Joint scheduling and routing optimization in time-sensitive networks

The rise of new technological applications, particularly in the domains of automation and autonomous vehicles, necessitates deterministic and reliable network communication with real-time responsiveness. However, existing methods frequently prioritize scheduling Time-Triggered (TT) traffic at the expense of optimizing lower- priority traffic, such as Best-Effort (BE), and often neglect joint routing considerations. Furthermore, the critical aspect of reliability in TT transmissions is often overlooked. This need has spurred the development and adoption of Time- Sensitive Networking (TSN), a set of evolving IEEE standards aiming to transform standard Ethernet for the demands of time-critical applications. This work introduces two novel approaches to address these limitations: Optimized Hybrid Deterministic Scheduling and Routing (OHDSR) and its enhanced version, OHDSR+. Both methods prioritize communication efficiency and jointly optimize scheduling and routing for TT traffic while accommodating the requirements of BE traffic. While OHDSR ensures timely delivery of both TT and BE communications, OHDSR+ extends this capability by guaranteeing the reliability of TT transmissions through redundant paths and time shift mechanisms. In the experimental setup, a CP-Sat Solver from Google's OR-Tools was utilized for optimization. This was implemented using Python 3.11 within the Visual Studio Code environment. A diverse dataset was generated via a custom Python algorithm that leveraged the NetworkX library to create realistic network topologies. Problem instances were encoded in Extensible Markup Language (XML) for seamless integration. The dataset varied in size and complexity, encompassing diverse parameters such as streams, bridges, end-systems, applications, and tasks. Experiments were conducted on a standardized platform featuring an Intel Core i7- 11370H CPU (3.30 GHz) with 8GB RAM, ensuring consistency and reliability in performance evaluations. The OHDSR and OHDSR+ frameworks incorporate a gate control mechanism to prioritize TT traffic and manage gate closing times, preventing excessive delays for BE traffic. This results in stable worst-case latency for both traffic types. Furthermore, OHDSR+ demonstrates superior efficiency in handling higher redundancy levels, maintaining performance even under extreme redundancy scenarios. The detailed analysis of its performance within TSN reveals significant improvements in scheduling and routing optimization compared to existing methods, particularly in larger and more complex network topologies. OHDSR excels in minimizing total latency while guaranteeing acceptable worst-case latency for both TT and BE traffic. Its scalability across various network sizes is noteworthy, significantly outperforming similar approaches in terms of response times. Building upon this foundation, OHDSR+ further enhances network reliability by incorporating redundant paths for TT traffic transmission and implementing a time shift mechanism to ensure temporal diversity. This effectively mitigates the impact of potential failures and ensures the continuous flow of critical data.

Text FK 2024 61 - Full Text.pdf Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (3MB) Text FK 2024 61.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (3MB)

Actions (login required)

View Item