The 5-GHz Airport Surface Area Channel—Part I: Measurement and Modeling Results for Large Airports

We describe results from a channel measurement and modeling campaign for the airport surface environment in the 5-GHz band. Using a 50-MHz bandwidth test signal, thousands of power delay profiles (PDPs) were obtained and processed to develop empirical tapped-delay line statistical channel models for large airports. A log-distance path loss model was also developed. The large airport surface channel is classified into three propagation regions, and models are presented for each of the regions for two values of bandwidth. Values of the median root-mean-square (RMS) delay spread range from 500 to 1000 ns for these airports, with the 90th percentile RMS delay spreads being approximately 1.7 ms. Corresponding correlation bandwidths (i.e., correlation value 1/2) range from approximately 1.5 MHz in non-line-of-sight (NLOS) settings to 17.5 MHz in line-of-sight (LOS) settings. Two types of statistical nonstationarity were also observed: 1) multipath component persistence and 2) propagation region transitions. We provide the multipath component probability of occurrence models and describe Markov chains that are used for modeling both phenomena. Channel tap amplitude statistics are also provided, using the flexible Weibull probability density function (pdf). This pdf was found to best fit fading tap amplitude data, particularly for frequently observed severe fading, which is characterized by fade probabilities that are worse than the commonly used Rayleigh model. Fading parameters equivalent to Nakagami-$m$-model values of $m$ near 0.7 were often observed (with $m = 1$ being Rayleigh and $m ≪ 1$ being worse than Rayleigh). We also provide channel tap amplitude correlation coefficients, which typically range from 0.1 to 0.4 but occasionally take values greater than 0.7.