Analysis of an efficient real-time generation of ES/N0 estimates for M-PSK signals

In Y. Linn [2002] a new method of estimating the E_S/N₀ ratio for M-PSK receivers was presented, which derives that estimate from the locked-state value of a new family of carrier lock detector structures. That method operates on the received symbols and generates an E_S/N₀ estimate at the output of the M-PSK demodulator, before error correction decoding or symbol decisions are made. The prevalent contemporary methods of E_S/N₀ estimation for M-PSK rely on measuring the pre- or post-decoder symbol error rate (SER) or bit error rate (BER) of the received signal, and estimating the E_S/N₀ from the relations between the SER or BER and the E_S/N₀ ratio. In this paper, a quantitative analysis shall be undertaken in which it will be shown that estimation of the E_S/N₀ using the method presented in Y. Linn [2002] not only requires significantly less hardware and/or software resources than estimation from the received signal´s error rate, but is also irrespective of the data sequence and necessitates far fewer symbols in order to arrive at an equally accurate estimate.