Continual Learning of Feedback-based Molecular Communication

arXiv cs.LG / 5/5/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces a continual learning (CL)-based performance estimation method for feedback-based molecular communication, using sequential simulation experiments to estimate performance.
As the communication protocol is tested across multiple experimental settings, the CL estimators learn new (previously unexperienced) estimation tasks while preserving performance on earlier tasks.
The approach is implemented on a standard neural network architecture by modifying the loss function with tailored regularization and replay strategies.
Experiments show that the estimators can learn from a continuous stream of simulation results and outperform a baseline neural network in estimation accuracy under different compute budgets.
Overall, the work aims to clarify and establish how continual learning techniques can be applied and benefit molecular communication research.

Abstract

This paper proposes and evaluates a new performance estimation method that leverages continual learning (CL) algorithms to carry out sequential simulation experiments for a feedback-based molecular communication protocol. As the protocol is sequentially examined in various experimental settings, the proposed CL-based performance estimators incrementally learn a series of unexperienced estimation tasks without compromising those that have been learned in the past. They are designed to work on a standard neural network architecture by customizing regularization and replay strategies in the loss function. Experimental results demonstrate that the proposed estimators can effectively learn on a continuous stream of simulation results and enhance the baseline neural network by improving estimation accuracy at a variety of computational costs. This paper's contribution is to establish the implications of CL in the field of molecular communication.