Monte Carlo Simulation with Error Classification for QAM Modulation under Rayleigh Fading Channel

Evaluation of bit error rate (BER) of digital communication systems is usually done via simulation using Monte Carlo (MC) method. For low BER, MC method requires huge sample sizes to achieve certain efficiency. To overcome this limitation, many variance reduction techniques such as importance sampling (IS) have been proposed. In this paper, a novel simulation method Monte Carlo simulation with error classification (EC-MC) is discussed. This method can reduce the estimation variance through dividing the total errors into many sub-categories and optimizing the simulation sample size for each sub-category. We apply this method for simulations of QAM modulation under Rayleigh fading channel. The simulation results demonstrate EC-MC method can achieve the same accuracy at smaller sample sizes and shorter simulation runtime, comparing with both conventional MC and IS methods, especially at high signal to noise ratios.