Power issues are not limited to devices, they are also affecting the wireless network operation cost. As shown in the previous post we have seen an increasing demand of power in base stations as we moved to more advanced cellular systems. Increased bandwidth, MIMO and smart antennas, all require more power (the decrease in power due to the use of optical fibre to replace copper in the connection of the towers to the base stations and core network is not enough to offset the increased demand of the other components). This power demand increase is expected to continue as the deployment of more sophisticated edge networks (see next post in this series) will take place in the coming years.
One additional reason for the increase in power demand by the base station is the need to manage more frequency bands.
As shown in the picture, the expectation is that over the next 4 years (by 2023) base stations having to manage 5 frequency bands or more will grow from 19% to 37% (they were just 3% in 2016). Notice that these figures are even more significant as you consider the total increase in the number of base stations expected in the next 3 years (the small-cell market growth in the coming 4 years has an estimated CAGR of 23% in the US, and this is similar to the growth in those networks that will be deploying 5G).
6G will further expand the number of frequency bands and this will further increase power consumption. Additionally, as discussed in the next post, we are going to see a growing intelligence at the edges, with massively distributed data bases/data centres and computing at the edges. Add to this, actually “multiply this” with the increased number of “base stations, antennas, nodes” that will be required by a 6G communications fabric and you start to get an idea of the power consumption problem.
The power demand for ICT is expected to reach one fifth of the overall electrical power consumption worldwide by 2030, 3 times more than its level in 2010. In absolute terms the growth is even larger, considering that the power demand is (slowly) shifting from fossil fuel to electricity (mostly due to a shift to electrical transport).
Looking at the graphics on the left one can see the impressive share of power taken by the communications networks (both wireline and wireless) and by data centres. These latter are the ones that are seeing the highest increase.
We can expect this trend to continue beyond 2030, with 6G giving further impulse to electrical power use at the edges (processing, data storage and transmission).
I am not actually confident in the data on consumer devices showing a decreased demand. What I saw in Italy has been an increased demand in this sector, from basically zero in 1990 to some 3 TWh in 2020. During this time the network power consumption for one of the largest telecom operator in Italy has remained stable at 2TWh (that Operator has halved, because of competition, the number of clients but it has increased the network traffic by an order of magnitude).
For an in depth analyses and forecast of future power demand in communications networks and the analyses of its various components take a look at this article. Independently of the various sources all seem to agree that in the future data centres will be the main user of power in this area,