An Improved Unwrap for Data Corrupted by Overflow

Overflow occurs when a value is represented in binary using an insufficient number of bits. Values that lie outside the bit range are misinterpreted as values within the bit range. Continuous signals that suffer from integer overflow can be correcting using unwrap functions, but this correction fails when significant noise is present. A new method of correcting signals is proposed which uses moving average filters to correct for integer overflow. The moving average is used to determine areas of active overflow. In the corrected signal, the areas of active overflow are removed, and jumps caused by overflow are corrected for. The unwrap method and the new moving average method are compared using various examples.