Abstract
A novel constant modulus algorithm (CMA) is proposed based on a uniformly spaced linear array by constraining its weight vector to a specific conjugate symmetric form. The constraint is derived from the beamformer with an optimum output signal to interference plus noise ratio (SINR). The effect of the additional constraint is equivalent to adding a second step to the original adaptive algorithm. With this constraint, the modified CMA will always generate a weight vector in the desired form for each update and the number of variables is effectively reduced by half, leading to a faster convergence rate and an improved SINR performance. The proposed idea is general and can be extended to many other variations of the standard CMA.