20 Metres From the House, a 5G Antenna - Rivariva?
By the close of the first quarter of 2026, global 5G subscriptions had passed 3.1 billion, after a single quarter added 162 million connections.
"no indigenous consultation under ILO Convention 169 took place"
Ericsson forecasts the total will more than double to 6.4 billion by the end of 2031, with two-thirds of all mobile subscriptions on the standard. Half of the world outside mainland China still lacks 5G coverage, though 2025 brought 400 million more people inside the footprint.
Network data traffic, counting both mobile and fixed wireless access, grew 22% between the first quarter of 2025 and the same period in 2026, pushing the monthly global total to around 210 exabytes. Video made up roughly 75% of mobile data at the end of last year. During 2025 the portion of mobile data handled by 5G climbed from 34% to 48%, with Ericsson projecting 85% by 2031. Behind those curves lies a physical change, that is, more antennas closer to where people live.
North America led 5G penetration at 79% by the end of 2025, ahead of North East Asia at 60% and Western Europe at 54%, while mainland China alone counted 1.2 billion 5G subscriptions. By the close of 2025, global 5G population coverage held at 60%, though mid-band remained patchy outside the leading markets. Some 390 service providers have launched commercial 5G, more than 90 of them on the standalone core. Australia, inside the South East Asia and Oceania grouping, is pushing toward 5G Advanced.
Across suburban Oceania, 5G mostly rides the same mid-band and low-band spectrum already familiar from 4G, though dense venues invite millimetre-wave cells around 26 to 28 gigahertz. Ericsson puts South East Asia and Oceania on a path toward roughly 670 million 5G subscriptions by 2031, with Australia adding autonomous networks.
What does the measurement record show? In Queenstown and Auckland, an independent assessor engaged by Manatū Hauora (NZ Ministry of Health) sampled seven sectors across five 5G sites in early 2020. While a download speed test loaded the 5G carrier, exposure from that carrier stayed at roughly one tenth of the total from the site. Against the public limit in the New Zealand standard NZS 2772.1, all readings fell far below the threshold. Full power on the installed transmitters would not change that verdict.
Around several hundred NZ cellsites, the highest exposure at most locations comes in below 1% of the national limit, with no site breaching it. Public limits themselves include a safety factor of 50 below the level where effects appear, a margin meant to protect the public. Manatū Hauora holds that nothing in how 5G encodes a radio wave makes it interact with the body differently from earlier generations, since millimetre energy barely penetrates past the skin.
Not all specialists accept that account, though. A 2020 letter in journal Radiation Protection in Australasia, signed by researchers from Australia and Aotearoa, argued that 5G had been deployed without adequate safety testing on humans. The authors had earlier reviewed 1955 peer-reviewed studies, reporting biological effects in 68% of them. They faulted the national radiation agency for assurances unsupported by primary data. Beam-forming and antenna density drew particular concern, alongside a millimetre-wave band once confined to radar and crowd-control uses.
In a 2023 literature review, the authors tried to weigh the same field. They screened 1269 records and kept 24 for synthesis, finding harmful effects in 10 of those 24 while judging the overall results ambivalent. Most observed changes were mild and reversible, limited to the cellular level, with temperature rises under 1°C, inside the limits set by the relevant engineering guidelines. Yet the reviewers cautioned that a few non-thermal effects should not be dismissed, and that lifetime, whole-population exposure had not been studied.
Where does that leave the corporate teams advising Oceania markets? Between a regulator reporting exposures far under the limit and a body of researchers calling the long-term picture unsettled, the message to a community cannot paper over the disagreement. A consent notice or a council submission must weigh in the confidence and the caveats, set down in plain wording, whether in Fijian or Bislama. A flattened caveat can tip a measured finding into alarm and mistranslate the hedge.
Chile shows how fast that pressure can arrive. After building more than 6,400 of its own antennas across the country, the operator WOM pledged a further US$200 million by year-end, on top of roughly US$1 billion already spent, and opened trials of 5G Advanced. Those trials follow the 3GPP Release 18, an intermediate step toward 6G, and lean on a Qualcomm 1024 QAM method that packs more data into a single signal, with latency gains theorised near 25% from the same spectrum. Half-year results showed revenue up 5.2% to more than US$182 million, with ARPU up 1.7%.
