View in EDS

User driven multiclass cell association in 5G HetNets for mobile IoT devices

Saved in:
Bibliographic Details
Title: User driven multiclass cell association in 5G HetNets for mobile IoT devices
Authors: Waheidi, Yaser M., Jubran, Mohammad K., Hussein, Mohammed
Publisher Information: IEEE
Publication Year: 2019
Collection: FADA - Birzeit University Open Access Repository
Subject Terms: Wireless communication systems, Mobile communication systems, 5G mobile communication systems, Ultra-dense networks, Cell association, Energy conversion, Energy harvesting, Game theory, Heterogeneous distributed computing systems
Description: An article published in journal : IEEE Access, Vol. 7, pp. 82991-83000 ; Fifth generation (5G) needs to support plenty of applications and services with a wide variety of quality of service requirements. The deployment of ultra-dense small cells’ networks as a part of the heterogeneous networks architecture is one of the key technologies to achieve this. In such a dense architecture, associating devices with the network is challenging. The traditional cell association algorithms use signal-to-interference-plus-noise ratio metric. However, this is not appropriate for 5G, especially with the reduction in the cells size and the growing number of the user equipment (UE) and the Internet-ofThings (IoT) devices. In this paper, we propose a distributed multiclass user-driven cell association algorithm based on the multi armed bandit game (CA-MAB) to connect devices with different requirements to the network. Here, we focus on two classes of devices: UE devices and low-power IoT devices. The proposed algorithm is evaluated in static and mobile environments, where the convergence and equilibrium are achieved. Our performance results are validated against the central cell association method that is complex and requires a huge amount of information exchange. The results show that CA-MAB throughput and energy efficiency are within 10% of the centralized solution. These values increase by less than 5% in the case of mobility. However, they reduce with more network densification
Document Type: article in journal/newspaper
Language: unknown
Relation: IEEE Access; http://hdl.handle.net/20.500.11889/6122; https://api.elsevier.com/content/abstract/scopus_id/85068718179; 2-s2.0-85068718179
DOI: 10.1109/ACCESS.2019.2924521
Availability: https://hdl.handle.net/20.500.11889/6122
https://api.elsevier.com/content/abstract/scopus_id/85068718179
https://doi.org/10.1109/ACCESS.2019.2924521
Accession Number: edsbas.C4F8EC5F
Database: BASE
Description
Abstract:An article published in journal : IEEE Access, Vol. 7, pp. 82991-83000 ; Fifth generation (5G) needs to support plenty of applications and services with a wide variety of quality of service requirements. The deployment of ultra-dense small cells’ networks as a part of the heterogeneous networks architecture is one of the key technologies to achieve this. In such a dense architecture, associating devices with the network is challenging. The traditional cell association algorithms use signal-to-interference-plus-noise ratio metric. However, this is not appropriate for 5G, especially with the reduction in the cells size and the growing number of the user equipment (UE) and the Internet-ofThings (IoT) devices. In this paper, we propose a distributed multiclass user-driven cell association algorithm based on the multi armed bandit game (CA-MAB) to connect devices with different requirements to the network. Here, we focus on two classes of devices: UE devices and low-power IoT devices. The proposed algorithm is evaluated in static and mobile environments, where the convergence and equilibrium are achieved. Our performance results are validated against the central cell association method that is complex and requires a huge amount of information exchange. The results show that CA-MAB throughput and energy efficiency are within 10% of the centralized solution. These values increase by less than 5% in the case of mobility. However, they reduce with more network densification
DOI:10.1109/ACCESS.2019.2924521