Fiber is the finish line. Wireless is how a rural county gets there.
This isn't fiber or wireless — it's the order you build them in. Here's an honest comparison: what each one really costs to own over time, how far modern wireless actually carries a household, and why a wireless network is the most realistic way to fund the fiber that eventually reaches the rest of the county.
Both run electronics that wear out. Only the glass is forever — and the glass isn't the expensive part.
There's a myth that fiber is "build it once and you're done." It isn't. Fiber needs powered electronics at both ends that get replaced on a cycle — just like wireless does. The part that lasts for decades is the buried glass and conduit. And in rural country, that buried plant is about 80% of the entire cost.
Lasts for decades (build once)
The truly long-lived assets
- Fiber: the glass strands and conduit — depreciated 25–50 years, with no known wear-out mechanism. Upgrades reuse it.
- Wireless: the towers and the fiber backhaul feeding them — also decades-long, and reusable for fiber later.
Cycles every ~5–10 years (both sides)
The powered electronics — neither technology escapes this
- Fiber: the home ONT and the headend OLT/optics — swapped each generation (GPON → XGS-PON → 50G-PON), roughly every 7–10 years.
- Wireless: the subscriber radio and the tower base node — refreshed roughly every 5 years.
So the comparison most people get wrong is "fiber = permanent, wireless = constant replacement." The honest version: both replace electronics on a schedule. Moving fiber to the next speed generation doesn't mean re-trenching — it means swapping the box on your wall and a card at the office, exactly the kind of upgrade wireless does too. The real, structural difference is the one-time civil construction — the trenching, boring, conduit, and pole work that wireless skips entirely.
| Cost component | Rural fiber | Fixed wireless |
|---|---|---|
| Outside plant (trench, conduit, poles, labor) | ~80% of total cost; the dominant rural expense | None — uses existing towers & backhaul |
| Home endpoint | ONT, ~$50–100 (≈1% of per-home cost) | Subscriber radio, ~$300 — more than an ONT |
| Headend / tower electronics | OLT port + optics | Base node (~$15–25K per site, shared across many homes) |
| Electronics refresh cycle | ~7–10 years | ~5 years (more often) |
| The permanent asset | glass + conduit (decades) | tower + fiber backhaul (decades) |
The honest bottom line on cost. Wireless's advantage isn't cheaper electronics — the radio on your house actually costs a bit more than a fiber ONT, and it's swapped more often. Wireless wins because it skips the trench, and out here the trench is nearly the whole bill: fiber runs about ~$40,000 per home in Okanogan versus a small fraction of that for wireless. Over 20–30 years the gap narrows as wireless re-spends on radios — even the most fiber-favorable study calls the long-run total "comparable," not a fiber win — but the upfront difference is enormous, and upfront is what decides whether a rural home gets connected this decade or not. The full per-home fiber math →
Sources: FCC depreciation schedules (conduit ~50 yr, fiber ~25 yr, central-office electronics ~7 yr); ITU-T PON standards (GPON/XGS-PON/50G-PON reuse the same fiber); CTC independent municipal FTTH estimate (~80% outside plant); FS.com ONT pricing; Cambium FWA CPE pricing. The "~5-year wireless refresh vs ~7–10-year fiber electronics" figure comes from a CWA-union-commissioned CTC analysis — credible, but an interested source, so we state it as the contested figure it is.
Modern wireless already delivers far more than a household actually uses.
The headline-speed race hides a simple fact: what a home is sold and what it uses are wildly different. Today's fixed wireless clears the bar real households live at — with room to spare — and it keeps improving.
Wireless stays ahead of what people actually do online — streaming, video calls, school, farm and clinic systems — and the gap is widening in wireless's favor in the near term, not against it. New spectrum is the engine: the FCC opened 6 GHz with automated coordination and the shared CBRS band, which is what lets these radios run wider, faster channels each year. The platform the District chose already has a drop-in next generation on its roadmap.
