Event WiFi : Planification et déploiement de réseaux sans fil temporaires

Executive Summary

Event WiFi is a distinct engineering discipline. Unlike permanent enterprise deployments, temporary wireless networks must absorb extreme client density within compressed timeframes, operate on borrowed or hired infrastructure, and meet compliance obligations — all while delivering a seamless user experience that reflects directly on the event brand. A failed network at a 3,000-person conference is not an inconvenience; it is a reputational and commercial incident.

This guide addresses the full deployment lifecycle: capacity modelling, hardware hire, backhaul provisioning, VLAN architecture, captive portal design, and on-site management. It is written for the IT professional who needs to make procurement and architecture decisions this quarter, not a theoretical overview of wireless standards. Where Purple's Guest WiFi and WiFi Analytics platform adds specific value — particularly around captive portal management, GDPR-compliant data capture, and post-event reporting — those integration points are called out explicitly.

Technical Deep-Dive

Why Event WiFi Is Different

The fundamental challenge of event WiFi is density combined with simultaneity. In a standard office deployment, you might have 100 devices spread across 1,000 square metres, with staggered connection times throughout the working day. At a conference keynote, you may have 2,000 devices attempting to associate within a five-minute window as attendees file into a hall. The RF environment, the DHCP infrastructure, and the authentication backend all need to be engineered for that peak load — not the average.

Three variables drive every architectural decision in an event deployment: client count, throughput requirement per user, and event duration. Get these wrong at the planning stage and no amount of on-site troubleshooting will recover the situation.

Capacity Planning: The Numbers That Matter

The industry baseline for high-density WiFi is one access point per 25–50 concurrent users, but this figure requires significant qualification. The ratio depends on the AP's radio capabilities, the expected mix of 2.4 GHz and 5 GHz clients, and whether the event involves heavy media consumption (live streaming, video calls) or lighter browsing and messaging traffic.

For throughput planning, a conservative estimate of 1–2 Mbps per active user is appropriate for general conference or exhibition use. For events with live streaming or broadcast-quality video requirements — such as product launches or press events — budget 5–10 Mbps per active user on the production VLAN. Your uplink must be sized to accommodate the aggregate of all VLANs simultaneously, with at least 20% headroom.

Backhaul: The Non-Negotiable Foundation

No amount of well-designed wireless infrastructure compensates for an inadequate backhaul. For events above 200 attendees, a dedicated leased line is the only appropriate uplink solution. A leased line provides a synchronous, uncontended connection with a guaranteed SLA — typically 99.95% uptime — which is fundamentally different from the shared, asymmetric broadband that most venues have installed for their own operations.

Leased line provisioning typically requires a four-to-six-week lead time. This is the single most common planning failure in event WiFi deployments: teams that begin network design two weeks before an event and discover they cannot get a dedicated circuit in time. For events where a leased line is genuinely impractical — outdoor festivals, temporary structures — a bonded 4G/5G solution using multiple SIM cards across different carriers provides a viable alternative, though with lower guaranteed throughput and higher latency.

Network Architecture and VLAN Design

Strict network segmentation is both a performance and a compliance requirement. The recommended minimum architecture for any event deployment uses three VLANs:

VLAN 10 — Guest WiFi: All public-facing attendee traffic. This VLAN connects to the captive portal for authentication and data capture. Client isolation must be enabled to prevent lateral movement between devices. DNS filtering should be applied to block malicious domains — see Purple's guide on protecting your network with strong DNS and security for implementation detail.

VLAN 20 — Staff and Point of Sale: Operational traffic for event staff, ticketing systems, and card payment terminals. If card payments are processed on this VLAN, PCI DSS scope applies and the VLAN must be fully isolated from the guest network with no routing between them.

VLAN 30 — AV and Production: Dedicated to broadcast equipment, presentation systems, and production crew. This VLAN typically requires the highest guaranteed throughput and lowest latency, and should be provisioned with QoS policies that prioritise it over guest traffic.

For larger events, additional VLANs for exhibitors, press, and security systems are common. Each SSID should map to a single VLAN, and inter-VLAN routing should be disabled at the core switch unless explicitly required.

Radio Frequency Planning

In high-density environments, the default behaviour of most enterprise APs — automatic channel selection and maximum transmit power — is actively harmful. Co-channel interference between adjacent APs on the same channel degrades performance far more than a slight reduction in coverage area.

The correct approach is to manually assign channels and reduce transmit power. On the 5 GHz band, use the non-overlapping channels available across the UNII-1 (36, 40, 44, 48), UNII-2 (52–64), and UNII-3 (149–165) bands. Reduce AP transmit power to 8–12 dBm in dense deployments. This creates smaller, cleaner cells with less interference, which improves aggregate throughput across the venue.

Band steering should be enabled on all APs to push 5 GHz-capable clients — which is the vast majority of modern smartphones and laptops — away from the congested 2.4 GHz spectrum. Reserve 2.4 GHz for legacy IoT devices and accessibility equipment that cannot connect to 5 GHz.

