayashi, Takaya Matsuura, Hao-Chung Cheng
Abstract
Universal channel coding enables communication without a complete description of the channel. For classical channels, universal codes can attain both capacity and the optimal high-rate reliability. We show that this compatibility fails for classical-quantum channels in general; that is, the optimal reliability in the channel-aware scenario is not always achievable with universal coding due to the ignorance of the unitary rotation of the output system. We exhibit a family of classical-quantum channels for which one cannot achieve the channel-aware optimal reliability by a fixed coding scheme. We further derive a converse bound on the reliability for unitary-invariant decoders, a natural assumption for the universal coding scheme, that can be strictly smaller than the optimal channel-aware error exponent. Conversely, we construct a channel-independent encoder-decoder pair and establish a universally achievable bound on the reliability that matches this converse bound in the high-rate regime, thereby characterizing the optimal universal reliability. Specifically, the channel-aware and universal exponents are governed by the Petz and sandwiched Rényi divergences, respectively. These divergences coincide for commuting outputs but differ for noncommuting ones, explaining why universality preserves optimal reliability classically but can reduce it quantumly. Our results showcase the fundamental reliability cost of performing the classical-quantum channel coding task universally.
AI Summary
Universal classical channel codes attain capacity and the optimal high-rate error exponent without knowing the channel. This paper asks whether the same is possible for classical–quantum (cq) channels, and identifies a reliability loss associated with an unknown common unitary rotation of the outputs. For binary pure-state qubit channels with outputs $\sqrt{1-p}|0\rangle\pm\sqrt p|1\rangle$, subject to an unknown phase rotation, it derives a worst-phase error-exponent upper bound $D(t\|p)$, where $h(t)=R$ and $0<t<p<1/2$. For every $0<R<h(p)$, this bound is strictly below an achievable channel-aware random-coding exponent. A further argument gives a fixed phase at which any phase-independent code sequence fails to attain that channel-aware exponent. For constant-composition codes with input distribution $P$, the paper constructs a channel-independent encoder and decoder achieving $\sup_{1/2\leq\alpha<1}\frac{1-\alpha}{\alpha}[\widetilde I^{\mathrm{Aug}}_\alpha(P;W)-R]$, where $\widetilde I^{\mathrm{Aug}}_\alpha$ uses sandwiched Rényi divergence. A matching high-rate converse holds for unitary-invariant decoders. The channel-aware expression instead uses Petz Rényi divergence; the quantities coincide for commuting outputs but can differ otherwise. The converse reduces coding to hypothesis testing of pinched output states and uses large deviations and minimax arguments. Achievability combines a measured-Augustin variational formula, quotient decoding, Schur–Weyl operator domination, and block regularization. The matching characterization is restricted to high rates, constant composition, and the unitary-invariant decoder class; low-rate optimality remains unresolved.
The authors used OpenAI’s ChatGPT models 5.6 Sol and 6 Astra to assist with exploring proof strategies, checking proofs, and preparing the manuscript.
Central claim may not hold: The claimed characterization of optimal universal reliability exceeds the proved converse, which assumes unitary-invariant decoders and constant-composition codes without establishing a reduction from general pointwise universal schemes. The fixed-phase lemma's proof contains an unjustified nesting assertion and an incorrect Poisson-kernel estimate, leaving its stronger conclusion unsupported by the displayed argument. The measured-Augustin duality theorem is false as printed, as an explicit two-state orthogonal ensemble demonstrates, although the later proposition suggests an intended sign correction. The universal achievability proof does not establish negligible padding or overhead: its blocklength schedule can produce zero usable blocks, and its stated growing-alphabet overhead scaling contradicts the required sublinear bound.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 5+23 pages