A study on usefulness of neural network equalizer to reduce impact of amplitude fluctuations in magnetic tape

Abstract As digital data grows explosively year after year, magnetic tape devices, with their high reliability and long-term storage capacity, are widely used for backup and archiving. Magnetic tape devices with larger capacity are required, and it is necessary to improve track density. We have previously constructed a simulator that considers inter-track interference (ITI) due to narrower tracks in barium ferrite (BaFe) magnetic tape drives and evaluated its performance. In this study, we develop a simulator which more accurately models the experimental data, considering amplitude fluctuations caused by the problem of poor tape running, and discuss the usefulness of a neural network equalizer to reduce the effects of amplitude fluctuations. As a result, we show that applying the neural network equalizer is useful in reducing the impact of amplitude fluctuations caused by the problem of poor tape running.