Model-Driven Deep Learning for Distributed Detection with Low-Bit Quantization

Abstract

Distributed detection plays a fundamental role in communication-efficient collaborative inference over large-scale networks, where spatially distributed nodes must cooperate using severely quantized messages due to bandwidth constraints. While low-bit quantization is essential for reducing communication overhead, it can significantly degrade detection performance, particularly when prior knowledge of observation noise is unavailable. To address this challenge, we propose a model-driven deep learning framework for distributed detection with low-bit quantization. For fixed local quantizers, we first derive the exact finite-sensor maximum a posteriori (MAP) detection error and prove that, under conditionally independent and identically distributed (i.i.d.) observations, independent local randomizations, and a common quantization-level constraint, identical quantizers preserve the supremum of the average network Chernoff information. Guided by these results, we develop a binary probabilistic quantizer with a trainable neural probability controller and design a tree-structured probabilistic quantizer for scalable multi-bit quantization. The exact finite-sensor MAP error provides the training objective for the binary quantizer, while we derive a computationally tractable Bhattacharyya surrogate objective for training the multi-bit quantizer. For both cases, we derive analytical MAP fusion rules based on the class-conditional probability mass functions (PMFs) of the quantized messages. When the observation statistics are unknown, these PMFs are estimated from held-out labeled data, yielding plug-in MAP detectors. Simulation results show that the proposed method approaches centralized-oracle performance at higher quantization resolutions and remains effective under changes in network size and test-time signal amplitude.

Adding Heavy-Ball Momentum to Distributed Constrained Optimization for Energy Resource Management Subject to Time-Delays and Quantization
Mohammadreza Doostmohammadian, Sergio Pequito
arXiv17 Oct 2026PreprintCC BY 4.0Full textDynamical systems Optimisation & controlarXiv
Deep Learning-Based Object Detection for Fashion, Uniform, and PPE Items: A Systematic Literature Review
Isa Aulia Almadani, Nurcahya Pradana Taufik Prakisya, Cucuk Wawan Budiyanto
Journal of Information Systems and Informatics24 Oct 2026Peer-reviewedCC BYFull textGeneral mathematicsDOAJ
Nonparametric Intensity Estimation for Covariate-Driven Poisson Processes Using Deep Learning
Atsutomo Yara, Yoshikazu Terada
arXiv2 Nov 2026PreprintarXiv licenceFull textStatisticsarXiv
Context-Guided Traversal with Deep Reinforcement Learning for Suspicious Subgraph Detection in Cryptocurrency Transaction Graphs
Seok-Jun Buu, Hae-Jung Kim
Mathematics27 Oct 2026Peer-reviewedCC BY 4.0General mathematicsCrossref
An Input-Frugal Deep Learning Framework for Weather-Driven National Crop-Yield Forecasting: A Case Study of Brazilian Soybean
Fernando Dupin da Cunha Mello, Prashant Kumar, Erick G. Sperandio Nascimento
arXiv29 Sep 2026PreprintarXiv licenceFull textStatisticsarXiv

Cite

BibTeX RIS

BibTeX

@misc{guo2026modeldriven,
  title = {{Model-Driven Deep Learning for Distributed Detection with Low-Bit Quantization}},
  author = {Wei Guo and Meng He and Chuan Huang and Hengtao He and Shenghui Song and Jun Zhang and Khaled B. Letaief},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.21759},
  eprint = {2610.21759},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.21759},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Model-Driven Deep Learning for Distributed Detection with Low-Bit Quantization
AU  - Wei Guo
AU  - Meng He
AU  - Chuan Huang
AU  - Hengtao He
AU  - Shenghui Song
AU  - Jun Zhang
AU  - Khaled B. Letaief
PY  - 2026
DA  - 2026/10/15
PB  - arXiv
DO  - 10.48550/arxiv.2610.21759
UR  - https://arxiv.org/abs/2610.21759
AB  - Distributed detection plays a fundamental role in communication-efficient collaborative inference over large-scale networks, where spatially distributed nodes must cooperate using severely quantized messages due to bandwidth constraints. While low-bit quantization is essential for reducing communication overhead, it can significantly degrade detection performance, particularly when prior knowledge of observation noise is unavailable. To address this challenge, we propose a model-driven deep learning framework for distributed detection with low-bit quantization. For fixed local quantizers, we first derive the exact finite-sensor maximum a posteriori (MAP) detection error and prove that, under conditionally independent and identically distributed (i.i.d.) observations, independent local randomizations, and a common quantization-level constraint, identical quantizers preserve the supremum of the average network Chernoff information. Guided by these results, we develop a binary probabilistic quantizer with a trainable neural probability controller and design a tree-structured probabilistic quantizer for scalable multi-bit quantization. The exact finite-sensor MAP error provides the training objective for the binary quantizer, while we derive a computationally tractable Bhattacharyya surrogate objective for training the multi-bit quantizer. For both cases, we derive analytical MAP fusion rules based on the class-conditional probability mass functions (PMFs) of the quantized messages. When the observation statistics are unknown, these PMFs are estimated from held-out labeled data, yielding plug-in MAP detectors. Simulation results show that the proposed method approaches centralized-oracle performance at higher quantization resolutions and remains effective under changes in network size and test-time signal amplitude.
ER  -

Abstract

Related

Cite