One of the main attributes of the heterogeneous cellular network is the composition of different cell sizes. From such heterogeneity, expanded network capabilities had sprouted with extensive computing power consumption and ultra-low latency constraints. Using a heterogeneous network will provide multiple paths in which the users' data can flow through the network depending on the users' available resources, remaining energy, etc. In this paper, we study the heterogeneous network model, which contains MeNB, SeNB, and Femtocells, and we propose a matching subcarrier resource allocation and offloading decision (MSRAOD) algorithm. MSRAOD aims at recourse allocation optimization to minimize the total energy consumption of mobile users' devices with acceptable latency requirements of the applications. We have evaluated MSRAOD through simulation, and when compared to non-optimized data offloading, the MSRAOD algorithm significantly enhances the average energy consumption of mobile users. Such results provide a promising roadmap towards the implementation of such an algorithm in offloading-heavy applications.