For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Detection of Traffic Congestion Based on Link-Level Metrics in IEEE 802.11 Networks
Jong-Ok KIM Peter DAVIS Tetsuro UEDA Sadao OBANA
IEICE TRANSACTIONS on Communications
Publication Date: 2011/04/01
Online ISSN: 1745-1345
Print ISSN: 0916-8516
Type of Manuscript: PAPER
Category: Terrestrial Wireless Communication/Broadcasting Technologies
congestion metric, link-level measurement, heterogeneous wireless network, transmission link switching, IEEE 802.11,
Full Text: PDF>>
In this paper, we address adaptive link switching over heterogeneous wireless access networks including IEEE 802.11. When an IEEE 802.11 link is congested, the transmission link of a terminal with multi-RATs (radio access technologies) is switched to another radio access systems. To this end, we propose link-level metrics of LC (link cost) and AC (access cost) for quantifying TCP congestion over IEEE 802.11 networks. The proposed metric can be easily measured at a local wireless terminal, and that enables each multi-RAT terminal to work in a distributed way. Through various indoor and outdoor experiments using a test-bed system, we verify that the proposed link level metrics are good indicators of TCP traffic congestion. Experimental results show that the proposed metrics can detect congestion occurrence quickly, and avoid the TCP throughput degradation of other neighboring terminals, when they are used for transmission link switching.