Saturday, December 10, 2022

Cellular Operators and Sustainability

From a sustainability perspective, the overall impact of the telecommunications industry is strongly positive. The GSM Association together with the Carbon Trust have estimated that in 2018 the total carbon-equivalent (CO2e) emissions of the mobile sector are around 0.4% of total global emissions, while savings achieved through ‘Smart Living and Working’, including a range of virtualised services supported by cellular connectivity (but not including IoT solutions), amounted to nearly 4x this amount.

At Transforma Insights, our own research based on forecasts and full-lifecycle impact analysis of 254 different IoT applications suggests that IoT solutions overall will enable net CO2e savings of around 1 gigatonne by 2030, equivalent to slightly more than 2.5% of today’s level of global emissions. 62% of this saving (equivalent to 1.6% of today’s global emissions) can be attributed to wide area wireless communications networks.

Clearly, the total net contribution that cellular networks will make to global carbon emissions will be very positive, but it’s still relevant to ask what the cellular industry is doing to minimise its footprint. To this end, it is helpful to provide an overview of one of the key industry initiatives in this space. ‘Green Future Networks’ is a project initiated by the NGMN (Next Generation Mobile Networks) Alliance to address climate action demands across the telecom ecosystem. It focuses on the identification and mitigation of environmental impacts generated by mobile communications networks.The NGMN Alliance was founded in 2006 as an open forum to evaluate candidate technologies to develop a common view of the evolution of wireless networks. Current membership comprises 80+ companies including A1 Telekom Austria Group, Airtel, Bell, BT, China Mobile, Chungwha Telecom, Deutsche Telekom, Hong Kong Telecom, KPN, NTT Docomo, Orange, Singtel, SK Telecom, T-Mobile, Tele2, Telecom Italia, Telefonica, Telia, Telus, Turkcell, UScellular, and Vodafone.

Increasing the energy efficiency of cellular networks is one of the key focus areas of the Green Future Networks initiative and is in turn comprised of three elements: equipment, network, and sites.

  • Equipment includes decarbonisation approaches such as the refurbishing of network equipment, Lean Smart Packaging (LSP) and eco design of networks, focussing on reducing the environment footprint of network equipment.
  • Networks, where network efficiency can be improved by enhancements in data transmission, improving efficiency in control messaging and signalling, and adding capabilities to go into sleep modes based on traffic and load conditions.
  • Sites, recognising that base station sites account for around 73% of total energy consumed by a typical mobile network. Actions to reduce the energy consumption of base stations include site modernisation such as reducing the use of air conditioners and cooling equipment.

In this context, one of the key techniques deployed to decrease energy consumption in networks is to stimulate the adoption of latest generation communications technologies and ultimately network rationalisation. 5G networks are significantly more energy efficient than 4G networks, and earlier generations, in terms of energy consumed per bit of traffic. To be offset against this increase in unit-efficiency, of course, is the overall increase in data traffic associated with the adoption of 5G technologies. However, in the longer term, cellular operators can achieve significant savings when infrastructure supporting legacy technologies (such as 2G and 3G) is completely decommissioned.

Improved sleep mode functions for radio power amplifiers can have a notable impact since amplifiers can consume significant amounts of electricity even when they are not actively supporting transmissions. Shutting down this component (as well as other components of a base station) can be an essential energy saving technique, as can mechanisms to allow for selective activation/deactivation of particular base stations depending on local load conditions, i.e., where the local load is limited only a subset of stations are activated. In the case of 5G energy savings can reach approximately 50% through a combination of turning off power amplifiers and putting certain parts of radio frequency stages and antennas into a ‘Sleep Mode’.

Artificial Intelligence (AI) techniques can also be deployed to make networks more efficient. The key techniques include:

  • AI in network automation. Dynamic parameter adjustments can be made to optimise networks based on analysis of historic data patterns and predicted traffic behaviour. Such parameters include the activation/de-activation of sleep modes and on demand network dimensioning.
  • AI to reduce data centre energy consumption. AI can be used in data centres to control the fan speeds, cold-water production, usage of cooling and the activation of different sleep modes in servers to reduce energy consumption. These decisions can be taken based on input data related to room temperature and humidity in different areas, air flows inside the technical room, the dissipated power, and the computational load.
  • AI for energy source management. Artificial Intelligence can be utilised to optimise the blend of electricity drawn from a grid and electricity generated by renewable sources.

In terms of network planning and deployment, sustainability footprints can be minimised by optimising the blend of radio spectrum used (with lower frequencies requiring more power but higher frequencies requiring more base stations). The virtualisation of radio access networks can also bring benefits. Simply disaggregating the hardware and software components in a RAN (Radio Access Network) allows for greater flexibility, functionality, and increased network efficiency. Shifting from traditional RAN equipment to a virtualised and cloud approach enables processors supporting radio software to run other applications as well during non-peak times, which traditionally hasn’t been possible in the case of traditional RAN hardware. Conversely, custom ASICs (Application-Specific Integrated Circuits) can be deployed to increase the efficiency of certain workloads.

At base station locations, cellular operators can adopt outdoor cabinets that have embedded cooling technologies such as siphon gravity heat pipes and can also deploy improved batteries with extended operating temperature ranges. Even moving power supply units closer to radio units can reduce power transmission losses.

Meanwhile, of course, one of the most impactful techniques that a cellular operator can adopt to reduce their sustainability footprint is to source (or locally generate) renewable energy.

The range, extent, and sheer ingenuity of some of these techniques being investigated and adopted by the cellular industry underlines the fact that combatting climate change is critical for all industry sectors, even those that can already claim to be making a significant net-positive contribution.