L'OFDMA expliqué : Comment le WiFi 6 gère les environnements à haute densité

Welcome to the Purple Technical Briefing. I'm your host, and today we are diving deep into the core technology that makes WiFi 6 a genuine game-changer for enterprise environments: OFDMA — Orthogonal Frequency Division Multiple Access. If you are an IT director, a network architect, or a venue operations manager dealing with high-density environments — whether that is a stadium, a hospital campus, a conference centre, or a retail chain — this is the technology you need to understand to future-proof your wireless infrastructure. Let us start with the context, because context is everything here. For years, the WiFi industry was almost entirely obsessed with top speed. Every new standard boasted a higher theoretical throughput figure. WiFi 4 gave us 600 megabits per second. WiFi 5 pushed that to 3.5 gigabits per second. And the marketing around each generation was relentlessly focused on that headline number. But here is the uncomfortable truth: in the real world, especially in dense venues, the problem has never really been speed. The problem is contention. It is too many devices trying to talk at the exact same time, on the same channel, fighting for the same airtime. And that is a problem that raw speed alone simply cannot solve. So let us talk about how we got here, and why OFDMA is the answer. In WiFi 5, or 802.11ac, and in all the standards before it, the underlying modulation technology was OFDM — Orthogonal Frequency Division Multiplexing. Now, OFDM is a genuinely brilliant piece of engineering. It divides a channel into many narrow subcarriers, each transmitting data simultaneously, which makes it highly resilient to multipath interference. But here is the critical limitation: OFDM is fundamentally a single-user technology. When an access point transmits using OFDM, it allocates the entire channel bandwidth to a single client for that transmission. The whole 20 megahertz, or 40, or 80 — all of it, to one device. Think of it like this. Imagine a motorway with a single lane. A lorry pulls out and takes up the entire lane. It does not matter whether that lorry is carrying a full load or just a single small parcel — it still occupies the entire lane. Every other vehicle has to wait. In a WiFi network, that lorry is your access point, and those small parcels are the tiny packets that make up the vast majority of real-world traffic: DNS lookups, TCP acknowledgements, IoT sensor pings, instant message notifications. Tiny payloads, but each one monopolising the entire channel for its transmission. In a home with three or four devices, this is barely noticeable. But in a hotel lobby with 300 guests, or a stadium concourse with 10,000 fans all trying to share a photo at half-time, or a hospital ward where dozens of medical devices are all polling for updates simultaneously — this contention overhead becomes catastrophic. Latency spikes. Throughput collapses. The user experience degrades, and no amount of additional access points will fully solve the problem, because the fundamental inefficiency is in the protocol itself. This is precisely the problem that OFDMA, introduced in the IEEE 802.11ax standard — WiFi 6 — is designed to solve. OFDMA takes the multi-carrier approach of OFDM and extends it into the multi-user dimension. Instead of allocating the entire channel to a single client, OFDMA divides the channel into smaller frequency allocations called Resource Units, or RUs. A single 20 megahertz channel can be subdivided into up to nine distinct Resource Units using what are called 26-tone RUs. This means a single access point can communicate with up to nine different clients simultaneously, all within a single transmission opportunity. To achieve this, WiFi 6 makes a fundamental change to the subcarrier architecture. In WiFi 5, subcarrier spacing was 312.5 kilohertz. In WiFi 6, this is reduced to 78.125 kilohertz — a fourfold reduction. This tighter spacing means longer symbol durations, which has a secondary benefit: improved robustness against multipath fading. In environments like warehouses, transport hubs, or large open-plan retail floors, where signals bounce off metal shelving, concrete pillars, and glass facades, this is a meaningful improvement in link reliability. Now, the mechanism that makes uplink OFDMA work is a new management frame called the Trigger Frame. In legacy WiFi, uplink transmissions were chaotic — clients would essentially compete for airtime using a contention-based mechanism called CSMA/CA, which stands for Carrier Sense Multiple Access with Collision Avoidance. It works, but it is inherently inefficient under load. In WiFi 6, the access point takes control. It sends a Trigger Frame to a group of clients, allocating specific Resource Units to specific devices, specifying the transmission power levels, and synchronising the timing so that all client signals arrive at the AP simultaneously. The AP is now the traffic controller, not just a passive receiver. This shift from a contention-based model to a scheduled, orchestrated model is the fundamental reason why OFDMA delivers such dramatic latency improvements in dense environments. In controlled tests, WiFi 6 networks with OFDMA enabled have demonstrated latency reductions of up to 75 percent compared to equivalent WiFi 5 deployments under high client load. That is not a marginal improvement — that is a qualitative shift in network behaviour. There is one more technology worth mentioning alongside OFDMA, and that is BSS Coloring. BSS stands for Basic Service Set, and coloring refers to a 6-bit identifier added to the PHY header of every WiFi 6 frame. This identifier allows radios to distinguish between transmissions from their own network — intra-BSS — and transmissions from neighbouring networks operating on the same channel — inter-BSS. In a dense deployment where multiple access points are operating on the same channel in adjacent areas, BSS Coloring allows devices to essentially ignore inter-BSS transmissions as background noise rather than