Quelle est la différence entre un réseau WiFi invité et votre réseau principal ?

PODCAST SCRIPT: "What Is the Difference Between a Guest WiFi Network and Your Main Network?" DURATION: ~10 minutes | VOICE: UK English, Male, Senior Consultant Tone --- [INTRO — 1 MINUTE] Welcome back. I'm going to cut straight to it today, because this is one of those topics that sounds deceptively simple — but gets organisations into serious trouble when it's not handled properly. The question is: what is actually the difference between a guest WiFi network and your main corporate network, and why does that distinction matter enormously from a security, compliance, and operational standpoint? Whether you're running a hotel chain, a retail estate, a conference centre, or a public-sector facility, the moment you offer WiFi to visitors, you've introduced a risk vector onto your infrastructure. How you manage that separation — at the SSID level, at the VLAN level, and through your authentication architecture — will determine whether your guest WiFi is a business asset or a liability. Let's get into it. --- [TECHNICAL DEEP-DIVE — 5 MINUTES] Let's start with the fundamentals. Your corporate network — what we'd call your main network — is the environment where your business-critical systems live. That's your domain controllers, your file servers, your POS terminals, your CCTV infrastructure, your ERP systems, your HR databases. Access to these resources should be tightly controlled, authenticated via IEEE 802.1X with certificates or credentials, and restricted to known, managed devices. Your guest wireless network, by contrast, is a shared, internet-only environment for visitors, customers, and contractors who need connectivity but have absolutely no business accessing your internal resources. The moment a guest connects, they should land in a completely isolated network segment with no visibility of — and no route to — anything on your corporate side. Now, here's where a lot of organisations go wrong. They think that simply having a separate SSID — a different network name — is sufficient isolation. It is not. An SSID is just a label. Without proper VLAN tagging at the switch and access point level, traffic from both SSIDs can still traverse the same Layer 2 broadcast domain. That means a device on your "GuestWiFi" SSID could, in theory, see traffic from your corporate SSID if the underlying switching infrastructure isn't correctly configured. The correct architecture is SSID-to-VLAN mapping. Your guest SSID maps to a dedicated VLAN — let's say VLAN 10 — which is trunked through your managed switches and terminated at a separate firewall interface or a DMZ zone. That VLAN has a route to the internet and nothing else. Your corporate SSID maps to VLAN 20, which routes through your main firewall with full access to internal resources. The two VLANs never exchange traffic unless you've explicitly configured inter-VLAN routing with appropriate ACLs — which, for guest traffic, you should not have. On the access point side, most enterprise-grade wireless controllers — whether you're running Cisco Meraki, Aruba, Juniper Mist, or Ruckus — support multiple SSIDs per radio with per-SSID VLAN assignment. This is standard functionality. What you need to ensure is that your access points are connected to trunk ports on your switches, not access ports, so that VLAN tags are preserved all the way back to your distribution layer. Now let's talk about authentication. For your corporate network, the gold standard is IEEE 802.1X with a RADIUS backend — ideally with certificate-based EAP-TLS rather than username-password methods. This ensures that only domain-joined or certificate-provisioned devices can authenticate. If you're running a RADIUS infrastructure, it's worth looking at RadSec — that's RADIUS over TLS — which encrypts the authentication traffic between your access points and your RADIUS server. There's a detailed guide on that at [RadSec: Securing RADIUS Authentication Traffic with TLS](/guides/radsec-radius-over-tls) if you want to go deeper. For your guest network, the authentication model is fundamentally different. You're not dealing with managed devices. You're dealing with personal smartphones, tablets, and laptops belonging to people you've never met. The standard approach here is a captive portal — a web-based login page that intercepts the guest's first HTTP or HTTPS request and redirects them to a registration or terms-of-service page. This is where platforms like Purple's guest WiFi solution add significant value: rather than just presenting a basic splash page, you're capturing first-party data — name, email, demographic information — with explicit GDPR-compliant consent, which feeds directly into your CRM and marketing automation workflows. On the encryption side, WPA3 