Project Title

Experiment Investigation of the impact of PHY Layer Rate Control and Frequency Selection in 802.11-based Ad-Hoc Networks

Abstract

This paper presents an experimental investigation of the performance impact of two important PHY layer design options that arise in 802.11 ad-hoc networks. In particular, throughput results are provided for multi-hop ad-hoc networks with and without PHY auto-rate control and for single vs. multiple frequencies. The study is motivated by the fact that default 802.11-based ad-hoc networks using commercially preset autorate PHY and a single frequency channel suffer from performance degradations caused by link quality fluctuations and MAC layer
self-interference respectively. A baseline ad-hoc network scenario is set up on the ORBIT radio grid testbed at Rutgers and is used to
determine end-to-end multi-hop flow throughput with default rate control and single channel operation. These results are then
compared with those obtained with multiple channels and alternative PHY-rate selection methods demonstrating the potential for significant performance improvements. We observed significant improvements in end-to-end flow throughput, as much as ~4x for multiple channel vs. single channel and ~3x for optimally controlled PHY rate vs auto-rate.

Publication

Zhibin Wu, Sachin Ganu, Ivan Seskar and Dipankar Raychaudhuri, "Experimental Investigation of PHY Layer Rate Control and Frequency Selection in 802.11-based Ad-Hoc Networks", in Proceedings of ACM SIGCOMM Workshops, August, 2005. 

Discussion

Some readers raise questions about the rate control algorithms adopted by the Atheros AR5212 driver used in our research in ORBIT testbed. It needs to be pointed out that our experiment is conducted in Feb to Mar, 2005. At that time, the default Madwifi driver use a algorithm different from the SampleRate algorithm. After summer 2005, the Samplerate algorithm is used by default in released Madwifi drivers. So, the experimental results might vary and not consistent with our results in the paper.