Self-Sustainable Dense Cellular M2M System With Hybrid Energy Harvesting and High Sensitivity Rectenna

Abstract

The Internet of Things (IoT) is envisioned as an essential enabler of the smart city, which 5G small cell is going to cover. In this paper, we model the dense cellular M2M communication system using a Ginibre determinant point process and study the corresponding green communication metrics, including grid energy consumption and grid energy efficiency (EE). A dedicated millimeter-wave-based hybrid energy harvesting (EH) mechanism is proposed to maintain self-sustainable communication. Two EH modes are discussed, of which Mode 1 is suitable for long-term operation, while Mode 2 is preferred in dealing with a temporary energy shortage. We first minimize the average grid energy consumption by solving a large-scale linear programming problem. When the battery at machine device is charged with sufficient energy at the initial time, closed-form formula describing the average grid energy consumption of both the modes can be derived under moderate throughput constraints, which is proved to have clear relationship with 1) initial battery energy level at machine gateway and machine device; 2) specific hybrid EH time-division strategy; 3) system operation time; and 4) static power consumption at machine gateway and machine device. The numerical solutions are obtained using the interior-point method, and the results are illustrated and discussed. When dealing with the originally non-convex EE maximization problem, we first transform it into a series of sub-problems using the successive convex approximation method and then iteratively solve them with a trick combining Taylor expansion and square error minimization. Non-orthogonal multi-access (NOMA) transmission is investigated in our model to further improve the EE performance. The experiments show that NOMA achieves nearly three times throughput promotion and two times EE improvement.