South Africa Just Unlocked 900MHz of New Spectrum - Does Fixed Wireless Now Rival Fibre?

On 22 May 2026, the Independent Communications Authority of South Africa gazetted a new set of regulations - Dynamic Spectrum Access and Opportunistic Spectrum Management - that opened two bands of radio spectrum to licence-exempt, database-driven shared use. We fetched and read a detailed regulatory breakdown of exactly what changed directly today rather than relying on a summary.

In practice, "licence-exempt" doesn't mean unregulated - it means operators don't need an individual Icasa licence to use the spectrum, but access is still coordinated through a database-driven sharing system (Icasa's own materials describe this as a "Unified Spectrum Switch") so multiple users don't collide on the same frequencies in the same place. That's a meaningfully lower barrier to entry than South Africa's traditional spectrum-licensing process, which is exactly why wireless ISPs - who typically can't afford to bid in a spectrum auction the way MTN or Vodacom can - are the ones most excited about it.

Icasa's own regulations name the intended beneficiaries directly: community networks, small businesses, and operators in rural or underserved areas. Nothing in the gazetted regulations names a specific home-broadband product, a coverage target, or a retail price - this is spectrum policy, not a product announcement.

On 20 August 2026, Paul Colmer - an executive member of the Wireless Access Providers' Association (WAPA), the industry body representing South African wireless ISPs - published an opinion piece making the case that this spectrum unlock changes fixed wireless's position relative to fibre. His central line, quoted directly: "Modern fixed wireless can now deliver the speed, capacity and reliability to be the first choice rather than the consolation prize."

His supporting argument, in brief: trenching fibre to dispersed rural populations is expensive per home reached ("no amount of enthusiasm changes the cost of trenching kilometres of road for a handful of customers," in his words), a single wireless site can serve several use cases at once (home broadband, business connectivity, CCTV backhaul, agricultural monitoring, IoT), and South Africa's relatively thin legacy copper network means the country never built the dense wired infrastructure that made wireless a consolation prize elsewhere.

It's a genuinely reasonable argument on its own terms - and it's worth being equally clear that it's an advocacy piece from an interested industry body, not independent testing or neutral reporting. WAPA's members are the operators who stand to benefit directly from more licence-exempt spectrum. That doesn't make the argument wrong; it does mean the claims deserve the same sourcing discipline this site applies to every operator announcement, price change or coverage claim it covers.

Strip away the advocacy framing and the underlying technology comparison between fibre and fixed wireless hasn't fundamentally changed - what's changed is how much spectrum wireless operators now have available to work with. Here's the practical comparison as it stands today, not as a hoped-for future state.

The honest summary: fibre keeps its edge in latency, consistency and symmetrical speed everywhere it's already built, because glass is physically unaffected by weather, interference or how many neighbours are online at once. Fixed wireless keeps its edge in how fast and cheaply it can reach a home fibre hasn't reached yet - and more available spectrum genuinely helps that case, by giving wireless operators more capacity to serve more simultaneous connections per site without everyone's speed dropping at peak time. That's a real technical benefit of the May 2026 spectrum opening. It isn't, on its own, evidence that fixed wireless now matches fibre's performance where fibre is already live.

This is the gap between the argument and the market: as of today, no operator has launched a named consumer product on either the lower 6GHz or N77 band that Icasa opened in May 2026. Consumer fixed wireless does already exist in South Africa - it's just running on different, previously licensed spectrum, and it's worth knowing what's actually buyable right now versus what's still an industry argument about future capacity.

In other words: if you're weighing fixed wireless against fibre for your own home today, the products actually available to you are Rain's or Vodacom's existing 5G home plans (or MTN's, where its own fixed-wireless push has gone live) - not a new product built specifically on the May 2026 spectrum opening. That spectrum matters for the wireless industry's future capacity and cost base, which can flow through to better service or lower prices over time, but it isn't a separate sign-up option on the market this week.

Icasa's May 2026 spectrum opening is real, confirmed against a detailed regulatory analysis, and genuinely useful to the wireless-ISP industry: 900MHz of combined new capacity across the lower 6GHz and N77 bands, available on a licence-exempt, database-coordinated basis, aimed squarely at community networks, SMEs and rural operators who can't compete for a traditional spectrum auction.

WAPA's argument that this makes fixed wireless "the first choice rather than the consolation prize" is a reasonable industry position, made by the industry body whose members benefit most directly from it being true - and it's not yet backed by a single named operator, base station count, coverage map or consumer price on either newly opened band. Until one of those things exists, the practical advice for a South African household hasn't changed: where fibre already reaches your address, it remains the stronger choice on the metrics that matter day to day. Where it doesn't, and especially in dispersed or rural areas, fixed wireless - today's live products, not a still-unlaunched one built on this specific spectrum - is a genuinely reasonable option worth comparing on price and coverage at your own address.