[Paper Review] Key technologies to accelerate the ICT Green evolution -- An operator's point of view

Azeddine Gati, Fatma Ezzahra Salem+18 more|arXiv (Cornell University)|Mar 22, 2019

Advanced MIMO Systems Optimization24 references19 citations

TL;DR

This paper presents a telecom operator's perspective on accelerating the ICT green evolution through key technologies across network infrastructure, software design, and mathematical optimization. It proposes energy-efficient architectures, dynamic resource orchestration, and multi-objective optimization to reduce energy consumption by up to 70% in access networks, with green orchestration as a central enabler for balancing conflicting KPIs like energy, latency, and revenue.

ABSTRACT

The exponential growth in networks' traffic accompanied by the multiplication of new services like those promised by the 5G led to a huge increase in the infrastructures' energy consumption. All over the world, many telecom operators are facing the problem of energy consumption and Green networking since many years and they all convey today that it turned from sustainable development initiative to an OPEX issue. Therefore, the challenge to make the ICT sector more energy-efficient and environment-friendly has become a fundamental objective not only to green networks but also in the domain of green services that enable the ICT sectors to help other industrial sector to clean their own energy consumption. The present paper is a point of view of a European telecom operator regarding green networking. We address some technological advancements that would enable to accelerate this ICT green evolution after more than 15 years of field experience and international collaborative research projects. Basically, the paper is a global survey of the evolution of the ICT industry in green networks including optical and wireless networks and from hardware improvement to the software era as well as the green orchestration.

Motivation & Objective

• Address the growing energy consumption in ICT networks driven by 5G and increasing data traffic, which has become a critical OPEX and sustainability challenge.
• Identify and evaluate key technological levers that can accelerate the transition to green networks across optical, wireless, and data center infrastructure.
• Propose a holistic framework integrating hardware efficiency, software optimization, and mathematical modeling to enable real-time, energy-aware network control and orchestration.
• Enable operators to balance conflicting KPIs—energy efficiency, quality of service, latency, and business revenue—through intelligent green orchestration.
• Position green ICT as a strategic differentiator by integrating sustainability into core network design and engineering education.

Proposed method

• Propose a shift from legacy central offices to a limited number of Next-Generation Central Offices (NGCO)/NGPoP to centralize and optimize energy use, especially for content delivery networks (CDNs).
• Implement dynamic virtual machine (VM) consolidation and sleep-mode activation to reduce data center energy consumption by maximizing idle resource utilization.
• Develop a green orchestration brain using multi-objective optimization and stochastic meta-models to manage trade-offs between energy, latency, and service quality.
• Introduce a pay-to-play model based on game theory to fairly allocate end-to-end energy consumption among over-the-top (OTT) services, which may account for up to 80% of network energy use.
• Explore network architecture simplification via separation of signaling and data planes to enable shared, energy-efficient signaling networks.
• Leverage data mining and mathematical tools to optimize content placement based on popularity and location, minimizing energy use while meeting latency requirements.

Experimental results

Research questions

• RQ1 How can energy consumption in access networks—responsible for ~70% of total ICT network energy use—be reduced through architectural and technological innovations?
• RQ2 What role can green orchestration play in balancing competing KPIs such as energy efficiency, quality of service, and business revenue in dynamic network environments?
• RQ3 How can game theory and multi-objective optimization be applied to fairly allocate energy consumption costs among OTT services and network operators?
• RQ4 In what ways can software eco-design and dynamic VM management reduce data center energy consumption without compromising performance or reliability?
• RQ5 What is the impact of content placement strategies on end-to-end energy efficiency, and how can they be optimized using data-driven and mathematical modeling techniques?

Key findings

• The access network segment accounts for approximately 70% of total ICT network energy consumption, making it the primary target for energy reduction efforts.
• Energy consumption in the access network is dominated by edge equipment such as Optical Line Terminals (OLTs), which exhibit nearly constant power draw regardless of load.
• The transport network shows stable energy consumption despite traffic growth, indicating that improvements in this segment offer less relative energy gain compared to access and core networks.
• Green orchestration using multi-objective optimization can enable real-time, non-aggressive control of network configurations to balance energy efficiency with service quality and latency.
• Mathematical modeling, particularly stochastic meta-models, can manage uncertainties in energy consumption data and improve the accuracy of energy response predictions.
• Fair energy cost allocation among OTT services via game theory-based models could account for up to 80% of total network energy use, enabling transparent and equitable sharing mechanisms.