DescriptionWe consider spatial and temporal coherence of the vehicle–to–vehicle (V2V) wireless communication channel with focus on a suburban residential highway. The dominant reflectors in such an environment are vehicles passing in the adjacent lane and houses
along the road. Instead of treating the reflectors as point targets, the V2V short range propagation environment requires partitioning of the illuminated reflector side into
sufficiently small tiles. The channel transfer function is obtained as a superposition of specular reflections from the tiles, the line–of–sight (LOS) component, and the ground reflection. The tile size is selected to ensure that the ratio of the tile area to the tile–to–receiver distance satisfies the far field conditions. The reflected power is described by the tile radar cross section (RCS). The bistatic physical optics RCS model is adapted to account for the tile’s orientation with respect to the ray geometry. We apply the superposition model to the numerical analysis of two general scenarios
for a 22MHz channel in the 2.4GHz band. The first scenario considers a single vehicle reflector passing in the lane adjacent to the V2V communication pair. Both the vector network analyzer (VNA) experiments and the tiling model analysis illustrate that repositioning of the reflector, the transmitter, or the receiver by a few centimeters results in change of the signal power by several decibels. The second scenario analysis characterizes the channel coherence statistics for the
suburban residential highway. We consider the V2V single lane LOS and non–LOS geometries, where in the latter the receiver is shadowed by a large vehicle. The reflectors are both houses and vehicles passing in the opposite direction. The measure of channel coherence is the normalized spatial covariance calculated by correlating transfer functions corresponding to feasible receiver position pairs and performing spatial smoothing. The area of feasible receiver positions is divided into contiguous squares
whose size ensures wide sense quasi–stationarity within the square. Irrespective of direction the correlation remains high and a typical sedan roof usually does not provide sufficient spacing to obtain average inter–antenna correlation lower than 0.5. The upper bound on coherence time extends over the transmission time of multiple packets for systems in the considered band, and does not allow for usable time diversity.