@article{LIN2026, 
author = {Weiwei LIN and Shaowei XU and Chuan LIN and Fuling YE and Dong ZHANG},
title = {Practice of polymorphic network design based on P4 programmable data plane},
year = {2026},
journal = {Experimental Technology and Management},
volume = {43},
number = {3},
pages = {214-221},
keywords = {research-to-teaching conversion, P4, programmable data plane, polymorphic network, experimental teaching},
url = {https://www.sciopen.com/article/10.16791/j.cnki.sjg.2026.03.027},
doi = {10.16791/j.cnki.sjg.2026.03.027},
abstract = {ObjectiveThe growing national and societal demand for talent in the field of polymorphic networks has highlighted the relevance of advancing network education in universities. As the primary institutions for cultivating network professionals, universities are placing increasing emphasis on the development of practical, forward-looking network curricula. Given the inherently hands-on nature of networking, designing polymorphic network experiments enables students to broaden their knowledge base, gain a deeper understanding of polymorphic network theory, and strengthen their innovative capabilities. However, traditional network teaching experiments are often limited by rigid protocols, which lack flexibility, programmability, and support for innovation. Although emerging experiments based on software-defined networking (SDM) and programmable data plane (PDP) offer improvements, they largely remain confined to single-modality scenarios. Therefore, overcoming the constraints of traditional methods and creating innovative experiments that support polymorphic network design has become an urgent and valuable research focus.MethodsInspired by the strong correlation between key modality processing technologies in polymorphic networks and the capabilities of the P4-based PDP, the flexible forwarding advantages of the PDP are leveraged to redesign the original programming network experiments. By integrating cutting-edge research achievements in PNs, we developed a comprehensive experimental project—PDP-based polymorphic network—and introduced it into undergraduate teaching. Three major challenges were addressed during the design process: redefining the experimental objectives, optimizing the design and content of the experiment, and implementing effective instructional guidance. The overall goal of the designed experiment is to construct and implement a polymorphic network topology based on P4 PDP. The experiment includes multiple progressive stages, namely, single-modality network implementation, dual-modality network implementation, polymorphic network implementation, modality scheduling, and arbitration verification, as well as supporting flexible adjustment of difficulty levels. In terms of instructional support, instructors are expected to provide continuous progress tracking and comprehensive guidance throughout the experimental process.ResultsTopology and link state correctness are validated using the Mininet platform’s command-line interface. Connectivity tests confirm that hosts on both sides of the polymorphic network can communicate, and packets sent from one host are encapsulated into different modalities and forwarded to the other host using distinct addressing and routing methods. In the security test, an ARP spoofing attack is launched from a server based on CENI, confirming that the polymorphic network provides greater security than traditional single-modality networks when facing network threats. Simulated packet transmission and reception experiments verify that the polymorphic network introduces some overhead. Finally, the experiment is validated through actual implementation in an educational setting.ConclusionsThis experiment is derived from cutting-edge research in polymorphic networking. By leveraging the PDP capabilities, the experiment designs and implements multiple network forwarding modalities and supports flexible adjustment of difficulty levels. As verified through actual teaching practice, the experiment can effectively help students broaden their knowledge in the field of networking, deepen their understanding of polymorphic networks, enhance their skills in network programming, and strengthen their capacity for polymorphic network innovation.}
}