It is an old story for both wireless and fixed broadband provision, the question of reaching the last few per cent of the population and reducing the urban/rural digital divide.
The US Competitive Carriers Association (CCA) is latest to weigh in with an estimate that it will cost an extra $36bn beyond what operators are willing to spend on infrastructure to establish true 5G ubiquity, as opposed to the nationwide coverage claimed by major MNOs on the basis of a certain proportion of the population reached.
Such ubiquity, the CCA would contend, is desirable not just to reach those relatively small numbers of people living in the most remote areas, but also to ensure that many others are able to obtain a 5G signal a greater proportion of the time while they are travelling. This last point shows up well in recent data from the mobile analytics group Opensignal on the major US operators. This reveals that among subscribers to 5G services, T-Mobile users are connected to 5G 36.3% of the time, AT&T users 22.5% and Verizon users just 10.5%. These figures are rising by around 2% to 3% per quarter at present, so it will be a long time on that basis before they would obtain 5G access nearly all the time during their day-to-day movements.
This Opensignal data also reveals the ambiguities that lie in defining ubiquitous coverage, especially with respect to 5G with its different variants. It shows for example that while T-Mobile’s highest range but lower capacity 5G service using low band 600 MHz spectrum reaches 300m USA citizens, its ‘Ultra Capacity 5G’ straddling midband and millimeter wave spectrum reaches 150m or half that number, although on track to touch 200m by the end of 2021.
This begs the question of whether ubiquity is achieved just when low band 5G spectrum has reached the whole country, in which case there would still be a digital divide. It is after all futile to assume that everybody can have access to precisely the same capacity or QoS wherever they are, especially over a wireless network with its variations in signal quality that can never be entirely eradicated.
In practice, the objective of rural broadband strategies should be to ensure a level of coverage and quality such that everybody has access to service of fundamental social and commercial importance, like video conferencing and online education. This is the view espoused recently by the Chinese government in highlighting how universal access to online education had helped mitigate the plain of lockdowns during the Covid-19 pandemic, trumpeting the virtues of nationwide cellular coverage.
Yet despite the country leading the world in 5G roll out, it was the old workhorse of 4G that was providing that coverage for well over half the population, and nearly all in rural areas. It might be cruising in 5G, but the rapid rollout is also widening its digital divide.
This was confirmed by figures from China’s Ministry of Industry and Information Technology in September 2020 in mid-pandemic, finding that although the country had by then deployed 480,000 5G base stations, twice the rest of the world combined, 30% of these were concentrated in Shenzhen, Beijing and Shanghai serving just 4% of the population.
This unbalanced roll-out is not atypical but meant that in the short term 5G had increased rather than diminished the urban/rural digital divide. So, although China might have established a base level of broadband connectivity with 4G, any high-tech business seeking to exploit cutting edge technologies such as Virtual and Augmented Reality would still be drawn to a major city.
There have been plenty of studies devoted to this conundrum of how to balance speed of wireless deployment in areas important for national competitiveness with the desire to keep rural areas sufficiently up to speed such that they can at least participate in most digital activities. We recall a 2018 paper outlining the pros and cons of different policy options for the UK, but equally applicable in any developed country (‘The cost, coverage and rollout implications of 5G infrastructure in Britain, Telecommunications Policy, 2018)
The paper concluded that under anticipated deployment plans, 90% of the country would be covered by 5G at that speed by 2027, but that the remaining 10% would be denied that bit-rate. The paper advocated a realistic approach where those final 10% would be guaranteed the considerably lower speed of 10Mbps, a discrepancy unacceptable to many. Nonetheless, the paper did a good job setting out alternative scenarios and milestones based on varying rates of both spatial and temporal 5G deployment.
There have been plenty more findings along those lines, with telecoms lobbying group ETNO in March 2021 reporting a study it had commissioned with consulting firm BCG, suggesting that the European Union would have to spend €300bn Euros ($355bn) on its telecoms infrastructure by 2025 to extend superfast 5G across the 27 member states.
This came just after former Google CEO Eric Schmidt had argued that one million extra cell sites would be needed at a cost of $70bn to extend 5G service to just 80% of the USA’s population, never mind geographical area. That is far more than the CCA’s more recent estimate, but may have been based on assumptions of more ambitious performance and capacity.
Schmidt also argued that the USA was repeating the mistakes made in Europe over 3G auctions in the noughties, where governments hobbled operators by extracting too much money, leading to what he called a lost digital decade.
Operators themselves have also piled into the debate over the divide, with Vodafone recently publishing a report commissioned with Deloitte starting with the finding that while 86% of all EU households currently have ‘Next Generation Access’ (NGA) connections, only 59% in rural areas do. These are mostly fixed broadband connections at present, and Vodafone defined NGA not by speed but by technology, being one of VDSL or VDSL2, FTTP (fiber-to-the-premise), DOCSIS 3.0 and DOCSIS 3, the latter two being data over cable TV standards. In practice this means at least 50Mbps, although often a lot more now.
Vodafone also noted the huge disparity between countries, with those such as Cyprus, Malta and Belgium that are small or densely populated enjoying near 100% rural NGA coverage, while others such as Lithuania ad Bulgaria, as well as Finland despite being technologically quite advanced below 30%. The key point though was that, in future, countries would increasingly be looking to cellular and particularly 5G to fill in rural coverage more cost effectively than fixed line with its dependence on getting fiber close to the premise, whether the service is provided by traditional telcos of cablecos.
This points towards FWA (fixed wireless access) as a direct substitute for wired connections, according to Vodafone, but the operator conceded that this would not completely heal the digital divide either. As Vodafone put it: “These alternative technologies also come with the same trade-offs. In the case of FWA, increasing coverage with the technology requires a reduction in the capacities and speeds that can be offered.”
While that is true, there are circumstances where FWA does come close to bringing remote premises closer to speed, and that is where terrain is quite flat so that line of sight connectivity, ort very close to it, can be achieved over distances up to 7 Kms. We have reported before on how that is happening in the Australian Outback for example as part of that country’s National Broadband Network (NBN), exploiting millimeter wave connectivity with beamforming to create wires through the air to reach remote farms and homesteads.
There are parts of the USA with similarly flat terrain, especially the plains of the Midwest, but many that are quite undulating. The CCA took such factors into account in its study “to help determine a realistic budget to meet the aspirations of the 5G Fund” and offer up a cost model based on real world data to “truly bring ubiquitous 5G coverage to rural America — to all of America,” according to CCA president and CEO Steve Berry.
Based on a model called CostQuest developed by the eponymous broadband consulting firm, the study “determines that the current budget for the 5G Fund is inadequate and additional support is needed to ensure all consumers – and not just those who will be covered by currently planned private investment – have access to ubiquitous 5G,” said Berry.
The CostQuest model looks at unserved areas, based on the latest LTE coverage maps from the FCC combined with information on roads and areas across the country. That led to the estimate that about 37,000 new sites will be needed to get 5G to all but a few, very remote areas of the USA, most of those being new towers costing around $1m each.