Modulation and signal selection for digital data systems

The statistical decision model of a communication channel is used to discuss a variety of signal modulation techniques intended to overcome the problems met in the use of voice-bandwidth channels for high-performance digital data transmission. The importance of proper transmission of timing and word-start information is emphasized. The equivalence of various modulation schemes to selection from an alphabet of signal waveforms is pointed out; also that, in the presence of additive white Gaussian noise, the performance of such simple-waveform systems as AM (amplitude modulation), phase reversal, and FSK (frequency shift keying) is the same to within a few db (decibels). Implications of the fact that the channel noise is not Gaussian, but impulsive in character, are briefly explored.