29 Jun 2023
QCMP accepted to FIRA 2023
Students demonstrate Reinforcement Learning on Programmable Switches
![](/media/12833/in-network-ml.jpg?width=850&height=450&format=webp&quality=85&rnd=133386332173530000)
Changgang Zheng, DPhil Student, and Benjamin Rienecker, 4th year MEng student, demonstrate for the first time reinforcement learning on commodity programmable switch ASIC.
The work uses Q-learning deployed within the data-plane for traffic load balancing. The algorithm, called QCMP, will be presented in September in the 2nd ACM SIGCOMM Workshop on Future of Internet Routing & Addressing (FIRA).