ASYMPTOTIC PERFORMANCE ANALYSIS FOR SIGNATURE WAVE-FORM ESTIMATION INSYNCHRONOUS CDMA SYSTEMS

In this correspondence, we present an asymptotic performance analysis of a subspace-based method proposed by Liu and Xu for blind signature waveform estimation in synchronous code-division multiple-access syste ms, In particular, we derive asymptotic, i.e., as the number of sample s becomes large, expressions for the covariance of the estimated chann el parameters. We also derive aln algorithm independent bound on the c ovariance of the estimated parameters. This bound can be used as a mea sure against the theoretically predicted algorithmic performance. Some insight into the achievable performance of this algorithm is obtained by numerical evaluation of the bound for two cases of interest, and t he results are compared to that obtained by numerical evaluation of th e theoretically predicted performance. Monte Carlo simulations are use d to verify the theoretical analysis.