The interaction capability provided by the Internet of Things (IoT) significantly increases communication between human and machine, changing our lives gradually. However, the abundant constructions of 5G small base stations (SBSs) and large-scaled access of IoT terminal equipment (TE) will surely cause a dramatic increase in energy expense costs of a wireless communication system. In this study, we designed a bilateral random model of TE allocation and energy decisions in IoT, and proposed a mixed energy supply algorithm based on a matching game and convex optimization to minimize the energy expense cost of the wireless communication system in IoT. This study divided the problem of minimizing energy expense cost of the system into two steps. First, the random allocation problem of TEs in IoT was modeled to a matching game problem. This step is to obtain the TE matching scheme that minimizes the energy consumption of the whole system on the basis of guaranteeing the quality of service of TEs. Second, the energy decision problem of SBS was modeled into a convex optimization problem. The energy purchase scheme of SBSs with the minimum energy expense cost of the system was obtained by solving the optimal solution of the convex optimization. According to the simulation results, the proposed mixed energy supply scheme can decrease the energy expense cost of the system effectively.
The future mobile communication system will face a challenge of explosive growth of access devices, which leads to a sharp increase of energy consumption at base stations (BSs). How to gain signal to interference and noise ratio (SINR) of more users while assuring system stability becomes a problem that has to be solved urgently. In this article, a multi-BS heterogeneous network system model which is equipped with renewable energy (RE) production devices is constructed and a hybrid optimization algorithm which is composed of a linear programming (LP) algorithm and multiobjective genetic algorithm is proposed. The hybrid algorithm considers RE output and existing inventory of batteries of BSs comprehensively and maximizes the sum of SINR (sum-SINR) of users in the system through user scheduling and beam forming. First, the hybrid algorithm relaxes the binary parameters of user scheduling into a continuous value during user scheduling, thus changing the mixed integer nonlinear programming problem into an LP problem. As a result, user scheduling scheme of maximizing sum-SINR of the system is solved. Subsequently, the algorithm changes the maximum sum-SINR problem of the system into a multiobjective optimization (MOP) problem of maximum SINR of different BSs during beam forming. The fast nondominated sorting genetic algorithm II that carries the elitist strategy gains the Pareto frontier of MOP based on semidefinite programming constraint and finds the beam forming scheme of maximum sum-SINR of the system. According to stimulation results, the proposed hybrid optimization algorithm can increase sum-SINR of system users effectively while assuring the system stability.