In this paper, the generalized Bayes estimator of mean vector parameter for multivariate normal distribution with Unknown mean vector and covariance matrix is considered. This estimation is performed under the balanced-LINEX error loss function. The generalized Bayes estimator by using wavelet transformation is investigated. We also prove admissibility and minimaxity of shrinkage estimator and we present the simulation study and real data set for test validity of new estimator.