Using additive white binary noise to marginalize read channels

Optimization of bit error rate (BER) is typically not a part of disk drive burn-in, due to the large amount of time required to measure bit error rates accurately. In this paper, theoretical calculations, numerical simulations, and experimental results are used to show that additive white binary noise can be used to marginalize the performance of a read channel. As measurements of the larger BERs due to marginalization are much faster than normal BER measurements, this technique makes practical the use of BER based optimization during drive burn-in.