Energy-Efficient vBBU Migration and Wavelength Reassignment in Cloud-Fog RAN

Author(s)

Tinini, Rodrigo Izidoro ^[1] ; Batista, Daniel Macedo ^[1] ; Figueiredo, Gustavo Bittencourt ^[2] ; Tornatore, Massimo ^{[3, 4]} ; Mukherjee, Biswanath ^{[4, 5]}

Affiliation

^[1] Univ Sao Paulo, Inst Math & Stat, BR-05508090 Sao Paulo, SP - Brazil

^[2] Univ Fed Bahia, Dept Comp Sci, BR-41170110 Salvador, BA - Brazil

^[3] Politecn Milan, Dipartimento Elettron Informaz & Bioingn, I-20133 Milan - Italy

^[4] Univ Calif Davis, Dept Comp Sci, Davis, CA 95616 - USA

^[5] Soochow Univ, Inst Broadband Res & Innovat, Suzhou 215006 - Peoples R China

• Document type: Journal article
• Source: IEEE TRANSACTIONS ON GREEN COMMUNICATIONS AND NETWORKING; v. 5, n. 1, p. 18-28, MAR 2021.
• Web of Science Citations: 0

Abstract

Cloud-Fog Radio Access Network (CF-RAN) is a new architecture that increases network capacity in Cloud RAN (CRAN) by moving some BaseBand Units (BBU) from cloud to fog nodes closer to Remote Radio Heads (RRH). However, fog nodes increases CapEx and OpEx. Moreover, tidal traffic fluctuations may lead to an energy-inefficient operation if resources becomes lightly loaded. To address this problem, BBUs of fog nodes could be dynamically activated, and following traffic fluctuations, migrated to cloud. By leveraging Network Functions Virtualization (NFV), virtualized BBUs (vBBUs) can be dynamically allocated, deallocated, and migrated from fog nodes to cloud. Moreover, considering a Time-and-Wavelength Division Multiplexed Passive Optical Network (TWDM-PON) fronthaul, traffic can be migrated among virtual Passive Optical Network (VPON) channels to optimize bandwidth usage. In this article, we propose an Integer Linear Programing (ILP) formulation and an algorithm based on its linear relaxation to solve this migration problem. Compared to an algorithm without migration capabilities, our proposals reduces blocking of RRHs demanding processing of 89% and achieves power savings of 38% by reducing activated processing resources and VPONs after migrations, while experiencing small rates of service interruption. Our relaxation-based solution approximates ILP optimality and reduces execution time of ILP up to 50x. (AU)

• FAPESP's process: 15/24485-9 - Future internet for smart cities
• Grantee: Fabio Kon
• Support Opportunities: Research Projects - Thematic Grants
• FAPESP's process: 14/50937-1 - INCT 2014: on the Internet of the Future
• Grantee: Fabio Kon
• Support Opportunities: Research Projects - Thematic Grants