Variance Reduction Techniques for Monte Carlo Simulation of Coincidence Circuits

The Monte Carlo simulation is widely used to choose the design of associated particle imaging devices or space radiation detectors containing coincidence circuits. Variance reduction techniques can significantly speed up Monte Carlo simulation. In present work, a variance reduction technique is proposed for estimating coincidence circuit responses using Monte Carlo simulation, accounting for false coincidences between particles of different branching trajectories. The importance sampling, Russian roulette, and splitting techniques are considered. The proposed estimate has a form of the sum of detector responses over all possible combinations of trajectory alternatives. The statistical weight of the combination is calculated as the product of trajectory weights. The estimate unbiasedness is proved. If the coincidence circuit is not heavily loaded, the proposed estimate can be simplified to reduce the number of combinations.