Energy Consumption of TCP Reno, Newreno, and SACK in Multi-Hop Wireless Networks

Authors: Harkirat Singh, Suresh Singh

Complete Citation

Harkirat Singh, Suresh Singh, Energy Consumption of TCP Reno, Newreno, and SACK in Multi-Hop Wireless Networks, ACM SIGMETRICS 2002. DOI Bookmark: http://doi.acm.org/10.1145/511334.511360

Abstract

In this paper we compare the energy consumption behavior of three versions of TCP – Reno, Newreno, and SACK. The experiments were performed on a wireless testbed where we measured the energy consumed at the sender node. Our results indicate that, in most cases, using total energy consumed as the metric, SACK outperforms Newreno and Reno while Newreno performs better than Reno. The experiments emulated a large set of network conditions including variable round trip times, random loss, bursty loss, and packet reordering. We also estimated the idealized energy for each of the three implementations (i.e., we subtract out the energy consumed when the sender is idle) and here, surprisingly, we find that in many instances SACK performs poorly compared to the other two implementations. We conclude that if the mobile device has a very low idle power consumption then SACK is not the best implementation to use for bursty or random loss. On the other hand, if the idle power consumption is significant, then SACK is the best choice since it has the lowest overall energy consumption.

Annotations

This paper talked how to measure the energy consumption of TCPs in wireless networks. Then, it compares the energy consumption of three versions of TCP - Reno, New Reno, and SACK.

Experimental setup: To measure the energy consumed, the sender's power supply is connected to a multimeter controlled by a separate notebook.

Results:

(1) The total energy consumed is inversely proportional to the throughput.

(2) SACK consumes the lowest energy in most cases except when loss rate is high. Reno has the highest energy consumption.

(3) For idealized energy (energy consumed during idle periods is not counted), SACK performs poorly compare to other two implementations.

-- YingxinJiang - 10 Jul 2008