REVeno: RTT Estimation Based Multipath TCP in 5G Multi-RAT Networks
Jaewook Jung, Changsung Lee, Jungsuk Baik, Jong‐Moon Chung · IEEE Transactions on Mobile Computing · 2022
5G networks were designed to provide sufficient throughput and reliable data communication services. To achieve the targeted QoS requirements of the 5G specifications, the use of mmWaves are essential. However, mmWavs are very vulnerable to signal blockages due their high frequency. Multi-path transmission control protocol (MPTCP) can deal with this problem by transmitting data through multiple subflows using TCP. The current congestion control scheme of MPTCP was designed for wired networks and is not suitable for use in wireless networks. In this paper, a new MPTCP congestion control algorithm named round trip time (RTT) estimation based Veno (REVeno) is proposed to better support wireless networks by distinguishing between loss due to congestion and wireless channel errors. The proposed REVeno scheme uses a novel backlog estimation formula that considers the total buffer size and parameters that distinguish loss due to congestion and loss due to the wireless channel. The REVeno scheme will use these estimations to minimizes the reordering delay by equalizing the equilibrium RTT of all subflows. The simulation results show that the proposed REVeno scheme performs better than the existing schemes in terms of goodput and latency without adversely affecting other concurrent TCP connections in the same network.