CS Paper Radar

A Tight Second-Order Converse Bound for Variable-Length Feedback Codes

Recep Can Yavas

cs.IT ·submitted 2026-09-10 11:02 UTC · v1

Abstract

We study variable-length feedback (VLF) codes over a discrete memoryless channel under average decoding-time and error-probability constraints. In the non-vanishing error probability regime, Polyanskiy, Poor, and Verdú (2011) derive achievability and converse bounds on the logarithm of the maximum achievable codebook size. These bounds establish the $ε$-capacity but leave an order-$\log N$ gap in the second-order expansion, where $N$ is the average decoding time. Yavas and Tan (2025) improve the coefficient of $\log N$ in the achievability bound from $-1$ to $-\frac{C}{C_1}$, where $C$ is the channel capacity and $C_1$ is the largest Kullback--Leibler divergence between two conditional output distributions. We derive a converse with the same coefficient, establishing the second-order fundamental limit for every positive-capacity discrete memoryless channel with finite $C_1$. The result also covers the moderate-deviations and error-exponent regimes, including polynomially decaying error probabilities. The converse uses Rényi entropy and the extrinsic Jensen--Shannon divergence. We also derive necessary properties of asymptotically optimal VLF codes. First-order-optimal codes must have an early-stopping branch, and second-order-optimal codes must additionally exhibit communication and confirmation behavior. Finally, for the binary erasure channel, we determine the exact minimum expected decoding time for every message-set size and admissible error probability.

Comment
26 pages, submitted to Information Theory Transactions

Closest papers

Nearest neighbours by meaning across the whole index.

The Sample Complexity of Channel Coding

cs.IT math.ST ·2026-11-16 ·Terence Viaud, Ioannis Kontoyiannis
similarity 0.75
Abstract · 31 pages, 3 figures

We introduce a new framework for the study and evaluation of the fundamental limits of channel coding, emphasizing genuinely nonasymptotic results. The {\em channel coding sample complexity} $n^*({\cal W},ε)$ of a channel ${\cal W}$ is defined as the smallest blocklength at which it is possible to transmit information through that channel with error probability below $ε$ and at a rate at least as high as that given by an appropriate constraint also in terms of $ε$. This formulation follows analogous developments in statistics, computer science and, more recently, in lossless source coding. For any discrete memoryless channel we derive nonasymptotic bounds on the sample complexity, with explicit constants. These bounds show that the sample complexity is characterized not by the usual channel capacity, but by the channel's Augustin--Csiszár capacity of order $1/2$, which coincides with the cutoff rate as well as with the channel's Sibson capacity of order $1/2$. The proof is partly based on a new channel-coding one-shot converse in terms of hypothesis testing. We also examine extensions of the sample complexity framework to general target rates. Moreover, we derive corresponding results for the sample complexity of the additive white Gaussian noise channel with a quadratic power constraint. Their proof relies on new, explicit upper bounds on the best achievable error probability, which may be of independent interest.

0.32

Second-Order Asymptotics for the Gaussian Multiple-Access Channel at Corner Points

cs.IT ·2026-08-18 ·Vincent Y. F. Tan
theory similarity 0.72
Abstract · 28 pages, 4 figures

We establish exact second-order coding rate regions at the two corner points of the capacity region of the two-user Gaussian multiple-access channel. For any average error probability $\varepsilon\in(0,1)$, we characterize the $n^{-1/2}$-scale fluctuations of achievable rates around each corner point, proving a converse that matches the previously known achievability bound. The proof first extracts a rectangular subcode that preserves the independence of the two transmitted codewords. The Polyanskiy--Verdú good-code output distribution theorem then yields log-determinant constraints that induce a spectral decomposition of a trimmed codebook into diffuse and exceptional subspaces. An entropic Brascamp--Lieb projection inequality controls the codeword inner product on the diffuse subspace, while variance-inflated Gaussian output distributions handle the low-dimensional exceptional subspace. Combining these two treatments yields the joint Gaussian limit required for the matching second-order converse.

0.47

Fano's Inequality and Channel Capacity in the Individual-Sequence Regime

cs.IT ·2026-11-04 ·Neri Merhav
theory similarity 0.70
Abstract · 31 pages; submitted for publication

We derive an individual-sequence analogue of Fano's inequality, replacing entropy by Lempel--Ziv (LZ) complexity and probability of error by empirical error frequency. In the setting considered, an arbitrary encoder maps a message to a channel input sequence $\{x_t\}$, the channel maps the input to an output $\{y_t\}$ in a completely arbitrary manner, and a block finite-state decoder produces an estimate of the channel input from the output. The individual-sequence Fano inequality lower-bounds the empirical symbol-error frequency in terms of the LZ complexity of the channel input conditioned on the channel output. This leads to a channel coding converse: reliable communication at a given rate is possible only if that rate does not exceed the individual-sequence LZ channel capacity, defined as the LZ mutual information maximized over channel inputs and minimized (worst-case) over a noise set. We also prove a matching achievability theorem for the class of \emph{symbol-wise invertible} channels, in which the channel output at each time $t$ is $y_t = f(x_t, v_t)$, $v_t$ being a `noise' variable, and $f(\cdot,v)$ is a bijection for every fixed $v$. This class includes group-additive channels $y_t = x_t \oplus v_t$ studied by Feder and coworkers. For such a channel, any rate below the LZ capacity is achievable with vanishing per-message error frequency, simultaneously for every $v^n$ in a given noise set $\mathcal{N}$, provided $\mathcal{N}$ is closed under block permutations; this condition is satisfied by any noise set that depends on $v^n$ only via its type class with respect to such blocks. The achievability uses a universal decoder that requires no knowledge of the channel and no feedback.