For outdoor events, the RF environment is fundamentally different. Without walls and ceilings to contain signal, coverage cells are larger and interference from adjacent deployments or consumer hotspots is harder to control. Directional sector antennas are preferable to omnidirectional APs in outdoor settings, as they allow you to focus coverage on specific zones — the main stage area, the food court, the registration queue — rather than broadcasting indiscriminately. All outdoor hardware must carry at minimum an IP55 ingress protection rating; IP67 is preferable for festival or exposed environments.

Captive Portal Architecture and GDPR Compliance

The captive portal is the user's first interaction with your event network and your primary mechanism for both compliance and data capture. A poorly designed portal that times out, fails to redirect correctly on iOS, or presents an unclear consent workflow will generate a disproportionate volume of support requests and undermine attendee confidence in the network.

From a GDPR perspective, any collection of personal data — email addresses, social login tokens, or device identifiers — requires a lawful basis, a clear privacy notice, and explicit consent for any marketing use. The consent must be granular: consent to use the WiFi is not the same as consent to receive marketing communications. Purple's Guest WiFi platform handles this consent workflow natively, presenting compliant opt-in flows and storing consent records with timestamps and IP addresses as required by Article 7 of GDPR.

The technical architecture of the captive portal matters for performance. A cloud-hosted portal that redirects authentication requests to an external server introduces latency into the login flow. At peak load — when hundreds of users are authenticating simultaneously — this latency can cause timeouts and failed logins. Purple's platform is architected for exactly this use case, with auto-scaling infrastructure that handles burst authentication loads without degradation.

Implementation Guide

Phase 1: Site Survey and Capacity Modelling (8 Weeks Before Event)

Begin with a physical site survey. Walk every area where attendees will be present and document ceiling heights, wall materials, structural obstructions, and existing infrastructure (conduit runs, power outlets, data ports). Use a WiFi survey tool — Ekahau Site Survey or iBwave are the industry standards — to model predicted coverage and identify dead zones before hardware is ordered.

At the same time, confirm the venue's existing network infrastructure. Identify available data ports, the location of the main distribution frame, and the capacity of any existing switches. Determine whether the venue's existing cabling can support PoE+ (802.3at) for the APs you intend to deploy, or whether you need to bring your own PoE switches and cabling.

Finalise your capacity model based on the expected attendee count, the event programme (a keynote session creates a very different load profile to a networking reception), and the throughput requirements of any production systems.

Phase 2: Hardware Procurement and Backhaul Ordering (6–8 Weeks Before Event)

Order your leased line immediately after the site survey. The four-to-six-week provisioning window is the critical path for the entire deployment. If the event venue already has a leased line, negotiate dedicated bandwidth allocation with the venue's IT team — do not assume that existing infrastructure will be made available.

For hardware, the choice between purchasing and hiring depends on the frequency of your events. For organisations that deploy event WiFi more than four times per year, ownership of a portable kit — enterprise APs, a managed PoE switch, a rack-mount router, and cabling — is more cost-effective than repeated hire. For one-off events, specialist event WiFi hire companies provide pre-configured hardware with on-site support, which reduces deployment risk significantly.

When specifying APs for hire or purchase, prioritise WiFi 6 (802.11ax) hardware for any deployment above 200 users. The OFDMA and BSS Colouring features of WiFi 6 provide meaningful performance improvements in high-density environments compared to WiFi 5 (802.11ac).

Phase 3: Pre-Event Configuration and Testing (1–2 Weeks Before Event)

Configure all network equipment in a staging environment before arriving on site. This includes VLAN configuration on the core switch, SSID-to-VLAN mapping on the wireless controller, DHCP scope configuration, and captive portal integration. Testing in a staging environment is far more efficient than troubleshooting on the day of the event.

For captive portal configuration, integrate Purple's platform at this stage. Configure the branded splash page, the authentication method (email, social login, or SMS), the consent workflow, and any post-authentication redirect. Test the full user journey on multiple device types — iOS, Android, Windows, and macOS all handle captive portal detection differently, and each has specific requirements for the redirect mechanism to work correctly.

Conduct a load test using a WiFi client simulator to validate that the DHCP scope, the authentication backend, and the uplink can handle the expected peak load. Tools such as Spirent or Ixia can simulate hundreds of concurrent WiFi clients for this purpose.

Phase 4: On-Site Deployment (Day Before Event)

Arrive on site with sufficient time to complete installation and testing before the venue opens to attendees. Mount APs according to the site survey plan — ceiling mounting is preferred for omnidirectional coverage; wall mounting is acceptable where ceiling access is not available. Run and label all cabling, and document the physical location of every AP with a photograph and a floor plan annotation.

Once all hardware is installed, conduct a post-installation survey using a laptop or dedicated survey device to validate coverage. Walk the entire attendee area and confirm signal strength of -65 dBm or better throughout. Identify and address any coverage gaps before the event opens.

Test the end-to-end user journey: connect a test device to each SSID, complete the captive portal authentication, and verify that internet access is available. Test card payment terminals on the staff VLAN. Confirm that AV equipment on the production VLAN can reach all required destinations.

