# Research on a Hybrid Scheduling Algorithm Based on Critical-Link Optimization for Large-Scale Time-Triggered Ethernet

## TL;DR

This paper introduces a new hybrid scheduling algorithm for large-scale Time-Triggered Ethernet networks to improve efficiency and reduce delays.

## Contribution

A novel hybrid scheduling algorithm combining critical-link optimization and an improved genetic algorithm for large-scale TTE networks is proposed.

## Key findings

- The proposed algorithm reduces computation time and increases message scheduling success rates.
- It improves TT message balance and reduces BE message transmission delay and jitter under higher loads.
- Simulation results show better performance compared to traditional algorithms in large-scale TTE scenarios.

## Abstract

With the rapid development of the Industrial Internet of Things (IIoT), the application scale of Time-Triggered Ethernet (TTE) technology in the IIoT has been increasingly expanding. To address the issues of rapidly increasing computation time and deteriorating scheduling quality in traditional scheduling algorithms for large-scale TTE applications, this paper proposes a hybrid scheduling algorithm based on critical-link optimization. A large-scale TTE message scheduling model is established based on the characteristics of Time-Triggered (TT) messages, and the constraints of TT scheduling are mathematically abstracted. After identifying the critical link of the network, a time slot balancing scheduling algorithm based on static priority is adopted for the link. The algorithm searches for the optimal scheduling time of current message by time-sliding within the current maximum time gap of TT messages from the center to both sides, maximizing the balance of TT message intervals to reduce the impact on Best-Effort (BE) message transmission performance. An improved genetic algorithm is proposed for the scheduling of the entire network to further enhance the global optimization capability, which takes the scheduling results of the critical link as the genes of initial population. The TT scheduling constraints are converted into the fitness function and the optimized genetic operators are developed for the genetic algorithm. Simulation results showed that the proposed algorithm can significantly reduce computing time and increase the success rate of message scheduling. At the same time, the scheduling results exhibit a better degree of TT message balance and can effectively reduce the transmission delay and jitter of BE messages as message load increases compared with traditional algorithms, making it better meet the scheduling requirements of large-scale TTE application scenarios.