Error statistics of magnetic recording in severe environments

Magnetic recording systems are frequently used in less than ideal environments. Unfortunately, all recording systems are prone to a variety of causative error mechanisms. Even in controlled environments digital recording systems are liable to error, due, for example, to additive noise or intersymbol interference. In stressed recording environments these mechanisms can be compounded by such environmental effects as shock, vibration, temperature variation and changing tape speed. However, with an understanding of system error statistics relating to these mechanisms, sufficient error detection and correction techniques can be employed to prevent errors in the recording process causing errors in the reproduced data. This paper concentrates on a real industrial application of digital instrumentation recording, where the equipment may be subjected to a severe environment. The research described in this paper involved off-line statistical experiments in which the effects of a simulated real-world environment, on a digital recording instrumentation system, were investigated. The paper presents various aspects of the results from this statistical analysis, in particular, the effects of dropouts are investigated. It is shown that a common cause of error is the `bit-slip´ or `synchronization-loss´ event, where bits are either lost or inserted into the data stream, such that bit-errors propagate throughout the remainder of the afflicted data block