Mobile innovative services are being introduced continuously to enhance people's life through facilitating the interaction between human and the rest of the world. The world of “Internet of Things” (IoT) is expanding everyday to include more things to be connected. Considering that more and more innovative mobile services are being introduced, the long-term evolution (LTE) telecom systems/networks will be more complicated, require more resources, and demand more challenging requirements. The LTE evolved packet core (EPC) network internal design is inadequate with regard to the resources reservation techniques used to carry out the guaranteed dedicated services. In fact, EPC does not have the capabilities to utilize properly the unused bandwidth of the guaranteed bearer when the reserved bandwidth is not fully used by the mobile service, the unused guaranteed bandwidth is considered as wasted resources and consequently the whole LTE/EPC network efficiency gets affected. In this paper, we propose an adaptive technique which enhances the resource reservation for the LTE Mobile guaranteed services, our solution provides techniques to: analyze the ongoing mobile guaranteed traffic usage, provide time-series models that mathematically represent the conducted data, forecast the mobile service guaranteed resource consumption, identify the wasted/unused resources, and utilize these resources by other services. Our experiments were conducted on a dataset captured on an LTE network, the experimental results show that our approach is feasible and beneficial as it enhances the resource allocation for the LTE mobile services and increases the overall throughput of the LTE/EPC networks.