Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Semantic Learning for Molecular Communication in Internet of Bio-Nano Things

Hanlin Cai, Ozgur B. Akan

TL;DR

The paper addresses the challenge of reliable, task-oriented communication in molecular channels for IoBNT by introducing an end-to-end semantic molecular communication (MC) framework that combines a deep semantic encoder–decoder with a probabilistic channel network to enable differentiable training under stochastic propagation. Semantic features are extracted from biomedical images, quantized into channel symbols, and decoded to yield diagnostic predictions, while a Gaussian mixture channel model captures ISI and noise for robust optimization. Key contributions include an explicit probabilistic quantization scheme, a five-layer CNN semantic feature extractor, a Gaussian mixture channel with end-to-end training, and experimental validation on the Kvasir GI image dataset showing at least 25% accuracy improvement over JPEG+LDPC baselines under resource constraints. The results demonstrate improved efficiency and robustness for task-oriented molecular communication in IoBNT, with potential impact on real-time biomedical diagnostics in resource-limited nanoscale networks.

Abstract

Molecular communication (MC) provides a foundational framework for information transmission in the Internet of Bio-Nano Things (IoBNT), where efficiency and reliability are crucial. However, the inherent limitations of molecular channels, such as low transmission rates, noise, and intersymbol interference (ISI), limit their ability to support complex data transmission. This paper proposes an end-to-end semantic learning framework designed to optimize task-oriented molecular communication, with a focus on biomedical diagnostic tasks under resource-constrained conditions. The proposed framework employs a deep encoder-decoder architecture to efficiently extract, quantize, and decode semantic features, prioritizing taskrelevant semantic information to enhance diagnostic classification performance. Additionally, a probabilistic channel network is introduced to approximate molecular propagation dynamics, enabling gradient-based optimization for end-to-end learning. Experimental results demonstrate that the proposed semantic framework improves diagnostic accuracy by at least 25% compared to conventional JPEG compression with LDPC coding methods under resource-constrained communication scenarios.

Semantic Learning for Molecular Communication in Internet of Bio-Nano Things

TL;DR

The paper addresses the challenge of reliable, task-oriented communication in molecular channels for IoBNT by introducing an end-to-end semantic molecular communication (MC) framework that combines a deep semantic encoder–decoder with a probabilistic channel network to enable differentiable training under stochastic propagation. Semantic features are extracted from biomedical images, quantized into channel symbols, and decoded to yield diagnostic predictions, while a Gaussian mixture channel model captures ISI and noise for robust optimization. Key contributions include an explicit probabilistic quantization scheme, a five-layer CNN semantic feature extractor, a Gaussian mixture channel with end-to-end training, and experimental validation on the Kvasir GI image dataset showing at least 25% accuracy improvement over JPEG+LDPC baselines under resource constraints. The results demonstrate improved efficiency and robustness for task-oriented molecular communication in IoBNT, with potential impact on real-time biomedical diagnostics in resource-limited nanoscale networks.

Abstract

Molecular communication (MC) provides a foundational framework for information transmission in the Internet of Bio-Nano Things (IoBNT), where efficiency and reliability are crucial. However, the inherent limitations of molecular channels, such as low transmission rates, noise, and intersymbol interference (ISI), limit their ability to support complex data transmission. This paper proposes an end-to-end semantic learning framework designed to optimize task-oriented molecular communication, with a focus on biomedical diagnostic tasks under resource-constrained conditions. The proposed framework employs a deep encoder-decoder architecture to efficiently extract, quantize, and decode semantic features, prioritizing taskrelevant semantic information to enhance diagnostic classification performance. Additionally, a probabilistic channel network is introduced to approximate molecular propagation dynamics, enabling gradient-based optimization for end-to-end learning. Experimental results demonstrate that the proposed semantic framework improves diagnostic accuracy by at least 25% compared to conventional JPEG compression with LDPC coding methods under resource-constrained communication scenarios.

Paper Structure

This paper contains 10 sections, 9 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Semantic Molecular Communication System
  3. Molecular Propagation Model
  4. Semantic Encoder and Decoder
  5. Semantic Feature Extraction
  6. Probabilistic Quantization
  7. Task-Specific Decoding
  8. Channel Network
  9. Experiment and Evaluation
  10. Conclusion and Future Work

Figures (3)

  • Figure 1: Illustration of the semantic molecular communication framework.
  • Figure 2: Temporal variations of SIR during the transmission of a continuous sequence of five ‘1’ symbol bits in two molecular communication scenarios.
  • Figure 3: Accuracy performance comparison between different methods in two molecular communication scenarios (BCR: Bandwidth Compression Ratio).