An Interactive Traffic Replay Method in a Scaled-Down Environment

Abstract

Network traffic replay plays an essential role in network security tests. However, performing traffic replay in a network is still a challenge due to the different scales of the live network and the replay network. In this study, we investigate three problems of the interactive traffic replay under a scaled-down replay environment and then propose a multi-node traffic replay method. In this method, a self-elected IP mapping algorithm is designed to construct the IP mapping between the target network and the live network, which reproduce the interaction between the nodes of the live network. We prove that the traffic aggregation problem is an NP-complete problem and design a divide and conquer algorithm to approximate an optimal solution. On this basis, a traffic replay algorithm based on minimum-delay forwarding mechanism is proposed to perform low-delay interactive traffic replay. The experimental results show that the method mentioned above enables us to better aggregate elephant flows and achieve high similarity in replay timing series and bandwidth, which can be employed to reproduce a real network scenario in network device tests and network security experiments.

Keywords

• Traffic replay
• IP mapping
• self-elected strategy
• traffic aggregation