AbstractThis paper presents an intensive study on the fading statistics of massive multiple‐input–multiple‐output vehicular communication channels. A geometry‐based channel model is proposed, where both ends of the link are considered to be equipped with large‐scale three‐dimensional antenna arrays, whereas the scattering volumes in their vicinity are modeled as flexible (ie, shiftable, scalable, and rotatable) along all the dimensions to adapt various dynamic vehicle‐to‐everything radio propagation scenarios. Both ends of the link are modeled as flexible to be static or mobile in different directions and with independent velocities, to adapt both the vehicle‐to‐vehicle and vehicle‐to‐infrastructure scenarios. Joint and marginal three‐dimensional angle‐of‐arrival and angle‐of‐departure statistics of the channel are studied. The analysis is further extended from plain angle‐of‐arrival/angle‐of‐departure statistics to the characterization of dispersion of energy in the angular domain by using various notable angular spread quantifiers. A correlation between the received signals on two different antenna elements of the array, at two different time instants, on two distinct frequencies, and at two different spatial positions along the node mobility route is thoroughly investigated. This analysis is further extended to analytically study the Doppler spectrum characteristics and second‐order fading statistics of the channel. The considered second‐order fading statistics quantifiers are fading rate variance, level crossing rate, average fade duration, and coherence distance. The impact of various physical (eg, link distance, direction, and speed of node mobility), geometric (eg, shape and orientation of scattering volume), and antenna (eg, beam pattern, beam width, and beam switching) parameters on the angular spread, received signal correlations, Doppler spectrum characteristics, and second‐order fading statistics of the channel is thoroughly investigated. The conducted analysis is highly significant for devising efficient interference mitigation methods, error correction codes, modulation schemes, antenna beams, and receiver designs for emerging massive multiple‐input–multiple‐output vehicle‐to‐vehicle communication links, which, in turn, serves the purpose of efficient spectrum utilization.