Cross-Layer Scheduling Optimization of 5G-TSN Integrated Networks Based on QoS Mapping and Joint Scheduling
TL;DR
Simulation results demonstrate that the proposed QoS-mapping-based no-wait latency-balanced joint scheduling (QMLB-JS) algorithm improves the end-to-end deterministic transmission capability of the integrated 5G-TSN network.
Abstract
Fifth-generation (5G) and time-sensitive networking (TSN) are widely recognized as the most promising technologies for future industrial networks. Quality of service (QoS) mapping and traffic scheduling mechanisms are critical to ensuring deterministic transmission in 5G-TSN integrated networks. However, the uncertainty of 5G air-interface delay significantly reduces the deterministic guarantee for traffic in 5G-TSN networks. A QoS mapping algorithm based on incremental mini-batch K-means++ stratified sampling (IMK-S3) is proposed to quickly and accurately determine the 5G QoS identifier (5QI) values of traffic flows, even under dynamic traffic variations. Based on this, combined with cross-layer scheduling optimization, a QoS-mapping-based no-wait latency-balanced joint scheduling (QMLB-JS) algorithm is proposed. QMLB-JS supports hold and forward buffer mechanism in DS-TT and NW-TT, and realizes accurate time-based gating management by orchestrating the time of time-triggered (TT) traffic injection into the network. Simulation results demonstrate that the proposed algorithm improves the end-to-end deterministic transmission capability of the integrated 5G-TSN network.