The reason for studying energy consumption in media delivery stems from a concern for the current state of the climate, and the need to mitigate the effects of human-induced climate change. These effects are due to greenhouse gas emissions associated with human activity, including the production of energy. In this regard, mitigation strategies revolve around 1) producing cleaner energy, and 2) using less energy. The latter is relevant for any sector, system or device not directly involved in producing energy, including those defined by 3GPP. However, with 70-80% of network traffic being media, media data centres, the transmission of media data and media consumption on UEs contribute significantly to the total energy consumed by mobile networks. Considering the 5G System, energy efficiency of each of its components as well as the system as a whole is required. In order to achieve increased energy efficiency - both at the component level and at the system level - the system needs first to be characterised. Such characterisation additionally enables reporting, thereby informing the various stakeholders of the system's energy performance. High-level measurements illustrated in clauses 4.1.2 and 4.1.3 are too coarse to allow system performance improvements, nor does it allow individual stakeholders, including Application Service Providers, network operators, and end users to know their own instantaneous energy use. Having access to fine-grained information on energy use, for instance on streaming an individual content asset, would allow the identification of potential energy hot spots, and it would facilitate government-mandated reporting which is increasingly prevalent in certain markets. This feasibility study therefore focuses on the possibility of putting infrastructure in place that would enable the measurement and reporting of energy consumption across the media delivery eco-system of 5G networks.

The terms power and energy are closely related, with power P being the rate at which work is done. It is measured in Watts or equivalently Joules per second (symbol W), or in derived quantities such as kW, MW or TW. Energy E is power integrated over time, measured in Joules (J ), or equivalently Watt-seconds (Ws). Larger quantities are often measured in kilo-Watt-hours kWh, mega-Watt-hours MWh or tera-Watt-hours (TWh). One kWh represents 3.6 × 10⁶ J. For the year 2020, the global annual electricity consumption (AEC) of mobile networks (including 2G up to 5G, as well as satellite communication) is estimated to have been 161 TWh, of which 146 TWh is spent by access networks, 6 TWh by the core network, and 9 TWh by support activities [8]. This represents 20 kWh per subscription per year [8]. In the period 2015-2018 this figure was estimated at 17 kWh per subscription per year [6]. To characterize the energy used to transmit data in a more fine-grained manner, energy-per-data figures are often reported, for example in kWh/GB. This suggests that a given network expends energy directly proportional to the amount of data communicated. This, however, has been shown to be an inaccurate measure due to the presence of significant fixed overheads. As an example, the servers in a data centre need to be cooled, irrespective of whether data passes through them or not. In addition, server and network switching hardware often has a fixed static base load energy consumption just for keeping it powered up. For this, and other reasons, the transmission of data incurs a base load which is related to the presence and maintenance of the infrastructure itself, plus a mark-up that depends on the amount of data being transmitted. Examples of power usage for 4G systems and use cases, taking into account such base load, are given in Table 4.1.2-1 based on a paper published in 2020 [7]. The figures tabulated represent one user out of an assumed 300 connected to a single base station that is capable of accommodating a maximum of 1000 users.

Where for a bit rate (in megabits per second, Mb/s), a is the base load (in W), b is the data-dependent term (in W/Mb/s), and P is the power consumped (in W). Based on the methodology of [7], as can be seen in Table 4.1.2-1, the fixed overhead a is relatively important at low bit rates. For larger bit rates (e.g. the file download example) the transmission rate dominates the power consumption.

According to [8] the global annual electricity consumption of smartphones and feature phones in 2020 is estimated to have been around 17 and 2 TWh respectively. According to the French ecological transition agency, ADEME [83], "5.6 metric tons of carbon dioxide (MtCO2e) emitted by the consumption of audiovisual content in France in 2022 which corresponds to linear TV, audio and video streaming on demand." ADEME adds that digital technology generates 3.5% of global greenhouse gas emissions […] and predicts that, given the rapid growth of uses in this field, this "carbon footprint" will have increased by 29% by 2030 if we follow the current trend (less live TV, but more video on demand and video streaming).

A study from Ember Climate [13] emphasises that electricity use may be mapped onto greenhouse gas via a conversion factor known as the carbon intensity measured in grams of CO₂ equivalent per kilowatt hour (g CO₂-e/kWh). The carbon intensity depends strongly on the method used to produce electricity and the energy source (e.g., coal, wind, solar, etc.), and therefore varies extensively across geographic locations. Currently, the carbon intensity ranges from under 100 CO₂-e/kWh to over 700 CO₂-e/kWh, with a global average of 436 CO₂-e/kWh (data from [13]). The measurement of greenhouse gas emissions is difficult if not impossible to perform directly based on energy sources alone, but through the locally and globally known carbon intensities, energy consumption measurements may be converted to estimates of greenhouse gas emissions. The energy consumption of a 5G network and its components could therefore be used in combination with the carbon intensity of its location-specific energy grid to estimate greenhouse gas emissions. In the case where components of mobile networks are operated via on-site electricity generators (e.g., due to grid blackouts and/or lack of electricity infrastructure) granular information about energy sources becomes critical for accurate emission accounting.

Larger companies in European member states are subject to corporate sustainability reporting under the Corporate Sustainability Reporting Directive (CSRD), and following European Sustainability Reporting Standards which are available under [12] supplementing Directive 2013/34/EU of the European Parliament and of the Council as regards sustainability reporting standards. This reporting law follows the Scopes defined by the Greenhouse Gas Protocol [10] (see clause 4.2.5). Even though the Greenhouse Gas Protocol has been widely adopted by many countries, other regions in the world may be subject to local and/or additional reporting requirements.