1: \begin{abstract}

2: The convergence of communication and computing has lead to the emergence of {\it \ac{MEC}}, where computing resources (supported by \acp{VM}) are distributed at the edge of the \ac{MN}, i.e., in \acp{BS}, with the aim of ensuring reliable and \mbox{ultra-low} latency services. Moreover, \acp{BS} equipped with {\it\ac{EH}} systems can decrease the amount of energy drained from the power grid resulting into energetically \mbox{self-sufficient} \acp{MN}. The combination of these paradigms is considered here.  

3: Specifically,  we propose an online optimization algorithm, called \ac{ENAAM}, based on foresighted control policies exploiting (\mbox{short-term}) traffic load and harvested energy forecasts, where \acp{BS} and \acp{VM} are dynamically switched on/off towards energy savings and QoS provisioning. Our numerical results reveal that ENAAM achieves energy savings with respect to the case where no energy management is applied, ranging from $57\%$ and $69\%$. Moreover, the extension of ENAAM within a cluster of BSs provides a further gain ranging from $9\%$ to $16\%$ in energy savings with respect to the optimization performed in isolation for each \ac{BS}. 

4: \end{abstract}

5: