Increasing data demand started to become a common challenge in mobile cellular networks. Various techniques, intelligent approaches and network architectures are being developed to tackle this difficulty, in order to provide high quality of service and to allocate mobile resources effectively. Adaptive modulation and coding is one of those techniques, that provides high quality of service while adapting modulation and transmission power with respect to erratic channel conditions. In this work, sufficient modulation and coding technique will be proposed to provide the necessary quality of service level while minimizing mobiles' consumed power by foreseeing actual data need of UMTS mobile users through digital divide analyses. As a consequence, with prior information on potential data requirements, it is shown here that power and energy efficiency can be improved by making use of convex optimization methods.