In this paper, an efficient bit-loading algorithm in the context of SVD-aided MIMO transmission for vehicle-to-infrastructure (V2I) communications is presented. This algorithm, which exploits the intrinsic properties of the eigen-channels resulting from SVD, tries to minimize the error probability of the system, assuming that the error rates per subchannel are equal, without considering the well-known Gap approximation. Furthermore, we examine the effects of increasing the number of antennas and spatial correlation when using a bit loading policy. Several simulation results for different transmission scenarios are presented and discussed. We show that when using a high number of antennas, bit and power allocation scheme is reduced to an adaptive power allocation scheme only which is due to the respect for the hypothesis on error rate, and for a reduced number of antennas, the effect of correlation becomes more pronounced while the allocation of ressources becomes sensitive to the variation of the number of antennas