0.22

Monotonicity of the Rényi channel capacity under non-signaling assisted channel simulation

cs.IT quant-ph ·2026-09-28 ·Tina Shariat, Matt Hoogsteder-Riera, Marco Tomamichel
similarity 0.69
Abstract · 25 pages, 1 figure

We give an alternative variational proof that non-signaling correlations shared between the sender and receiver cannot increase the Rényi capacity of a classical channel. This monotonicity follows from the data-processing inequality for channel divergences established by Oufkir, Yao and Berta. Our approach works directly with the information-radius representation of the capacity. The constraints defining non-signaling boxes are linear in the box, while the Rényi capacity depends on the channel nonlinearly. We use a Legendre transform to express the capacity as an extremum over a family of affine functionals in which the channel appears linearly. Combining this method with separate arguments for the Shannon capacity and endpoint orders, we recover monotonicity at every order from zero to infinity. We then derive consequences for standard coding exponent functions: non-signaling simulation cannot increase the random-coding or sphere-packing exponent below capacity, or decrease the strong-converse exponent above capacity.

0.35

MacKay-Neal Codes Achieve Capacity under MAP Decoding

cs.IT ·2026-10-07 ·Kenta Kasai
theory similarity 0.68
Abstract · 22 pages

Statistical-mechanical analyses predict that MacKay-Neal codes can achieve channel capacity at fixed degrees. We prove this prediction for the uncoupled $(\ell,3,3)$ MN socket ensemble for every fixed integer $\ell\ge4$. For each binary-input memoryless symmetric channel of capacity strictly greater than $3/\ell$, the ensemble-average block error probability under maximum a posteriori (MAP) decoding is $O(\log N/N)$, where $N$ is the transmitted blocklength. The actual transmitted rate converges to $3/\ell$. The result includes both punctured and transmitted variables and does not condition the sparse square matrix on invertibility. The proof establishes an entropy inequality for six bits subject to even parity by an analytic argument that reduces the domain by symmetry, restricts its interior stationary points, where both partial derivatives vanish, to the diagonal, and controls the resulting scalar functions by explicit polynomial bounds. Exact configuration counts then bound the conditional entropy on the binary symmetric channel. A standard comparison by mutual information extends the entropy bound to general symmetric channels; independent output erasures and a minimum-distance estimate for the transmitted code yield the bit and block error bounds.

0.43

Information-Spectrum General Formulas for the Capacity Region of Classical-Quantum Multiple Access Channels

cs.IT quant-ph ·2026-11-05 ·Takuya Kubo, Hiroshi Nagaoka
similarity 0.68
Abstract · 11 pages

For classical-quantum multiple access channels (cq-MACs) with two senders, we give several general expressions for the capacity region, valid for arbitrary sequences of channels. In the case of asymptotically vanishing error ($\varepsilon=0$), the capacity region is expressed in terms of spectral inf-mutual information rates. For every $0\le\varepsilon<1$, the $\varepsilon$-capacity region is expressed in terms of a quantity defined through a linear combination of three alternative hypotheses. In particular, the outer bound of Kubo and Nagaoka (2015) is tight. Our results include the quantum counterpart of the general formula of Han (1998) for the classical MAC. The most important ingredient is the one-shot inner bound of Sen (2021), which we apply to each member of an arbitrary channel sequence.

0.25

Fine asymptotics for approximate simulation of symmetric channels

cs.IT ·2026-10-11 ·Valentinian Lungu, Aaron B. Wagner, Ioannis Kontoyiannis
similarity 0.68
Abstract

We study the approximate channel simulation problem in the normal approximation regime. Our starting point is the fact established by Cao et al.~\cite{Cao_channel_simulation}, that the minimal codebook size $M_ε^*(W)$ for a fixed error tolerance $ε>0$ is closely related to the $ε$-smooth channel max-information, $I_{\max}^ε(W)$. Using this, we refine earlier large-blocklength approximations to $M_ε^*(W^{\otimes n})$ for discrete memoryless channels. Specifically, we provide tight third-order expansions for Gallager-symmetric channels and demonstrate that the third-order term is channel-dependent. As in corresponding channel coding expansions, we identify a substantial rate gap between singular and nonsingular channels. Notably, the direction of this gap is reversed relative to channel coding: The third-order term is $-\frac{1}{2}\log n$ for singular channels but $O(\log\log n)$ for nonsingular channels.

0.21

An Upper Bound to the Undetected Error Exponent with Erasure Decoding at Positive Rates

cs.IT ·2026-10-11 ·Seyed AmirPouya Moeini, Marco Dalai, Albert Guillén i Fàbregas
theory similarity 0.67
Abstract

We study the undetected-error exponent of discrete memoryless channels with erasure decoding. Assuming that the undetected-error and erasure probabilities vanish, we derive an upper bound on the undetected-error exponent at positive rates. The bound uses Somekh-Baruch's genie-aided argument where an auxiliary output provides the decoder with a list of codewords. It recovers Telatar's tight rate-zero converse and is at least as tight as the straight-line bound. A numerical example shows that the improvement over the straight-line bound can be strict.

0.16