As 5G and future cellular networks evolve to support critical real-time applications like telesurgery, remote robotics, and autonomous driving, balancing extreme reliability with ultra-low latency becomes paramount. However, traditional packet recovery methods often introduce delays that are unacceptable for the Tactile Internet.

A new collaborative study, published in IEEE, introduces Flexible Network Coding (FlexNC) and implements a Recoder in the Network (RecNet) to solve this fundamental bottleneck within cloud-native 5G environments.

The paper, titled “Flexible Network Coding and Recoding in Cloud-Native 5G Environments” [or insert exact paper title if different], was co-authored by Osel Lhamo, Tung V. Doan, Elif Tasdemir, Mahdi Attawna, Giang T. Nguyen, and Patrick Seeling—representing researchers from the ComNets chair at TU Dresden and key initiatives like CeTI and 6G-life.

The Challenge: The Latency Penalty of Lost Data

In standard communication networks, if a packet of data is lost due to wireless interference, the system must request a retransmission. While this ensures the data eventually arrives (high reliability), it introduces a massive delay (high latency). For human-in-the-loop applications—where haptic feedback and video streams must synchronize in real-time under tight milliseconds limits—traditional retransmission is simply too slow and can disrupt the physical control loop.

Research Highlights: In-Network Computing to the Rescue

To bypass these traditional delays, the research team designed and implemented a proactive, low-latency framework:

  • In-Network Processing (RecNet): Instead of waiting for a distant server to retransmit lost data, the team deployed a “Recoder in the Network” (RecNet). By leveraging modern in-network computing, network switches can instantly reconstruct missing packets on the fly.

  • Flexible Network Coding (FlexNC): This protocol dynamically adapts how data packets are grouped and coded, allowing the network to recover lost packets seamlessly under noisy or unstable wireless conditions without requesting retransmissions.

  • Real-World Validation: Rather than relying solely on simulations, the authors validated their framework on a real-world, cloud-native 5G testbed. They utilized the open-source platform OpenAirInterface combined with commercial programmable switching hardware to prove the system’s real-time feasibility.

The Impact for Industry 4.0 and Digital Health

Whether it is a surgeon operating on a patient miles away or an engineer controlling cooperative robots on a factory floor, reliable and instant communication is non-negotiable. This research demonstrates how the integration of flexible network coding and in-network computing can significantly boost reliability for remote users while maintaining the strict timing requirements of tactile and visual traffic. This work lays a vital foundation for resilient, global control loops in the next generation of mobile communication.

The study is published in the IEEE library. Dive into the technical details and testbed results here.

Authors: Osel Lhamo, Tung V. Doan, Elif Tasdemir, Mahdi Attawna, Giang T. Nguyen, and Patrick Seeling

This work is a collaboration within the ComNets chair, the Cluster of Excellence CeTI, and the 6G-life hub at TU Dresden.