Impact of localization errors on wireless channel prediction in mobile robotic networks

In this paper we are interested in understanding and mathematically characterizing the impact of localization errors on wireless channel prediction in mobile robotic networks. We take the increase of the channel prediction error variance due to localization errors as our metric and analyze it in different communication environments. In particular, we show that this metric is monotonically increasing with respect to the path loss exponent and/or the standard deviation of shadowing, and is monotonically decreasing with respect to the standard deviation of multipath. Moreover, this metric is monotonically increasing with respect to the decorrelation distance of shadowing when it (the decorrelation distance of shadowing) is small, and is monotonically decreasing with respect to it when it is large. We also briefly discuss how the localization errors in the positioning of the channel a priori samples affect the estimation of the channel underlying parameters. Finally, we verify our mathematical analysis in a simulation environment.