Bias correction for the estimation of discretized diffusion processes from an approximated likelihood function

We consider the parameter estimation of a discretized diffusion process based on an approximation of the likelihood function. We suppose the observation step fixed and we propose to correct the bias of such estimators by using simulations of the process.