Abstract: From the point of view of blind signal separation,this paper introduces briefly analytical constant modulus algorithm(ACMA).By means of analytical method,it computes the complete set of beamformers all at once,which is its significant difference from other iterative CMAs that search for the extremum of cost function and by means of iterative methods,to recover the constant modulus signals one by one.This feature presents many other advantages.What's important is that it can recover all CM signals without the influence of signal power difference.In this paper,with theoretical demonstration and computer simulation,we validate the asymptotic properties of ACMA.In Gaussian noise,it converges to the Wiener beamformer when the number of samples increases to infinity.Meanwhile,we utilize practical field data to test and analyse its performance of signal separation,and validate its perfect ability of signal separation.