TY - GEN
T1 - Deadline-aware Multipath Transmission for Streaming Blocks
AU - Zuo, Xutong
AU - Cui, Yong
AU - Wang, Xin
AU - Yang, Jiayu
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Interactive applications have deadline requirements, e.g. video conferencing and online gaming. Compared with a single path, which may be less stable or bandwidth insufficient, using multiple network paths simultaneously (e.g., WiFi and cellular network) can leverage the ability of multiple paths to service for the deadline. However, existing multipath schedulers usually ignore the deadline and the influence from subsequent blocks to the current scheduling decision when multiple blocks exist at the sender. In this paper, we propose DAMS, a Deadline-Aware Multipath Scheduler aiming to deliver more blocks with heterogeneous attributes before their deadlines. DAMS carefully schedules the sending order of blocks and balances its allocation on multiple paths to reduce the waste of bandwidth resources with the consideration of the block's deadline. We implement DAMS with the inspiration of MPQUIC in user space. The extensive experimental results show that DAMS brings 41%-63% performance improvement on average compared with existing multipath solutions.
AB - Interactive applications have deadline requirements, e.g. video conferencing and online gaming. Compared with a single path, which may be less stable or bandwidth insufficient, using multiple network paths simultaneously (e.g., WiFi and cellular network) can leverage the ability of multiple paths to service for the deadline. However, existing multipath schedulers usually ignore the deadline and the influence from subsequent blocks to the current scheduling decision when multiple blocks exist at the sender. In this paper, we propose DAMS, a Deadline-Aware Multipath Scheduler aiming to deliver more blocks with heterogeneous attributes before their deadlines. DAMS carefully schedules the sending order of blocks and balances its allocation on multiple paths to reduce the waste of bandwidth resources with the consideration of the block's deadline. We implement DAMS with the inspiration of MPQUIC in user space. The extensive experimental results show that DAMS brings 41%-63% performance improvement on average compared with existing multipath solutions.
UR - https://www.scopus.com/pages/publications/85133257652
U2 - 10.1109/INFOCOM48880.2022.9796942
DO - 10.1109/INFOCOM48880.2022.9796942
M3 - Conference contribution
AN - SCOPUS:85133257652
T3 - Proceedings - IEEE INFOCOM
SP - 2178
EP - 2187
BT - INFOCOM 2022 - IEEE Conference on Computer Communications
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 41st IEEE Conference on Computer Communications, INFOCOM 2022
Y2 - 2 May 2022 through 5 May 2022
ER -