Browse Prior Art Database

Processing Oversampled Data in the form of a Serial Bit Stream Disclosure Number: IPCOM000201826D
Publication Date: 2010-Nov-24
Document File: 1 page(s) / 412K

Publishing Venue

The Prior Art Database

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 100% of the total text.


Processing Oversampled Data in the form of a Serial Bit Stream

When transmitting data using “noisy” mechanisms, like radio, it may become necessary to remove, ignore or correct “noise” found in the data.  One technique for “correcting” data is oversampling and voting.  The incoming data is samples at a rate that is a multiple of the transmission rate.  The oversamples are then “voted” to remove any “noise”.

When large data stream are transmitted without a synchronizing clock, drift can occur in the data stream as the data advances or retards in time between the transmitter and the receiver.  Correcting drift (synchronizing the clocks) is desirable to prevent alignment errors from causing improper voting of oversamples.

In some cases the oversampled data may result in sample sizes that cannot be represented (and therefore analyzed) using data types that are native (natural) to the processor.  That is, data may oversample 13 times, this “oversample” does not evenly fit into an 8, 16, 24 or 32 bit values that are natively available in most processors.

This method allows for oversampled data to be represented in a platform neutral way and analyzed on most processors.  The method also allows voting and drift correction for data being processed “on the fly”.