This thesis deals with antenna arrays when used in mobile communication environment, but the emphasis is on the reverse data link of IS-95. Two estimators are proposed: one for the direction-of-arrival (DOA) and other for the data. Initially, study of the antenna arrays is presented together with the main details of the IS-95 System, especially the details related to the code division multiple access (CDMA). After that, the problem of dirction- of-arrival estimation is discussed. It is shown that mostly of the known estimators, such as Delay-and-Sum, Capon, MUSIC and ESPRIT, do not offer reasonable performance when used in CDMA environment. Therefore, a new Maximum- Likelihood (ML) method is proposed using the so-called Expectation Maximization (EM) algorithm. The main advantage of using such approach is due to the fact that multidimensional optimization required by ML estimation can be split into several problems of lower dimension, thus simplifying the solution. The proposed estimator, besides the DOA estimation, also allows the estimation of attenuation and phase shift of the arriving signals. Simulations show that EM-estimator offers a fair performance when compared to other methods. In order to properly access the statistical significance of this study, the of DOA-estimation erro´s variance is compared with the Cramér-Rao Lower Bound (CRLB). Under a multipath environment, four versions of a data estimator are proposed, and one of them do employ trellis. In the development of such estimators, vector basis are used to represent the Walsh functions and the module-2 operations are replaced by convenient ordinary arithmetic operations. By means of simulation, it is shown that these versions present different levels of performance and computational cost thus allowing to choose the method according to the environmentn conditions.