Experimental modeling of the effect of adjacent lane traffic on the vehicular channel

Abstract

Our experimental study measures the effects of a stationary and a moving vehicle on the 22 MHz 802.11b and the 20 MHz 802.11a channel using a vector network analyzer (VNA). The experiments with the vehicle slowly passing by a transmitter-receiver pair correspond to a case study where the communication session occurs in a single lane of a multi-lane highway with a vehicle moving in the adjacent lane. We propose a corresponding ray-tracing model based on the bistatic radar equation which predicts the link power for a given car geometry and the position of transmit and receive antennas. The model converts a near field propagation problem to a superposition of a set of far field sub-problems by representing the vehicle as a set of (sufficiently small) ideally conductive flat tiles. Hence, the channel transfer function is determined as a sum of the line-of-sight (LOS) component and the rays reflected from the tiles. The ray strengths are a function of the effective tile radar cross sections (RCSs). The carefully selected RCS model allows for a good match between the measurements and the resulting ray-tracing model. Both the model and the measurements illustrate that the change in the propagation geometry on the order of centimeters, created by a car passing in the proximity (on the order of meters), results in the change of the signal power at the receiver on the order of several decibels.