is now the recommended standard for both networks. For your guest network, WPA3-SAE — Simultaneous Authentication of Equals — provides forward secrecy, meaning that even if the pre-shared key is compromised, past session traffic cannot be decrypted. For your corporate network, WPA3-Enterprise with 192-bit mode provides the highest level of protection for sensitive environments. One more thing on the technical side: client isolation. On your guest VLAN, you should enable wireless client isolation — sometimes called AP isolation — which prevents guest devices from communicating with each other on the same SSID. Without this, a guest device could attempt to probe or attack other guest devices on the same network. This is particularly important in high-density environments like hotel lobbies, conference centres, and retail stores where hundreds of devices may be connected simultaneously. --- [IMPLEMENTATION RECOMMENDATIONS AND PITFALLS — 2 MINUTES] Right, let's talk about what goes wrong in practice. The most common mistake I see is organisations deploying guest WiFi on consumer-grade or unmanaged hardware that doesn't support proper VLAN tagging. If your access points can't trunk VLANs, you cannot achieve proper network segmentation. Full stop. This is a non-negotiable infrastructure requirement. The second pitfall is bandwidth contention. Without QoS policies, a single guest streaming 4K video can saturate your uplink and degrade performance for your corporate users. You need rate limiting on the guest VLAN — typically a per-client download cap of somewhere between 5 and 20 megabits per second depending on your uplink capacity — and traffic prioritisation that ensures corporate traffic always takes precedence. Third: DNS and DHCP. Your guest VLAN should have its own DHCP scope with a separate IP range — something like 192.168.100.0/24 — and should use a public DNS resolver like 8.8.8.8 or 1.1.1.1, not your internal DNS server. If guests are resolving DNS through your internal server, you've created an information leakage vector and potentially a DNS rebinding attack surface. Fourth, and this is critical for hospitality and retail: PCI DSS compliance. If your payment card infrastructure — your POS terminals, your payment gateways — shares any network segment with your guest WiFi, you are almost certainly in violation of PCI DSS requirements. The cardholder data environment must be completely isolated. A properly segmented guest VLAN with no inter-VLAN routing to your POS network is a foundational requirement for PCI compliance. Finally, logging and monitoring. Your guest network should have its own NetFlow or syslog feed into your SIEM. You need to be able to demonstrate, for GDPR and legal intercept purposes, who was connected, when, and what traffic they generated. Purple's analytics platform captures connection events, dwell time, and visit frequency data that feeds directly into this audit trail. --- [RAPID-FIRE Q&A — 1 MINUTE] Quick-fire questions I get asked regularly: "Can I use the same physical access points for both networks?" — Yes, absolutely. That's the whole point of multi-SSID with VLAN tagging. One AP, multiple logical networks. "Do I need separate internet connections for guest and corporate?" — No, but you do need separate firewall policies and ideally separate WAN interfaces or sub-interfaces to enforce QoS and traffic shaping independently. "What about IoT devices — where do they go?" — They get their own VLAN, separate from both guest and corporate. IoT is a third network segment, not a subset of either. "Is WPA2 still acceptable for guest networks?" — It's functional but WPA3 is strongly preferred. WPA2 with TKIP is deprecated. If you're still running TKIP anywhere, fix that today. --- [SUMMARY AND NEXT STEPS — 1 MINUTE] To wrap up: the difference between a guest WiFi network and your main network is not just a matter of a different password or a different SSID name. It's a fundamental architectural separation implemented at the VLAN level, enforced by your switching and firewall infrastructure, with distinct authentication models, QoS policies, and monitoring requirements for each. Get this right and you've got a guest WiFi deployment that's secure, compliant, and — with the right platform on top — a genuine first-party data asset for your marketing and operations teams. Get it wrong and you've got a lateral movement risk sitting in your lobby, broadcasting on 2.4 and 5 gigahertz. If you want to go deeper on the authentication side, check out the RadSec guide. And if you're evaluating guest WiFi platforms, Purple's solution at purple.ai covers the full stack — from captive portal and GDPR-compliant data capture through to WiFi analytics and venue intelligence. Thanks for listening. We'll see you on the next one. --- END OF SCRIPT