On the Performance of Phase-Shift-Keying Systems

Coherent phase-shift keying (CPSK) and differential phase-shift keying (DPSK) are two techniques often used in digital communications over channels such as satellite, terrestrial radio, and voiceband telephone. The literature abounds in analyses of their performance under a variety of conditions. A sample collection of some of this literature may be found in Ref. 1. The chief reasons for the widespread use of these techniques are simplicity of implementation, superior performance over the additive Gaussian noise channel, minimal bandwidth occupancy, and minimal envelope variation. The relative performance of CPSK and DPSK systems is well understood only in the presence of additive Gaussian noise. In this case, the detection efficiency of DPSK is known to be about 1 dB (in s/n) below that of CPSK for binary modulation and this degradation approaches 3 dB for multilevel systems. In applications where a 3-dB loss in s/n is important, such as in down-link satellite, space communications, and terrestrial radio under deep fading conditions, CPSK is the preferred method. In CPSK, however, the generation and extraction of a local carrier-phase reference at the receiver is required. A coherent phase estimate is usually obtained by using phase-locked loop (PLL) techniques, and because of frequency instabilities and phase jitter inherent in transmitter and receiver systems, carrier recovery loop bandwidths cannot be made arbitrarily small. Consequently, in practice a noisy phase estimate is obtained and only partial coherent reception can be claimed.