Approximating the Probability Distribution of OFDM Symbol Errors
Given an N subcarrier orthogonal frequency division multiplexing (OFDM) system transmitting over a slow fading Rayleigh channel, the distribution of b, the number of received symbol errors, is Poisson binomial. Hence, (N b) terms are required to calculate each probability for b = 0, 1, . . ., N. When N is large, as in most OFDM systems, the Poisson binomial distribution is often approximated by the Poisson distribution. We show that, for large N, the total variation distance between the approximation and the true distribution is lower and upper bounded by random variables with fully known probability density functions. The bounds on the total variation distance indicate that the distribution of OFDM symbol errors is well approximated by a Poisson distribution.