In this paper, the concept of cooperative diversity is introduced in future wireless networks. An important question for cooperative wireless networks is how cooperative terminals are assigned and managed in a multi-terminal network. In other words, how is it determined which users cooperate with each other, and how often are relaying terminals reassigned. We propose a discrete stochastic approximation algorithm to adaptively select a better cooperative subset using criteria such as minimum error rate. The proposed algorithm is significantly more computationally efficient. Through Monte-Carlo simulations we verify both our analytical results and the performance of the algorithms. We also present numerical examples to show the convergence of the algorithm