Engineering for the Unknown: Open-Source 6G Experimentation Toolbox for Protocol Innovation and High-Fidelity Testbed Evaluation

Presentation at the II xGMobile International Workshop (Inatel, Brazil)

Abstract: The evolution toward 6G introduces profound uncertainty regarding future protocol stacks, data-plane abstractions, cognitive core functions, and AI-assisted control mechanisms. Anticipated 6G Core innovations, such as semantic- and intent-aware network functions, predictive mobility and slicing controllers, real-time telemetry loops, and data-driven APIs, require experimental platforms capable of rapid prototyping and validation. To address these emerging needs, this talk presents an expanding open-source toolbox for programmable networking research, testing, and experimentation. Building on P4/Tofino hardware and software-based data-plane programmability, we discuss recent advances in hardware-accelerated emulation, high-fidelity traffic generation, and custom 6G protocol prototyping. The talk highlights ongoing research efforts on open-source tools, such as the P4 Programmable Patch Panel (P4), the PIPO-TG and P4R traffic generators, for high-fidelity, experimentally driven research on 6G Core network functions. Together, these tools create a unified experimental environment for evaluating new protocol headers, adaptive pipelines, and complex traffic patterns, ranging from TSN flows and stateful TCP/QUIC to VR/XR, holographic media, and point-cloud streaming workloads. Beyond supporting cutting-edge research, these platforms have been adopted in educational labs and courses, enabling students and researchers to design, deploy, and test realistic 5G/6G scenarios. The resulting open and extensible ecosystem provides a practical pathway for iterative 6G innovation, allowing the community to explore programmable data planes, validate early 6G Core function designs, integrate AI/ML-based network intelligence, and test standardization ideas before they mature.