Energy Efficient Congestion Control for Multipath TCP in Heterogeneous Networks

Abstract

In this paper, we propose a receiver energy efficient congestion control algorithm based on multipath transmission control protocol (MPTCP) to enable the battery powered mobile devices receive more data than MPTCP with the same energy consumption. First, based on the receiver energy consumption model of wireless interface, a constrained optimization problem to maximize energy efficiency is formed. With genetic algorithm, a rate distribution vector is obtained as a near-optimal solution. Second, we adjust the congestion windows based on the acquired vector to schedule packets over each path directly. Jointly considering energy efficiency, round trip time and path loss rate, a novel congestion control algorithm is proposed to adjust the increment of congestion window when an acknowledgement is received in the congestion avoidance phase. The energy efficiency term is obtained by extending the energy efficient rate distribution vector into the window-based congestion control algorithm via fluid model. The simulation results demonstrate that the proposed algorithm shifts part of traffic from the higher energy consumption paths to the lower ones. It improves throughput greatly and achieves higher energy efficiency, almost twice the size of MPTCP.