SnapAI: a Simulator for Network Algorithms and Protocols with Artificial Intelligence integration

Simulating networked scenarios is vital for validating algorithmic and communication systems. To address this challenge, we introduce SnapAI, an open-source simulator for network algorithms and protocols, enhanced with native integration of Artificial Intelligence (AI) techniques for advanced network analysis, and designed in line with Software Defined Systems (SDS) principles to offer flexibility and programmability. SnapAI is structured as a modular platform that facilitates the analysis of node placement, connectivity, and message exchange. By integrating AI techniques such as classification, clustering, multidimensional embedding, and Graph Neural Networks (GNNs), SnapAI empowers developers to conduct AI-enhanced network analyses. To ensure comprehensive functionality and adaptability, SnapAI’s architecture comprises six core components: Nodes, Models, Messages, Timers, Simulator, and Graphical Interface. SnapAI’s capabilities are demonstrated through four use cases involving message exchanges and use of timers, integrating classical algorithms like Dijkstra’s and advanced AI techniques such as clustering and embedding. SnapAI’s intuitive design broadens applicability across civilian, military, and other critical networked domains.