Among the obligatory localities in that national 5G concession was Rapa Nui, the Polynesian island 3,500 kilometres off the Chilean coast, long tied to a single carrier and patchy links. In the Hahanau sector, a resident learned of the plan only when excavation started for a 20-metre antenna 20 metres from his house. Neighbours soon found other parts of the island contesting four more antennas. Their core grievance is procedural, since no indigenous consultation under ILO Convention 169 took place before a build-out on territory protected by the State as a Historic Monument.
The regulatory path drew further objections. Before the courts, the environmental assessment service noted that WOM had filed a pertinence query about whether the project needed a full study, then let the process lapse by missing a deadline for follow-up questions. WOM treated its consultation duty as met by approaching CODEIPA, the island development commission. Residents counter that the body only advises, without power to authorise telecom sites or to replace an indigenous consultation. Entel, by contrast, had earlier placed its towers in remote spots, raised and camouflaged to spare the island's scenic value.
What about the bees? Beekeepers on the island warn that several antennas rise near active apiaries, citing Chilean and overseas research on how electromagnetic fields from masts and power lines disturb bees and thin out nearby pollination. In 2024 the agricultural authority certified Rapa Nui as the world's first zone free of the bee diseases common elsewhere, a status that makes the local honey prized. A beekeepers' committee has petitioned the agriculture service and the presidential delegation to apply the precautionary principle.
The islanders are not asking for a rubber-stamp consultation after the fact. They want construction halted and the installed masts dismantled, with a real consultation and impact study before any project returns. They also note that Starlink had already eased the connectivity gap once cited as absolute. A rights advocate, Tai Riroroco, framed the territory as ancestral and protected, dismissing any firm's bid to override the community with connectivity as the excuse. For the firms drafting those notices, consent rendered into Rapa Nui and Spanish must come across as consent, never as a settled deal dressed up for the record.
But look further out. The pressure only builds. Ericsson expects total mobile network traffic, fixed wireless included, to grow about 2.5 times to 515 exabytes a month by 2031, with 6G specifications due by 2028 or 2029. Smartglasses and other wearables put the antenna closer to the head than a handset ever did. As that hardware creeps toward the body, the safety argument will keep crossing language lines from Samoan to Rapa Nui. Handled with skill, a contested mast still earns an informed hearing in Hanga Roa as readily as in Apia.
Network data traffic, counting both mobile and fixed wireless access, grew 22% between the first quarter of 2025 and the same period in 2026, pushing the monthly global total to around 210 exabytes. Video made up roughly 75% of mobile data at the end of last year. During 2025 the portion of mobile data handled by 5G climbed from 34% to 48%, with Ericsson projecting 85% by 2031. Behind those curves lies a physical change, that is, more antennas closer to where people live.
North America led 5G penetration at 79% by the end of 2025, ahead of North East Asia at 60% and Western Europe at 54%, while mainland China alone counted 1.2 billion 5G subscriptions. By the close of 2025, global 5G population coverage held at 60%, though mid-band remained patchy outside the leading markets. Some 390 service providers have launched commercial 5G, more than 90 of them on the standalone core. Australia, inside the South East Asia and Oceania grouping, is pushing toward 5G Advanced.
Across suburban Oceania, 5G mostly rides the same mid-band and low-band spectrum already familiar from 4G, though dense venues invite millimetre-wave cells around 26 to 28 gigahertz. Ericsson puts South East Asia and Oceania on a path toward roughly 670 million 5G subscriptions by 2031, with Australia adding autonomous networks.
What does the measurement record show? In Queenstown and Auckland, an independent assessor engaged by Manatū Hauora (NZ Ministry of Health) sampled seven sectors across five 5G sites in early 2020. While a download speed test loaded the 5G carrier, exposure from that carrier stayed at roughly one tenth of the total from the site. Against the public limit in the New Zealand standard NZS 2772.1, all readings fell far below the threshold. Full power on the installed transmitters would not change that verdict.
Around several hundred NZ cellsites, the highest exposure at most locations comes in below 1% of the national limit, with no site breaching it. Public limits themselves include a safety factor of 50 below the level where effects appear, a margin meant to protect the public. Manatū Hauora holds that nothing in how 5G encodes a radio wave makes it interact with the body differently from earlier generations, since millimetre energy barely penetrates past the skin.
