Cooperative communications have been gaining huge attention lately, due to the increase in the number of mobile devices and the advancement in their capabilities. In this framework, a commonly suggested approach to benefit from cooperation is the formation of virtual groups of mobile terminals. Mobility-aware algorithms are commonly proposed, where nodes form cooperative clusters according their basic mobile characteristics. However, more detailed mobility pattern information is not used to improve the performance of these systems. The reason is the complexity of current mobility pattern recognition algorithms that require previous set up and advanced knowledge on the scenario. This paper presents a novel cognitive algorithm for the identification of mobility patterns that can be used for mobile group formation and that requires no configuration. The proposed algorithm, based on data mining techniques and cooperative optimization, can evaluate the number of mobility patterns in the scenario and classify nodes according these patterns in a distributive fashion. The algorithm is mathematically presented and tested with extended simulations in both Matlab and ns-2.