Sampling Signals Part 2 (9/10) – Pulse Coded Modulation

Another way of embedding sample value information into a pulse stream is via Pulse Coded Modulation (PCM). This modulation scheme requires sample values to be quantized to discrete levels. Each discrete level is assigned a bit label. These bit labels are then transmitted as positive and negative pulses. By observing the amplitude of the pulses, one can determine the quantization level and thus the original sample value. The number of quantization levels is important. While few levels require little memory to store or transmit information, the quantization error introduced would be large and reconstruction of the original continuous-time signal would yield a signal that has errors relative to the original continuous-time signal. Increasing the number of levels reduces the quantization error and improves the similarity between the reconstructed signal and the original signal at the cost of requiring more bits-per-sample to encode the information.