As opposed to analog. It is a method of representing real-life signals (which are generally effectively infinitely variable) by using discrete numbers, usually binary numbers (a pattern of 1s and 0s). Holding discrete values, rather than the infinite number of intermediate levels used by analog, makes digital signals relatively resistant to distortion and noise. The reason is that if any inaccuracy creeps in, unless very severe it will not affect the signal enough to throw it off. Consider a binary system. If all data is represented as either 0 volts or 1 volt, then it doesn't matter if some interference causes the 1 volt level to be sometimes 1.1 volts, sometimes 0.9 volts. The receiving module will regard any voltage as greater than 0.5 volts as 1 volt and treat it accordingly.