
84 Simulation Technologies in Networking and Communications
3.4.2 progressive ase algoriThm
ASE simulation is approximately SE for moderate sizes of N. However, from (3.45), it is clear that
ASE simulation will eventually lose the 1/N
2
convergence rate as long as ϵ ≠ 0.
In [12], Biglieri suggested computing
’s iteratively to improve the accuracy of the estimation.
As opposed to the original ASE algorithm, which has xed accuracy for the entire simulation, we
proposed a PASE algorithm that keeps improving the accuracy of
’s as the chaotic MC simula-
tion goes on. The idea is to use the samples B(x
i
) generated in the main chaotic MC simulation ...