Not all specialists accept that account, though. A 2020 letter in journal Radiation Protection in Australasia, signed by researchers from Australia and Aotearoa, argued that 5G had been deployed without adequate safety testing on humans. The authors had earlier reviewed 1955 peer-reviewed studies, reporting biological effects in 68% of them. They faulted the national radiation agency for assurances unsupported by primary data. Beam-forming and antenna density drew particular concern, alongside a millimetre-wave band once confined to radar and crowd-control uses.
In a 2023 literature review, the authors tried to weigh the same field. They screened 1269 records and kept 24 for synthesis, finding harmful effects in 10 of those 24 while judging the overall results ambivalent. Most observed changes were mild and reversible, limited to the cellular level, with temperature rises under 1°C, inside the limits set by the relevant engineering guidelines. Yet the reviewers cautioned that a few non-thermal effects should not be dismissed, and that lifetime, whole-population exposure had not been studied.
Where does that leave the corporate teams advising Oceania markets? Between a regulator reporting exposures far under the limit and a body of researchers calling the long-term picture unsettled, the message to a community cannot paper over the disagreement. A consent notice or a council submission must weigh in the confidence and the caveats, set down in plain wording, whether in Fijian or Bislama. A flattened caveat can tip a measured finding into alarm and mistranslate the hedge.
Chile shows how fast that pressure can arrive. After building more than 6,400 of its own antennas across the country, the operator WOM pledged a further US$200 million by year-end, on top of roughly US$1 billion already spent, and opened trials of 5G Advanced. Those trials follow the 3GPP Release 18, an intermediate step toward 6G, and lean on a Qualcomm 1024 QAM method that packs more data into a single signal, with latency gains theorised near 25% from the same spectrum. Half-year results showed revenue up 5.2% to more than US$182 million, with ARPU up 1.7%.
Among the obligatory localities in that national 5G concession was Rapa Nui, the Polynesian island 3,500 kilometres off the Chilean coast, long tied to a single carrier and patchy links. In the Hahanau sector, a resident learned of the plan only when excavation started for a 20-metre antenna 20 metres from his house. Neighbours soon found other parts of the island contesting four more antennas. Their core grievance is procedural, since no indigenous consultation under ILO Convention 169 took place before a build-out on territory protected by the State as a Historic Monument.
The regulatory path drew further objections. Before the courts, the environmental assessment service noted that WOM had filed a pertinence query about whether the project needed a full study, then let the process lapse by missing a deadline for follow-up questions. WOM treated its consultation duty as met by approaching CODEIPA, the island development commission. Residents counter that the body only advises, without power to authorise telecom sites or to replace an indigenous consultation. Entel, by contrast, had earlier placed its towers in remote spots, raised and camouflaged to spare the island's scenic value.
What about the bees? Beekeepers on the island warn that several antennas rise near active apiaries, citing Chilean and overseas research on how electromagnetic fields from masts and power lines disturb bees and thin out nearby pollination. In 2024 the agricultural authority certified Rapa Nui as the world's first zone free of the bee diseases common elsewhere, a status that makes the local honey prized. A beekeepers' committee has petitioned the agriculture service and the presidential delegation to apply the precautionary principle.
The islanders are not asking for a rubber-stamp consultation after the fact. They want construction halted and the installed masts dismantled, with a real consultation and impact study before any project returns. They also note that Starlink had already eased the connectivity gap once cited as absolute. A rights advocate, Tai Riroroco, framed the territory as ancestral and protected, dismissing any firm's bid to override the community with connectivity as the excuse. For the firms drafting those notices, consent rendered into Rapa Nui and Spanish must come across as consent, never as a settled deal dressed up for the record.
But look further out. The pressure only builds. Ericsson expects total mobile network traffic, fixed wireless included, to grow about 2.5 times to 515 exabytes a month by 2031, with 6G specifications due by 2028 or 2029. Smartglasses and other wearables put the antenna closer to the head than a handset ever did. As that hardware creeps toward the body, the safety argument will keep crossing language lines from Samoan to Rapa Nui. Handled with skill, a contested mast still earns an informed hearing in Hanga Roa as readily as in Apia.
Huri Translations
Tel. +689 89 205 483
[email protected]
PO BOX 365 Maharepa
98728 Mo'orea
French Polynesia
N°TAHITI 876649