Co-Opting Linux Processes for High-Performance Network Simulation (2022)
3 hours ago
- Network experimentation tools are vital for developing and testing distributed systems, but state-of-the-art simulation tools face efficiency and architectural challenges.
- The paper presents Phantom, a novel tool that combines a discrete-event network simulator with unmodified Linux processes, using process control, system call interposition, and data transfer methods.
- Phantom outperforms existing tools in large-scale benchmarks: up to 2.2× faster than Shadow, up to 3.4× faster than NS-3, and up to 43× faster than gRaIL, with comparable performance to Shadow in Tor network simulations.
- The paper received the Best Paper award at USENIX ATC 2022 and is freely available under Open Access.