Coverage probability estimation in single frequency networks in presence of correlated useful and interfering components

In OFDM-based single frequency networks (SFN) used in digital audio broadcasting (DAB) and digital video broadcasting (DVB) the received signal can be a superposition of multiple useful and interfering lognormal components. In many cases considerable correlations exist among these components. These correlations are typically neglected in the system analysis. In this paper we show that in certain signal constellations the estimation error can be considerable. We also provide a simple framework to estimate the correlation coefficient between the total useful and total interfering power, which is used as a correction factor in estimating the mean and standard deviation of the SIR, modeled by a lognormal distribution. Using Monte Carlo simulation we show that our method yields good accuracy in the estimation of the coverage probabilities in single-frequency networks