Traffic Shaping at the Edge: Enabling Bounded Latency for Large-scale Deterministic Networks
Guoyu Peng, Shuo Wang, Yudong Huang, Ru Huo, Tao Huang, Yunjie Liu
Abstract
With the advent of 5G and the evolution of Internet protocols, the industrial internet is driving the need for deterministic Quality of Service (Qos) guarantees. Recent researches have leveraged Cyclic Queuing and Forwarding (CQF) mechanism to achieve the deterministic transmission in Time-sensitive Networking (TSN). However, most of them are difficult to implement in Large-scale Deterministic Networks (LDNs) due to long link propagation latency and global time synchronization.In this paper, we design large-scale deterministic network flow shaping mechanism (LFS), a simple yet effective solution to guarantee deterministic worst-case latency and zero packet loss for time-sensitive (TS) flows in LDNs. Based on Cycle Specified Queuing and Forwarding (CSQF) mechanism of multi-queue circular forwarding, LFS specifies the start time for TS flow injected into the Wide Area Network (WAN) to improve the bandwidth resource utilization and mitigate the impact of additional delays caused by flow convergence. Compared with the solution without using the LFS mechanism, experimental results demonstrate that LFS had better schedulability and resource utilization while enabling bounded latency for LDNs.