And the ceiling keeps rising — gigabit-class is already here on the best links. The District's own field test recorded links past 900 Mbps, and independent monitoring shows the top tier of today's 6 GHz links clearing ~1.2 Gbps — so gigabit-plus is real now in good conditions, even if the typical busy-hour link sits lower. Each new generation and spectrum band lifts the bar again: wider 6 GHz channels today, a next-generation radio already shipping (the maker rates it around 6.4 Gbps per sector — its figure, not yet independently field-verified), bigger antenna arrays, and new mid-band spectrum on the horizon for the early 2030s. The trajectory is up, not flat.
And here's the honest limit, because we don't hide them. Those gains come with a trade-off: more speed means more spectrum, and more spectrum lives at higher frequencies that travel shorter distances and don't punch through trees as well. So the biggest headline numbers belong to short, clear links — a home miles out through forest will always see less. Wireless also shares finite spectrum, so capacity is divided among users and fades with distance, and its upload is its weaker direction just as upload demand is rising. Fiber's ceiling, by contrast, is effectively unlimited: the same buried glass has been pushed to hundreds of terabits in the lab by swapping only the electronics. So wireless is not a permanent equal to fiber — and we won't claim it is. What it is: more than enough for how rural households actually live today and for years to come, climbing with every new band and generation, deliverable now, and upgradeable as the gear improves.
Sources: Preseem passive telemetry (2025), median Tarana link rates across many ISPs; OpenVault Broadband Insights (Q4 2025) household usage; FCC 6 GHz/AFC and CBRS orders; Shannon–Hartley limit; NICT 402 Tb/s standard-fiber demonstration. Vendor roadmap figures (e.g., next-gen "6.4 Gbps per sector") are vendor-stated and not yet independently field-verified.
Wireless isn't the dead end. It's the on-ramp that pays for fiber.
A working wireless network does something a fiber-only plan can't: it connects people in months, at a fraction of the cost — and the customers, revenue, and backhaul it builds are exactly what make future fiber affordable. This is how rural co-ops and providers across the country actually do it.
This phased "wireless now, fiber where it pencils" approach is mainstream practice, not a theory. WISPs and electric co-ops across rural America run both and migrate as density grows; one Kansas provider explicitly reinvested its fixed-wireless profits into building fiber. Closer to home, Pend Oreille County PUD already runs fiber and fixed wireless on its open-access network — the same model, in Washington, working now. How other PUDs do it →
Said carefully: we're not claiming wireless profits automatically pay every fiber bill — most rural fiber is still grant- and capital-funded, and a wireless network has to keep reinvesting in its own gear. The defensible, proven claim is narrower and stronger: wireless is the fastest, cheapest way to connect people now, and it builds the customers, revenue, and backhaul that make extending fiber realistic later. Pausing the wireless network doesn't bring fiber sooner — it removes the very on-ramp that gets you there.
Sources: Fiber Broadband Association and trade reporting on WISP-to-fiber migration (Nextlink, Rise, Vistabeam, Visionary; KwiKom's reinvestment); community-broadband case studies (Butler Electric/KS, RS Fiber/MN); NRECA hybrid-broadband doctrine; Pend Oreille County PUD service pages; CostQuest/BEAD high-cost-location modeling (the costliest ~2.5% of locations ≈ ~28% of fiber spend). We deliberately keep this argument inside the broadband business — wireless revenue funding fiber — not electric ratepayers subsidizing broadband.
Fiber where it pencils, wireless everywhere else — that's not a compromise, it's the only arithmetic that connects a county this size. Wireless costs far less upfront because it skips the trench, it already carries more than rural homes use, and it's the engine that funds fiber's reach over time. Letting the public wireless network age out doesn't move the county toward fiber. It just disconnects the people fiber can't yet reach. What the upgrade is → · Why fiber costs 9× more → · See it on the map →