An improved error probability evaluation method for digital communication systems with applications to sequential decoding

It is often necessary to use Monte Carlo (MC) simulations for evaluating digital communication system performance in terms of bit error rates. The calculations of these error rates using a brute-force MC are difficult. An improved importance sampling (IS) technique for estimating the performance of coded communication systems which employ sequential decoders is presented. This technique uses an efficient simulation scheme for reducing the required number of simulation runs. Reduction is achieved by the use of `biased´ sampling distribution, which tends to increase the relative frequency of the rare error events