A survey on network simulators, emulators, and testbeds used for research and education

Abstract

Listen

Translate article

Network operators and researchers constantly search for platforms to evaluate future deployments and test new research ideas. When experimenting, they usually face challenges in deciding on an appropriate platform to validate the advantages and limitations of their proposed system. These challenges include finding an experimentation environment that balances traffic realism, scalability, and cost. An experimenter can evaluate systems, protocols, and security implementations using simulators, emulators, or testbeds to validate the expected behavior of the proposed idea. Simulators and emulators provide a controlled environment to conduct reproducible experiments but lack realism. Testbeds provide realism and scale depending on the available resources. However, real equipment can be costly and unavailable for many experimenters. The inability to test networking ideas in a realistic environment at a large scale presents a barrier for companies, institutions, and network vendors to implement new features, thus, slowing down innovation. In the past few decades, the networking community developed new platforms to test new ideas and deployments at scale, with realism, and at lower costs. These platforms also enable the instruction of networking concepts, cybersecurity, distributed computing, storage systems, and science applications. From the learner’s side, practical hands-on experience is required to internalize concepts and improve troubleshooting skills. Learning these concepts can be challenging due to the multidisciplinary nature of networking instruction, where a learner must have a background in several computing areas (e.g., operating systems, programming languages, and computer architecture). This paper presents experimentation platforms used to conduct research in computer networks and evaluates the potential of these platforms for instructing networking courses. This paper examines the literature and presents a taxonomy of network experimentation platforms. It also discusses challenges, analyzes the limitations, and suggests future perspectives by providing an overview of the tools, a description of the underlying resources (i.e., hardware and software), and a summary of the supported experiments. The paper aims to assist experimenters and educators in deciding which platform is more suitable for their experimentation needs and discuss the challenges and future directions related to the network experimentation platforms.