DPSAF: Forward Prediction Based Dynamic Packet Scheduling and Adjusting with Feedback for Multipath TCP in Lossy Heterogeneous Networks
Kaiping Xue,Jiangping Han,Dan Ni,Wenjia Wei,Ying Cai,Qing Xu,Peilin Hong
DOI: https://doi.org/10.1109/tvt.2017.2753398
IF: 6.8
2017-01-01
IEEE Transactions on Vehicular Technology
Abstract:As multihomed terminals are equipped with multiple interfaces and allowed to access heterogeneous networks, transferring data simultaneously through all the available paths becomes possible and also brings many benefits. Multipath TCP (MPTCP) has been proposed to distribute an application stream over different TCP connections. However, due to the disparate latencies of different paths, the problem of existing out-of-order packets usually occurs at the receiver. Large number of these packets exhaust the limited receiving buffer and make the receive window be stalled, which greatly degrade the throughput. Thus, an efficient scheduling mechanism will play an important role to keep in-order delivery. However, almost all of the previous intelligent scheduling mechanisms ignored packet losses, and didn't consider window changes of the congestion control algorithm and utilize the feedback information, which cannot perform well in the lossy heterogeneous networks. In this paper we propose a new scheduling algorithm: Forward Prediction based Dynamic Packet Scheduling and Adjusting with Feedback (DPSAF). DPSAF first utilizes maximum likelihood estimation in TCP modeling to estimate the data amount sent on other paths simultaneously, which takes packet loss rate and time offset into consideration, then gets feedback information from SACK options and fixes the scheduling value. From the simulation, we can see that our mechanism obviously improves throughput and reduces cache occupancy at receiver in lossy heterogeneous networks.