Signal-to-Noise and Idle Channel Performance of Differential Pulse Code Modulation Systems - Particular Applications to Voice Signals

Digital techniques for transmitting analog signals such as voice, television, or facsimile have been known for a long time, and technology has reached the point where some of these methods are commercially feasible. Since cost is a critical factor in determining applicability of these systems, there has been from the beginning an attempt to improve the efficiency of analog-to-digital conversion by reducing the bit rate required for a given accuracy of reproduction. One of the principal methods involves removing inherent signal redundancy through the use 1123 1124 T H E BELL SYSTEM TECHNICAL J O U R N A L , S E P T E M B E R 1966 of feedback around the quantizer, and has led to a wide variety of schemes which may all be classed as differential systems. The origins of differential pulse code modulation (DPCM) stem from patents by the N. V. Phillips Company in 19511 and by C. C. Cutler in 1952.2 The ideas also appear in several papers of about that time.3 4-5 Since that time, considerable research and development work has been reported, and one has only to look at our reference list, which is certainly not complete, to be convinced that the problems have been examined at great length. The work to be reported here is the result of a fairly extensive investigation of the potential advantages and pitfalls of voice transmission by practical DPCM systems and by alternatives which are essentially variations on the basic theme of PCM or DPCM. The problems are handled analytically as far as is possible.