arXiv subjects

Alexander Sauter

Publications and source records attributed to Alexander Sauter.

4 recordsLinked to original sources

arXiv2026

Polar and Convolutional Codes for the Unequal Message Protection Problem

This paper proposes the design of polar and convolutional coset codes for the unequal message protection (UMP) in the short blocklength regime, to overcome the rate loss introduced by preamble-based solutions. After providing conditions to ensure message class disjointness, a two-step decoding architecture is proposed: it first identifies the message class via a likelihood ratio test--computable exactly for convolutional codes and approximated for polar codes--and subsequently performs maximum (or near) likelihood decoding among the codewords of the chosen message class. Numerical results show that our construction closely tracks finite-length benchmarks. Specifically, the analyzed CRC-aided polar codes perform comparable to existing polar code approaches, without requiring specific code design, while offering a robust and spectrally efficient solution for UMP scenarios.

cs.IT↗

arXiv2025

Error Detection Based on Generalized Successive Cancellation List Decoding for Polar Codes

Successive cancellation list (SCL) decoding has been widely adopted for polar codes, which allows near maximum likelihood performance with sufficiently large list size. In this work, we show that, if the list size is $2^\gamma$, where $\gamma$ is the fundamental quantity called mixing factor, then a modification to SCL decoding can implement Forney's generalized decoding rule. Hence, it provides an efficient means to discard unreliable decisions. The performance achieved by short polar codes under the proposed generalized SCL decoding is analyzed via Monte Carlo simulations.

cs.IT↗

arXiv2025

Undetected Error Probability in the Short Blocklength Regime: Approaching Finite-Blocklength Bounds with Polar Codes

We analyze the trade-off between the undetected error probability (i.e., the probability that the channel decoder outputs an erroneous message without detecting the error) and the total error probability in the short blocklength regime. We address the problem by developing two new finite blocklength achievability bounds, which we use to benchmark the performance of two coding schemes based on polar codes with outer cyclic redundancy check (CRC) codes -- also referred to as CRC-aided (CA) polar codes. The first bound is obtained by considering an outer detection code, whereas the second bound relies on a threshold test applied to the generalized information density. Similarly, in the first CA polar code scheme, we reserve a fraction of the outer CRC parity bits for error detection, whereas in the second scheme, we apply a threshold test (specifically, Forney's optimal rule) to the output of the successive cancellation list decoder. Numerical simulations performed on the binary-input AWGN channel reveal that, in the short-blocklength regime, the threshold-based approach is superior to the CRC-based approach, both in terms of bounds and performance of CA polar code schemes. We also consider the case of decoding with noisy channel-state information, which leads to a mismatched decoding setting. Our results illustrate that, differently from the previous case, in this scenario, the CRC-based approach outperforms the threshold-based approach, which is more sensitive to the mismatch.

cs.IT↗

arXiv2022

A Deep Variational Approach to Clustering Survival Data

In this work, we study the problem of clustering survival data $-$ a challenging and so far under-explored task. We introduce a novel semi-supervised probabilistic approach to cluster survival data by leveraging recent advances in stochastic gradient variational inference. In contrast to previous work, our proposed method employs a deep generative model to uncover the underlying distribution of both the explanatory variables and censored survival times. We compare our model to the related work on clustering and mixture models for survival data in comprehensive experiments on a wide range of synthetic, semi-synthetic, and real-world datasets, including medical imaging data. Our method performs better at identifying clusters and is competitive at predicting survival times. Relying on novel generative assumptions, the proposed model offers a holistic perspective on clustering survival data and holds a promise of discovering subpopulations whose survival is regulated by different generative mechanisms.

cs.LG↗