Researchers Develop Flexible Scheduling Framework to Boost Network Capacity for Time-Sensitive Communications

FCP researchers from the Singapore University of Technology and Design (SUTD) and collaborators have introduced a new approach that could significantly improve the efficiency of future industrial and mission-critical communication networks.

Many emerging applications, from smart factories and autonomous vehicles to 5G/6G ultra-reliable communications, rely on the timely delivery of data packets with strict deadlines. Existing systems typically reserve network resources in a fixed, periodic pattern, which lowers the number of data streams allowed when data streams operate with different cycle lengths.

In this work, the team proposes ๐—›๐˜†๐—ฝ๐—ฒ๐—ฟ๐—ฐ๐˜†๐—ฐ๐—น๐—ฒ-๐—Ÿ๐—ฒ๐˜ƒ๐—ฒ๐—น ๐—™๐—น๐—ฒ๐˜…๐—ถ๐—ฏ๐—น๐—ฒ ๐—ฆ๐—ฐ๐—ต๐—ฒ๐—ฑ๐˜‚๐—น๐—ถ๐—ป๐—ด (๐—›๐—™๐—ฆ), a new scheduling framework that enables more flexible network resource allocation while maintaining the deterministic performance required by time-sensitive applications, such as industrial automation, avionics communications, and interactive communication services. Rather than forcing every packet in a data stream to follow the same schedule, HFS intelligently adapts resource reservations across a larger scheduling horizon, creating many more opportunities to fit traffic efficiently.

The researchers show that HFS can schedule ๐—ฎ๐—ฝ๐—ฝ๐—ฟ๐—ผ๐˜…๐—ถ๐—บ๐—ฎ๐˜๐—ฒ๐—น๐˜† ๐Ÿญ.๐Ÿฑ ๐˜๐—ถ๐—บ๐—ฒ๐˜€ ๐—บ๐—ผ๐—ฟ๐—ฒ ๐—ฐ๐—ผ๐—บ๐—บ๐˜‚๐—ป๐—ถ๐—ฐ๐—ฎ๐˜๐—ถ๐—ผ๐—ป ๐—ณ๐—น๐—ผ๐˜„๐˜€ than conventional fixed cyclic scheduling approaches while maintaining strict timing guarantees. They also developed an efficient least-load-first algorithm (HFS-LLF), that achieves at least 90% of the optimal performance while running more than 10,000 times faster than exact optimisation methods.

As factories, autonomous vehicles, and connected machines become increasingly dependent on real-time communications, smarter network scheduling techniques such as HFS could help support more devices and services without requiring additional network capacity.

๐˜—. ๐˜ž๐˜ข๐˜ฏ๐˜จ, ๐˜š. ๐˜š๐˜ฐ๐˜ถ๐˜ณ๐˜ข๐˜ท, ๐˜‰. ๐˜Š๐˜ฉ๐˜ฆ๐˜ฏ, ๐˜. ๐˜“๐˜ช, ๐˜. ๐˜ž๐˜ข๐˜ฏ๐˜จ, ๐˜ข๐˜ฏ๐˜ฅ ๐˜. ๐˜ก๐˜ฉ๐˜ข๐˜ฏ๐˜จ, โ€œ๐˜›๐˜ฉ๐˜ฆ ๐˜›๐˜ฉ๐˜ณ๐˜ฐ๐˜ถ๐˜จ๐˜ฉ๐˜ฑ๐˜ถ๐˜ต ๐˜Ž๐˜ข๐˜ช๐˜ฏ ๐˜ฐ๐˜ง ๐˜๐˜บ๐˜ฑ๐˜ฆ๐˜ณ๐˜ค๐˜บ๐˜ค๐˜ญ๐˜ฆ-๐˜“๐˜ฆ๐˜ท๐˜ฆ๐˜ญ ๐˜™๐˜ฆ๐˜ด๐˜ฐ๐˜ถ๐˜ณ๐˜ค๐˜ฆ ๐˜™๐˜ฆ๐˜ด๐˜ฆ๐˜ณ๐˜ท๐˜ข๐˜ต๐˜ช๐˜ฐ๐˜ฏ ๐˜ง๐˜ฐ๐˜ณ ๐˜›๐˜ช๐˜ฎ๐˜ฆ-๐˜›๐˜ณ๐˜ช๐˜จ๐˜จ๐˜ฆ๐˜ณ๐˜ฆ๐˜ฅ ๐˜Š๐˜ฐ๐˜ฎ๐˜ฎ๐˜ถ๐˜ฏ๐˜ช๐˜ค๐˜ข๐˜ต๐˜ช๐˜ฐ๐˜ฏ,โ€ ๐˜๐˜Œ๐˜Œ๐˜Œ ๐˜›๐˜ณ๐˜ข๐˜ฏ๐˜ด๐˜ข๐˜ค๐˜ต๐˜ช๐˜ฐ๐˜ฏ๐˜ด ๐˜ฐ๐˜ฏ ๐˜Š๐˜ฐ๐˜ฎ๐˜ฑ๐˜ถ๐˜ต๐˜ฆ๐˜ณ-๐˜ˆ๐˜ช๐˜ฅ๐˜ฆ๐˜ฅ ๐˜‹๐˜ฆ๐˜ด๐˜ช๐˜จ๐˜ฏ ๐˜ฐ๐˜ง ๐˜๐˜ฏ๐˜ต๐˜ฆ๐˜จ๐˜ณ๐˜ข๐˜ต๐˜ฆ๐˜ฅ ๐˜Š๐˜ช๐˜ณ๐˜ค๐˜ถ๐˜ช๐˜ต๐˜ด ๐˜ข๐˜ฏ๐˜ฅ ๐˜š๐˜บ๐˜ด๐˜ต๐˜ฆ๐˜ฎ๐˜ด, 2026. ๐˜‹๐˜–๐˜: 10.1109/๐˜›๐˜Š๐˜ˆ๐˜‹.2026.3727381.


#SUTD #FCP #IMDA #NRF #IndustrialIoT #6G #NetworkOptimisation #IEEE