This paper introduces Text2Net, an innovative system designed to transform plain English descriptions into dynamic, interactive network simulations within the Emulator Virtual Engine–Next Generation (Eve-NG) environment. By integrating SOTA technologies from Natural Language Processing (NLP), Large Language Models (LLMs), and proposed adaptor software, Text2Net bridges the technical knowledge gap, enabling both technical and non-technical users to effortlessly create and interact with complex network topologies within a simulation environment. The system architecture combines an intuitive chat interface, GPT4 LLM to interpret user inputs, NLP key-value extraction, a simulation adaptor, and the EVE-NG engine. TextNet democratizes network emulation, empowering educators to efficiently construct simulations for interactive learning. It also benefits industrial prototyping and testing configurations.CCS CONCEPTS•Applied computing → Interactive learning environments; Computer-assisted instruction; IT architectures;•Networks → Network design principles; Programming interfaces; Topology analysis and generation; Logical / virtual topologies; Network manageability; Programmable networks; Network management; Network monitoring.

The goal of this research is to model the spatio-temporal dependencies of radon gas generation and movement underground. In this regard, we have embarked upon an interdisciplinary research effort that involves studying the dependencies of radon gas emanation with soil and environmental parameters. To this end we design, implement and deploy an innovative real-time sensor network, OpenRadon Lab, to develop a radon prediction model that maps its distribution along space and time. This network constitutes a soil-to-cloud wireless computing framework to enable machine learning assisted soil radon prediction, and optimized data offloading to conserve computing resources on the sensing devices.