Optimal Resource Allocation in Energy-Efficient Internet-of-Things Networks With Imperfect CSI

• Published in: IEEE internet of things journal Vol. 7; no. 6; pp. 5401 - 5411
• Main Authors: Ansere, James Adu, Han, Guangjie, Liu, Li, Peng, Yan, Kamal, Mohsin
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.06.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Combinatorial analysis; Computational geometry; Computer simulation; Convexity; Efficiency; Electronic devices; Energy consumption; Energy conversion efficiency; Energy efficiency maximization; Greenhouse effect; Internet of Things; Internet of Things (IoT); Iterative algorithms; Lagrangian dual decomposition; Nonlinear programming; Optimization; power allocation and user selection; Power management; Quality of service; Resource allocation; Resource management; Smart devices; Transmitting antennas
• ISSN: 2327-4662; 2327-4662
• DOI: 10.1109/JIOT.2020.2979169

Internet of Things (IoT) is an emerging networking paradigm that enhances smart device communications through Internet-enabled systems. Due to massive IoT devices connectivity with economic and greenhouse emission effects, the energy-efficiency poses critical concerns. Under imperfect channel state information (CSI), this article investigates joint optimization of user selection, power allocation, and the number of activated base station (BS) antennas of multiple IoT devices considering the transmit power and different Quality-of-Service (QoS) requirements in combinatorial mode to maximize energy-efficiency. The optimization problem formulated is a nonconvex mixed-integer nonlinear programming, which is NP-hard with no practical solution. The primal optimization problem is transformed into a tractable convex optimization problem and separated into inner and outer loop subproblems. This article proposes a joint energy-efficient iterative algorithm, which utilizes a successive convex approximation technique and the Lagrangian dual decomposition method to achieve near-optimal solutions with guaranteed convergence. The simulation results are provided to evaluate the proposed algorithm and its significant performance gain over the baseline algorithms in terms of energy-efficiency maximization.