DeepQuic
Long Zhang, Bo He, Jingyu Wang, Qi Qi, Haifeng Sun, Jianxin Liao · 2022
In the complex 5G network environment, the number of multiple access devices increases sharply, which requires a more efficient multi-path scheduling algorithm. However, traditional multi-path heuristic algorithms can only work in static scenarios without time-varying path conditions due to their poor robustness. In this paper, we propose a learning-based multi-path scheduling algorithm, DeepQuic, to determine the multi-path scheduling for dynamic and heterogeneous 5G networks. DeepQuic works upon the multi-path QUIC (MPQUIC) protocol and focuses on increasing the transmission goodput of signaling in the 5G core network.