Wireless Link Bandwidth Estimation from Signal-to-Noise-Ratio

Dynamics of link bandwidth of a wireless link, which changes frequently and abruptly due to the dynamic channel sharing, fading, and mobility, is of interest to adaptive applications and protocols in wireless communication environments. This paper presents a novel approach to model wireless link bandwidth based on radio signal-to-noise ratio (SNR). Its salient feature is that unlike the traditional methods that use probe packets, our method is non-intrusive to the wireless network since wireless device drivers normally provide SNR information. The method is evaluated on both indoor and outdoor IEEE 802.11b wireless networks, and the estimation error for a model using BP-neural-network is within 10% to 15% range.