Canaux DFS : ce qu'ils sont et quand les éviter

DFS Channels: What They Are and When to Avoid Them A Purple WiFi Intelligence Briefing — Approximately 10 Minutes --- INTRODUCTION AND CONTEXT — approximately 1 minute Welcome to the Purple WiFi Intelligence Briefing. I'm your host, and today we're going deep on a topic that trips up even experienced wireless engineers: DFS channels. Dynamic Frequency Selection. If you've ever had a venue's WiFi suddenly drop clients mid-session, seen access points go silent for sixty seconds with no obvious cause, or had a hotel guest complain that their connection vanished during check-in — there's a reasonable chance DFS was involved. This briefing is aimed at IT managers, network architects, and venue operations directors who need to make a decision about DFS channels this quarter. We're not going to spend time on theory for its own sake. We're going to cover what DFS actually is, why regulators mandate it, where it causes operational pain, and — critically — how to build a channel plan that protects your guest experience and your SLA commitments. Let's get into it. --- TECHNICAL DEEP-DIVE — approximately 5 minutes So, what is DFS? Dynamic Frequency Selection is a regulatory mechanism defined under IEEE 802.11h and mandated by bodies including Ofcom in the UK, the FCC in the United States, and ETSI across Europe. The core requirement is straightforward: any WiFi device operating in the 5 GHz band between 5250 and 5725 megahertz — that's channels 52 through 144 — must be capable of detecting radar signals and, if detected, vacating that channel within ten seconds. Why does this exist? Because those frequencies are shared with primary users: weather radar systems, military radar, air traffic control, and maritime navigation. WiFi is a secondary user. The primary users have absolute priority, and DFS is the mechanism that enforces that. Now, the operational implications of this are significant. Before an access point can transmit on a DFS channel, it must complete what's called a Channel Availability Check — a CAC. During the CAC period, the AP listens passively for radar signals. It cannot transmit. It cannot serve clients. The CAC period is typically 60 seconds for most DFS channels, but it extends to 600 seconds — that's ten minutes — for channels in the 5600 to 5650 megahertz range, which overlap with weather radar. Those channels are 120, 124, and 128 in the standard channel numbering. Think about what that means operationally. If an AP detects radar and is forced off a DFS channel, it must switch to an alternative channel and complete a new CAC before it can resume service. During that window, every client associated to that AP is disconnected. In a hotel with 200 rooms, that's potentially hundreds of guests losing connectivity simultaneously. In a retail environment, it could mean point-of-sale terminals going offline. In a conference centre during a keynote presentation, it means the presenter's laptop drops off the network at the worst possible moment. The 5 GHz band is divided into what are called UNII sub-bands. UNII-1, covering channels 36, 40, 44, and 48, is entirely DFS-free. These are your safe channels — no radar detection requirement, no CAC, no risk of sudden channel evacuation. UNII-3, covering channels 149 through 165, is also DFS-free in most jurisdictions, though there are some country-specific exceptions worth verifying. The problem is that UNII-1 and UNII-3 together give you only nine non-overlapping 20 MHz channels. When you're deploying in a high-density venue — a stadium, a convention centre, a large hotel — nine channels is not enough to build a clean, non-overlapping cell plan. That's the tension at the heart of DFS channel planning. DFS channels give you access to an additional 475 megahertz of spectrum — channels 52 through 144 — which is enormously valuable for capacity planning. But that spectrum comes with operational risk that varies dramatically depending on your venue's physical environment. The key variable is radar proximity. If your venue is within approximately 30 to 50 kilometres of a weather radar installation, military base, or major airport with approach radar, your DFS channels will trigger. Not occasionally — regularly. The UK has a dense radar footprint. Ofcom's radar database shows weather radar installations across the country, and many major cities — including London, Manchester, Birmingham, and Edinburgh — have radar systems operating in the DFS bands within that radius. There's also a less obvious source of DFS triggers that catches many engineers off guard: false positives. Certain types of equipment generate RF signatures that DFS algorithms misidentify as radar. FHSS devices, some industrial wireless systems, and even poorly shielded microwave ovens have been documented as DFS false-trigger sources. In a venue with a commercial kitchen — a hotel, a conference centre, a hospital — this is a real operational risk. The DFS detection algorithm itself has evolved. Modern access points from vendors like Cisco, Aruba, Ruckus, and Juniper Mist implement what's called Enhanced DFS, or EDFS, which uses more sophisticated pulse pattern recognition to reduce false positives. But even EDFS is not immune, and the regulatory requirement to vacate within ten seconds means the impact is immediate regardless of whether the trigger was a genuine radar pulse or a false positive. One more technical point worth covering: channel width and