This project develops a variety of code and scenarios for training and deploying neural network (NN) modules in Internet of Bio-Nano Things (IoBNT) applications.

This project includes various examples for training and developing NN modules across a variety of layers of IoBNT networks. The scenarios concern molecular communication (MC) channels, where transmitters and receivers operate with molecules as information carriers, as illustrated in Fig. 1 below.

Fig. 1: Representation of a molecular communication link.

The code examples include NN-based models for

• Distance estimator among cells using a feedforward NN,
• A digital synchronizer using a Reinforcement Learning agent,
• A digital decoder for binary transmissions using various architectures: Autoencoders, RNN, BiRNN, CNN, and feedforward NN architectures,
• Explainable methods of NNs based on ICE and SHAP methods,
• Simulator for modeling a microfluidic pipe.

Each example is developed within a separate folder.

The installation steps are specified within each different folder.

Usage of this code is directly described within each folder.

• Realistic model for Molecular Communication channels: The code on each file includes models and testbed data for the training and deployment of the NN Architectures.
• Tested code for training and deployed NNs: The examples in each folder include a variety of architectures such as Autoencoders, RNN, BiRNN, CNN, and feedforward NN.

Interested contributors can contact the project owners below. Please refer to the Contact Information below. We identify further developments for more complex scenarios and the integration of many other architectures.

We want to acknowledge the support provided by funding agencies

• DFG with grant numbers 16KIS1986K 16KIS1994 and 16KISK001K.
• BMBF with grant numbers DR 639/21-1.
• Germany’s Excellence Strategy with grant number EXC 2050/1.
• University SAL Labs.

Code developed in this repository followed the research work in:

Jorge Torres Gómez, Pit Hofmann, Lisa Y. Debus, Osman Tugay Basaran, Sebastian Lotter, Roya Khanzadeh, Stefan Angerbauer, Bige Deniz Unluturk, Sergi Abadal, Werner Haselmayr, Frank H. P. Fitzek, Robert Schober and Falko Dressler, “Communicating Smartly in Molecular Communication Environments: Neural Networks in the Internet of Bio-Nano Things,” arXiv, Jun. 2025. Link

• Name: Jorge Torres Gómez

  

  

  

  

• Name: Lisa Y. Debus

  

  

  

  

• Name: Roya Khanzadeh

  

  

  

  

• Name: Osman Tugay Başaran