Measurements and modelling of land-mobile satellite signal statistics

This paper describes measurement on an 870 MHz signal having variations caused by multipath fading and shadowing, typical of a land mobile satellite link in a rural environment. Spectral densities, envelope and phase distributions are computed from these measurements. Preliminary results show that the probability densities of the phase of the signal and its rate of change (random FM) can be modelled by Gaussian random processes, An application of the random FM model shows that random FM has negligible effect on the probability of error of an MSK signal at 2400 bits per second.