The main challenge for airborne radars in discovering ground targets is the removal of Clutter and Jammer. The adaptive space-time processing method is the newest method of processing against these signals. The space-time processor creates a two-dimensional filter, which sets zero (null) angular and Doppler frequencies where Jammer and clutter are signaled. In computing the optimal weight vector for this filter, the covariance matrix of the interference signals is required. In practice, the covariance matrix of the signal is not interfering and should be estimated. The estimation of covariance matrix is based on recursive signals and according to other target cells other than the target cell (secondary data). The main problem in estimating this matrix is the limitation of the number of secondary data that is homogeneous and reliable. Different processing methods have been devised to increase convergence and reduce computational volume. The use of de finite filters facilitates the convergence and faster updating of the weighted vector and eliminates the processes caused by the inversion of the covariance matrix. In this paper, the recursive method has several parallel processing branches associated with weight reduction and drop-down matrix updating approach which, in comparison with previous recursive methods, results in improved performance by having a limited number of secondary data.