A Multiplication Scheme With Variable Weights For Ensemble Monte Carlo Simulation Of Hot-Electron Tails
The use of the Monte Carlo (MC) technique for simulation of semiconductor devices is very expensive in terms of computational resources. This is mainly due to the high statistical error in the modeling of rare events. Several variance-reduction schemes have been proposed in the literature, based on the concept of splitting or repetition of trajectories, and averaging of the results.1–6 Many of the methods are only applicable to single-particle simulations, because they assume steady state and do not account for time-dependent Coulomb interaction between particles.1,2,3 Schemes suitable for ensemble simulation are basically extensions of the one-particle procedure, and require some tuning to the particular device structure being simulated.4 In this work we present an extension of the splitting/gathering scheme of Ref. 5. A flexible procedure is derived that allows the simulation of particles of different weights to obtain a balanced sampling of the phase space in a time-dependent, self-consistent framework. Results are presented for bulk, one-dimensional, and two-dimensional simulations.
Unable to display preview. Download preview PDF.
- A. Phillips, Jr. and P. J. Price, Monte Carlo calculations on hot electron energy tails, Appl. Phys. Lett. 30(10):528(1977).Google Scholar
- E. Sangiorgi, B. Ricco, and F. Venturi, MOS2: An efficient Monte Carlo Simulator for MOSdevices, IEEE Trans. Comput.-Aided Design Integrated CircuitsCAD-7(2): 259 (1988).Google Scholar
- F. Venturi, R. K. Smith, E. Sangiorgi, M. R. Pinto, and B. Riccò, A general purpose device simulator coupling Poisson and Monte Carlo transport with applications to deep submicron MOSFET’s, IEEE Trans. Comput-Aided Design Integrated CircuitsCAD-8(4): 360 (1989).Google Scholar
- D. Liebig, P. Lugli, P. Vogl, M. Claassen, and W. Harth, Tunneling and ionization phenomena in GaAs PIN diodes, Microelectronic Engineering19: 127 (1992).Google Scholar