Joint Access and Resource Allocation in Ultradense mmWave NOMA Networks With Mobile Edge Computing

• Published in: IEEE internet of things journal Vol. 7; no. 2; pp. 1531 - 1547
• Main Authors: Nouri, Nima, Abouei, Jamshid, Jaseemuddin, Muhammad, Anpalagan, Alagan
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.02.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Computation offloading; Decision analysis; Delay; Delays; Edge computing; Electronic devices; Energy consumption; Internet of Things; Internet of Things (IoT) networks; IoT; Iterative algorithms; Iterative methods; Millimeter waves; millimeter-wave (mmWave); Mobile computing; mobile edge computing; Nonorthogonal multiple access; nonorthogonal multiple access (NOMA) technique; Optimization; Queues; Queuing theory; Resource allocation; Resource management; Servers; successive convex approximation (SCA); Task analysis; ultradense (UD); Wireless networks
• ISSN: 2327-4662; 2327-4662
• DOI: 10.1109/JIOT.2019.2956409

This article considers a two-tier heterogeneous network consisting of conventional sub-6-GHz macrocells along with millimeter-wave (mmWave) small cells, where mobile devices (MDs) can connect to either macrocell or small cells opportunistically via the nonorthogonal multiple access (NOMA) protocol. We employ the queuing theory in our network model to conduct an assessment on the execution delay, energy consumption and the total cost of offloading tasks in a mobile-edge computation offloading (MECO) system. The main goal is to design an energy-efficient MECO decision algorithm in an ultradense Internet of Thing (UD-IoT) network to analyze the tradeoff between execution delay and energy consumption. The proposed scheme jointly optimizes the communication and computation resource management, subject to the energy and delay constraints. Due to the mixed-integer nonlinear problem (MINLP) for resource allocation and computation offloading, an iterative algorithm along with the successive convex approximation (SCA) is proposed to achieve the optimum local frequency scheduling, power allocation, and computation offloading. The superior performance of the proposed MECO algorithm in our UD-IoT network is verified by the extensive numerical results.