Mobile edge computing (MEC) is a promising technology which is able to provide cloud computing (CC) capabilities in the vicinity of the users in a fifth-generation network. In this way, the users can overcome the constraints of their devices such as battery, computational and storage capacity, and also the long delay of CC. In this paper, we first introduce the proposed system model assuming that network users can accomplish their processing in two ways. The first method is local processing in the users' equipment and the second method is offloading the processing in the edge server (ES) which is embedded in the base station (BS). The main goal of this paper is to minimize the total cost of the network which includes a weighted combination of the consumed energy and the latency of accomplishing the users' processing of the whole network. Weighting these functions will be performed based on the percent of the remained energy in each device. Then, we model this problem as a constraint optimization problem and since the resulting problem is a non-convex and also mixed integer non-linear programming (MINLP) problem, we use the successive convex approximation (SCA) algorithm to find the optimal solution for this problem. Simulation results show that this joint optimization of the parameters leads to a reduction in the total cost of the network.