DFS interaction. When you're running 80 MHz or 160 MHz wide channels — which you need for Wi-Fi 6 and Wi-Fi 6E throughput targets — the probability of a DFS trigger increases proportionally. An 80 MHz channel occupies four 20 MHz sub-channels. If any one of those sub-channels detects radar, the entire 80 MHz channel must be evacuated. This is why many experienced wireless architects running high-density deployments on Wi-Fi 6 will deliberately constrain channel width to 40 MHz on DFS channels, or avoid DFS entirely and rely on 6 GHz for the wide-channel throughput. --- IMPLEMENTATION RECOMMENDATIONS AND PITFALLS — approximately 2 minutes Right, let's move to practical guidance. Here's how I'd approach DFS channel planning for a new deployment. Step one: radar environment assessment. Before you configure a single access point, check the radar footprint around your venue. In the UK, Ofcom publishes radar data. Cross-reference with your venue's coordinates. If you're within 35 kilometres of a weather radar or military installation, treat DFS channels as high-risk and plan accordingly. Step two: build your non-DFS baseline first. Channels 36, 40, 44, 48, 149, 153, 157, 161, and 165 are your foundation. In a high-density deployment, design your cell plan around these channels first. Only introduce DFS channels where you have a genuine capacity requirement that cannot be met with non-DFS spectrum alone. Step three: if you do use DFS channels, implement a fallback channel plan. Every AP operating on a DFS channel should have a pre-configured fallback channel on non-DFS spectrum. Most enterprise-grade controllers support this natively. The fallback channel should be pre-scanned and pre-validated so the AP can transition with minimal client disruption. Step four: monitor continuously. A WiFi analytics platform that provides real-time channel utilisation data, DFS event logging, and client association metrics is not optional in a high-density venue — it's essential. You need to know when DFS events are occurring, how frequently, and which APs are affected. Without that visibility, you're operating blind. Step five: validate your DFS configuration against your regulatory domain. This is a common pitfall — access points shipped with a default regulatory domain of US or worldwide may behave differently from APs configured for the UK or EU regulatory domain. The DFS requirements, CAC timers, and permitted transmit power levels differ by jurisdiction. Always verify your regulatory domain setting before deployment. The biggest pitfall I see in practice is engineers enabling DFS channels to solve a capacity problem without first assessing the radar environment. They get clean performance in the lab or during initial testing — because the CAC completes successfully — and then go live in a venue that's 20 kilometres from a weather radar installation. Within days, they're getting client complaints about intermittent disconnections that are almost impossible to diagnose without proper logging. Purple's hardware-agnostic platform integrates with your existing infrastructure to provide exactly that visibility — correlating DFS event logs with client experience metrics so you can identify whether a connectivity issue is DFS-related or something else entirely. --- RAPID-FIRE Q AND A — approximately 1 minute A few quick questions I get asked regularly. Can I just disable DFS entirely? Yes, on most enterprise controllers you can restrict the AP to non-DFS channels only. In high-risk radar environments, this is often the right call. Does Wi-Fi 6E solve the DFS problem? Largely, yes. The 6 GHz band has no DFS requirement. If you're deploying Wi-Fi 6E access points, you can run wide channels on 6 GHz without any radar detection risk. This is one of the most compelling operational arguments for accelerating Wi-Fi 6E adoption in high-density venues. What about the 6 GHz band and AFC? Automated Frequency Coordination in the 6 GHz band is a different regulatory mechanism — it's not DFS. AFC uses a database-driven approach rather than real-time radar detection, and the operational impact is significantly lower. Does Purple's platform support DFS event alerting? Yes — Purple's WiFi analytics layer can surface DFS-related connectivity events through its dashboard, helping operations teams correlate network events with guest experience data. --- SUMMARY AND NEXT STEPS — approximately 1 minute To wrap up: DFS channels are a double-edged sword. They give you access to valuable spectrum that can significantly expand your capacity in high-density deployments. But they come with regulatory obligations — CAC timers, mandatory channel evacuation — that create real operational risk in venues with radar proximity. The decision framework is straightforward. Assess your radar environment first. Build on non-DFS channels as your foundation. Introduce DFS only where capacity demands it and where you have proper monitoring and fallback configuration in place. And if you're deploying Wi-Fi 6E, prioritise 6 GHz to sidestep the DFS problem entirely. For a deeper look at channel planning tools, Purple has a guide on the best WiFi analyser tools for troubleshooting channel overlap — worth reading alongside this briefing. And if you're evaluating your guest WiFi platform's ability to surface these operational insights, Purple's analytics platform is worth a conversation. Thanks for listening. Until next time. --- END OF SCRIPT Total approximate duration: 10 minutes