treating them as potential collisions. This spatial reuse mechanism works in concert with OFDMA to significantly reduce co-channel interference. Now let us move to implementation, because understanding the technology is only half the battle. Deploying it effectively is where the real work happens. The single most important factor in realising OFDMA's benefits is client ecosystem readiness. OFDMA requires 802.11ax hardware on both the access point and the client device. If a client is a legacy WiFi 4 or WiFi 5 device, the access point must revert to standard OFDM to communicate with it. In a venue where 60 or 70 percent of connected devices are legacy hardware — which is entirely realistic in a hotel, a hospital, or a retail environment — the access point will spend the majority of its time in legacy mode. The OFDMA capability exists but is rarely exercised. This is why profiling your client ecosystem before committing to an architecture refresh is not optional — it is essential. Tools like Purple's WiFi Analytics platform give you that visibility. The second critical implementation decision is channel width. This is counterintuitive for many engineers who have spent years chasing wider channels for higher throughput. In a dense deployment, wider channels are often actively harmful. An 80 megahertz channel occupies four times the spectrum of a 20 megahertz channel. In the 5 gigahertz band, there are a limited number of non-overlapping channels available. If every access point in a dense venue is configured for 80 megahertz, the number of available non-overlapping channels drops dramatically, and co-channel interference becomes severe. The recommendation for ultra-dense deployments — stadiums, auditoriums, conference halls — is to standardise on 20 megahertz channels. A 20 megahertz channel with OFDMA serving 50 concurrent clients will deliver better aggregate throughput and far lower latency than an 80 megahertz channel struggling under contention. The third consideration is power infrastructure. Modern WiFi 6 access points are sophisticated devices. They have multiple radios, dedicated scanning radios for security and analytics, and powerful processors for OFDMA scheduling. They require more power than their predecessors. Many enterprise WiFi 6 APs require 802.3at PoE Plus, which delivers up to 30 watts, or even 802.3bt PoE Double Plus, which delivers up to 90 watts. If these APs are connected to legacy 802.3af switches, which cap at 15.4 watts, the APs will enter a power-saving mode. They will disable spatial streams, reduce transmit power, or shut down secondary radios. The result is a WiFi 6 AP performing at WiFi 5 levels, or worse. Before any WiFi 6 deployment, a full audit of the switching infrastructure is mandatory. Let me give you a rapid-fire Q and A to address the most common questions we hear from clients. Question: Will OFDMA improve the range of my network? Answer: Not significantly. OFDMA is about capacity and spectral efficiency, not coverage. It allows more devices to operate smoothly within the existing coverage area. If you need to extend coverage, you need more access points or higher transmit power — OFDMA will not solve a coverage problem. Question: Do I need WiFi 6E to benefit from OFDMA? Answer: No. OFDMA is a core feature of WiFi 6 and operates across the 2.4 gigahertz and 5 gigahertz bands. However, WiFi 6E extends the standard into the 6 gigahertz band, which is entirely free from legacy WiFi 4 and WiFi 5 clients. In the 6 gigahertz band, every connected device is WiFi 6E capable, which means OFDMA can operate at peak efficiency from day one. For mission-critical applications — think operating theatre communications or real-time venue management systems — WiFi 6E is worth the investment. Question: Is WPA3 required for WiFi 6? Answer: Yes. WPA3 is mandatory for WiFi 6 certification. It introduces Simultaneous Authentication of Equals, which provides significantly stronger protection against offline dictionary attacks compared to WPA2. For organisations subject to PCI DSS or GDPR, this is not just a nice-to-have — it is a compliance requirement. Question: What is the most common reason OFDMA is not performing as expected in a newly deployed WiFi 6 network? Answer: Legacy clients. Almost every time. When we audit a WiFi 6 deployment that is underperforming, the root cause is a high percentage of legacy devices forcing the access points into OFDM mode. The fix is a combination of client profiling, aggressive band steering, and in some cases, accelerating the refresh cycle for legacy endpoint hardware. To summarise everything we have covered today. OFDMA is the foundational technology of WiFi 6 that shifts the focus from peak single-user throughput to multi-user spectral efficiency. It divides channels into Resource Units, allowing an access point to serve multiple clients simultaneously, dramatically reducing latency and contention overhead. It is the reason WiFi 6 feels so much more responsive in dense environments, even when the headline speed numbers are not dramatically higher than WiFi 5. To realise its benefits in your deployment, you need to profile your client ecosystem and understand what percentage of your devices are WiFi 6 capable. You need to design for capacity rather than coverage, using 20 megahertz channels in high-density areas. And you need to ensure your wired infrastructure can deliver the power that modern WiFi 6 access points require. For your next steps, I would recommend starting with a wireless site survey and client ecosystem audit. Use that data to build a phased migration plan that prioritises the highest-density areas first — your conference spaces, your lobbies, your concourse areas. And ensure that your network management platform gives you the visibility to monitor OFDMA utilisation, client distribution, and channel efficiency in real time. Thank you for joining this Purple Technical Briefing. For detailed deployment guides, architecture templates, and vendor-neutral best practice documentation, visit the Purple resources centre. Until next time.