Phase 5: On-Site Management and Monitoring

During the event, monitor the network in real time using the wireless controller's management dashboard. Key metrics to watch are: AP association counts (flag any AP that exceeds 80% of its recommended client capacity), channel utilisation, DHCP pool utilisation, and uplink throughput. Purple's WiFi Analytics platform provides an additional layer of visibility into user behaviour — dwell time, peak connection periods, and portal conversion rates — which is valuable both for real-time management and for post-event reporting.

Have a clear escalation process for network issues. Designate a single point of contact for all network-related support requests from event staff, and ensure that the on-site network engineer has remote access to all equipment via an out-of-band management connection that is independent of the guest network.

Best Practices

The following recommendations represent vendor-neutral best practices derived from large-scale event deployments across hospitality , retail , and conference environments.

Disable SSID broadcasting for staff and production networks. There is no operational reason for these SSIDs to be visible to attendees. Hiding them reduces the attack surface and prevents accidental connections.

Set aggressive DHCP lease times on the guest VLAN. A lease time of 30–60 minutes ensures that IP addresses from disconnected devices are reclaimed promptly. This is particularly important at multi-day events where the attendee population changes significantly between sessions.

Implement 802.1X authentication on staff and production VLANs. WPA3-Enterprise with 802.1X provides per-user authentication and eliminates the risk of a shared pre-shared key being compromised. For guest networks, WPA3-Personal or an open network with a captive portal is the standard approach.

Use DNS-over-HTTPS or DNS filtering on the guest VLAN. Public event networks are a target for DNS hijacking and phishing attacks. Applying DNS filtering — either through your upstream provider or through a dedicated DNS security service — provides a meaningful layer of protection for attendees. Purple's platform integrates with DNS security providers to apply this filtering at the captive portal layer.

Document everything. Create a network diagram, a cabling schedule, and an AP placement map before you arrive on site. This documentation is invaluable for troubleshooting during the event and for planning future deployments at the same venue.

For airport and transport hub deployments, additional security considerations apply — Purple's guide on airport WiFi security covers the specific threat model and mitigation strategies relevant to high-footfall public environments.

Troubleshooting and Risk Mitigation

DHCP Pool Exhaustion

This is the most common failure mode in event WiFi. Symptoms include devices that connect to the WiFi but cannot obtain an IP address, or that receive an APIPA address (169.254.x.x). The fix is to increase the DHCP scope size and reduce the lease time. Prevention is straightforward: size your DHCP scope to at least twice the expected peak client count and set lease times to 30–60 minutes.

Authentication Server Overload

At peak load, a large number of simultaneous authentication requests can overwhelm an on-premises RADIUS server or captive portal backend. This manifests as slow or failed logins. Cloud-hosted platforms like Purple auto-scale to handle burst loads, which is a significant architectural advantage over on-premises deployments for event use cases.

Co-Channel Interference

If multiple APs are operating on the same channel in close proximity, performance degrades significantly. Symptoms include low throughput despite good signal strength, and high retry rates visible in the wireless controller. The fix is to review channel assignments and ensure that adjacent APs are on non-overlapping channels. Reducing transmit power also helps by shrinking the interference radius of each AP.

Captive Portal Redirect Failures

Different operating systems use different mechanisms to detect captive portals. iOS uses a dedicated CNA (Captive Network Assistant) that makes HTTP requests to specific Apple URLs. Android uses a similar mechanism with Google's connectivity check servers. If your captive portal does not respond correctly to these probes, the portal will not open automatically and users will need to manually navigate to the portal URL. Ensure your captive portal is configured to intercept and respond to these specific probe requests.

Uplink Failure

A single point of failure on the uplink is the highest-impact risk in an event deployment. Mitigate this by provisioning a 4G/5G backup connection that activates automatically if the primary leased line fails. Most enterprise routers support dual-WAN failover with sub-second switchover times. Test the failover mechanism during the pre-event setup, not during the event itself.

ROI and Business Impact

Event WiFi is increasingly recognised not just as a utility but as a data asset. Every attendee who connects to your event network and authenticates through a captive portal is providing first-party data — email address, demographic information, and behavioural data — that has significant commercial value for event organisers, venue operators, and sponsors.

Purple's WiFi Analytics platform quantifies this value directly. Post-event reports provide data on total unique connections, peak concurrent users, average session duration, portal conversion rates, and opt-in rates for marketing communications. For a 2,000-attendee conference with a 70% portal opt-in rate, that represents 1,400 new, consented marketing contacts captured in a single event — a cost per acquisition that is difficult to match through any other channel.

For venue operators in the hospitality sector, the analytics layer provides additional value through footfall analysis and dwell time mapping. Understanding which areas of a venue attract the most engagement — and for how long — informs layout decisions, F&B placement, and sponsor positioning for future events.

The ROI calculation for event WiFi investment should account for three categories of return: operational (reduced support costs from a well-designed network versus an ad-hoc one), commercial (first-party data capture and marketing opt-ins), and reputational (the brand value of a reliable, fast network that enhances the attendee experience). For large-scale events, the commercial return alone typically justifies the infrastructure investment within two or three events.