3.2 WAVEFORM CODING TECHNIQUES

Waveform coding itself consists of two fundamental operations. The first is sampling, in which the analog signal is measured at regular intervals in time. Sampling converts a continuous-time waveform into a sequence of discrete-time samples while preserving all of the information required for reconstruction, provided the sampling rate satisfies the Nyquist criterion introduced in Chapter 2. The second operation is quantization, in which each sample is assigned to one of a finite number of amplitude levels. The resulting sequence of quantized values can then be represented by binary numbers and transmitted as a digital bitstream.

The various waveform coding techniques differ primarily in how they perform these operations and how efficiently they represent the sampled signal. Some methods, such as pulse code modulation (PCM), encode the absolute value of every sample. Others, including differential PCM (DPCM) and adaptive differential PCM (ADPCM), exploit the fact that successive samples are often very similar, transmitting only the changes between them. Delta modulation (DM) goes one step further by representing only whether the signal has increased or decreased since the previous sample. Each technique represents a different compromise between transmission bit rate, implementation complexity, bandwidth, and reconstructed signal quality.

In the sections that follow we begin by examining the fundamental processes of sampling and quantization before introducing the principal pulse modulation techniques. We then study PCM in detail, including quantization noise and companding, before considering differential coding methods and delta modulation. Together, these techniques form the foundation of digital waveform coding and provide the basis for many of the speech, audio, and multimedia coding systems used in modern communications networks.