Bandwidth choice of log-periodogram estimators with short-term noise in long memory series

This paper makes research on two log-periodogram estimators of long memory and takes importance to the short-term noise effects on the bandwidth choice. It turned out that under finite samples, both the short-term and long-term impact on the performance of the log-periodogram estimators. Experiments found that it caused difficulty in choosing an optimum bandwidth with the presence of short-term noise, however, the minimum RMSE criteria is a relatively acceptable choice.