Convergence analysis of the CMA blind equalizer

The convergence behaviour of the constant modulus algorithm (CMA) is analytically evaluated in terms of the mean and the variance of the equalized sequence conditioned on the transmitted symbol, thus characterizing the dispersion pattern of the equalized samples around the QAM constellation points. This extends to the CMA and improves the approach of Weerackody et al. (see ibid., vol.40, no.6, p.1383, 1992) , where the MSE trajectory of the generalized Sato algorithm (GSA) was analysed.<>