11.2 High-Density, Location Services and Guest Access
Key Takeaways
- High-density WLAN design uses more APs, lower transmit power, smaller cells, a preference for 5 GHz and 6 GHz, and often directional antennas so each BSS serves fewer stations with less co-channel contention
- Raising AP power to fix a dense room enlarges cells, increases co-channel contention, and is the classic density trap
- Location services and RTLS need a denser AP grid than coverage-only design, an attenuation or calibration map, and a method such as RSSI fingerprinting; angle-of-arrival arrays are more accurate at CWNA awareness depth
- Guest and BYOD designs isolate traffic with VLANs and client isolation, use a captive portal and onboarding path, and keep guests off employee subnets
- 2.4 GHz cannot scale AP count the way 5 GHz and 6 GHz can, because it offers only a few non-overlapping 20 MHz channels in typical regulatory domains
Northbridge’s problems are no longer only “voice versus data.” Tuesday grand rounds pack 200 phones into a lecture hall. Biomed wants room-level locations for infusion pumps. Visitors need Internet without a path into EHR servers, and clinicians bring personal phones (BYOD) that must not share the badge SSID. CWNA-109 Objective 4.3.2 groups those specific applications: location services, high density, and guest access / BYOD. The RF habits that help a quiet clinic office will fail all three if you only turn power up.
This OpenExamPrep section is independent CWNA-109 study material by OpenExamPrep. It does not claim CWNP approval, partnership, or exact equivalence with CWNP training.
High density: shrink the cell, do not shout
A high-density space is any room where client count per square meter (and offered load) exceeds what one or two omnidirectional cells can airtime-budget. Lecture halls, waiting rooms during a portal outage, conference keynotes, and cafeterias at lunch are the usual WLAN examples. The design pattern is consistent:
- More APs so each BSS serves a slice of the crowd, not the whole room.
- Lower transmit power so the cell edge is a few meters, not half the building.
- Smaller cells as the result of (1) and (2), plus disabled low basic rates so the cell does not stretch at 1 or 6 Mbps.
- 5 GHz and 6 GHz preference because those bands have far more non-overlapping channels than 2.4 GHz.
- Directional antennas (patch, sector, wall-plate, downtilt ceiling) that put energy into the seating and not into the AP across the atrium on the same channel.
2.4 GHz cannot follow the same AP-count curve. In typical regulatory domains you have about three non-overlapping 20 MHz channels (1, 6, 12/13 depending on the region). If you deploy twenty 2.4 GHz radios in one hall, most of them land on the same three channels and contend. High-density plans therefore often reduce 2.4 GHz AP count, disable 2.4 GHz in the hall, or use 2.4 GHz only for a handful of IoT SSIDs. Capable phones should be steered or allowed to prefer 5 GHz and 6 GHz (band steering is a proprietary feature family from Objective 4.4; the design idea is still “put load where channels exist”).
Channel width is a reuse decision. 80 MHz (and 160 MHz where it exists) is wonderful in a sparse office and hungry in a packed hall: each wide channel consumes more spectrum, leaving fewer unique channels to assign. Many high-density designs stay on 20 MHz (sometimes 40 MHz) in 5 GHz so they can paint a reuse map. Wi-Fi 6 OFDMA helps efficiency inside a BSS; it does not create extra 2.4 GHz channels.
| Lever | High-density setting | Why it helps |
|---|---|---|
| AP count | Higher | Splits clients across BSSIDs |
| AP power | Lower | Shrinks the contention domain |
| Antenna | Directional / downtilt | Keeps energy in the intended seating |
| Band | 5 GHz and 6 GHz first | Many more reuse channels than 2.4 GHz |
| Channel width | Often 20 MHz (sometimes 40) | More distinct channels for reuse |
| Low rates | Disabled after validation | Stops the cell from stretching |
| 2.4 GHz radios | Few or off in the hall | Avoids three-channel CCC collapse |
The trap: raising power to fix density
The help desk instinct is “the room is busy, turn the APs up.” Higher power makes each AP audible across more seats. Clients at the back stick to a far BSSID. Adjacent APs on the same channel hear each other and defer. SNR can look acceptable while retries explode because the noise-plus-contention floor rose. Fix density with more, quieter, better-aimed APs, not with maximum EIRP. If a corner is actually a coverage hole, add an AP or aim an antenna — do not solve a 200-client airtime problem with a 3 dB power bump.
Secondary high-density tools (awareness, not a full controller course): client load balancing, BSS coloring / spatial reuse on 802.11ax where supported, multicast-to-unicast for the presenter’s slide stream, and disabling legacy rates so a single 802.11b probe does not tax the hall (section 11.1). None of those replace a reuse map.
Worked density sketch (illustrative)
Grand rounds: 200 phones, mostly 5 GHz. If one AP holds all 200, even modest 150 kbps signaling plus beacons and retries will saturate airtime. Ten APs at reduced power, 20 MHz channels, and a reuse plan that does not put same-channel neighbors facing each other, cuts the per-BSS count toward 20. That is still busy, but it is a different problem from one omni shouting at 200 radios. Exact AP count is a predictive-design and survey job (later CWDP-style work). CWNA-109 wants the direction of the knobs.
Location services and RTLS
Real-time location systems (RTLS) and indoor positioning estimate where a tag, phone, or asset is. Coverage-only AP placement — APs down the hallway centerline — is a poor location geometry. You need the device heard by multiple APs from different angles, including the perimeter of the floor, and you need a model of how walls eat RF.
Common building blocks at CWNA awareness depth:
- Denser AP grid than coverage-only, including perimeter APs, so lateration or fingerprinting has geometry.
- Attenuation maps / calibration walks: the system records how this building actually attenuates, not a free-space myth.
- RSSI fingerprinting: during a survey you record a vector of RSSI values from visible APs at known points (the radio map). At run time the live RSSI vector is matched to that map. Furniture moves and a missing AP degrade the match.
- RSSI lateration: convert RSSI to a rough distance from several APs and intersect. RSSI versus distance is noisy indoors, so room-level claims need humility.
- Angle of Arrival (AoA): an AP with an antenna array estimates the direction of the transmitter. Combined with another AoA or a distance estimate, accuracy can beat fingerprinting. CWNA-109 expects awareness that AoA exists and why arrays help, not vendor calibration menus.
- Tags versus associated clients: Wi-Fi tags, RFID, and BLE beacons often blink without a full user association. Location is not only “associated STA on the EHR SSID.”
| Method | What it uses | Typical CWNA takeaway |
|---|---|---|
| Coverage heatmap only | One RSSI vs location | Finds holes; does not give room-level RTLS |
| RSSI lateration | Distance-from-RSSI to several APs | Coarse; walls break the distance model |
| RSSI fingerprinting | Calibrated radio map | Common software approach; needs upkeep |
| AoA | Antenna array angle | Higher accuracy potential; special hardware |
Trap: promising room-level pump tracking on a coverage design that placed APs only in corridors. Another trap: turning all AP power to maximum “so location can hear everything,” which smears fingerprints and increases CCC for the data network that shares those radios.
Northbridge should treat location as its own RF program: extra APs, calibration, and a tag technology choice — then confirm the same radios can still meet voice and density goals.
Guest access and BYOD
Guest means a person and device you do not fully control. BYOD means a personal device that may need more than raw Internet (email, clinical apps) but still is not a corporate-owned endpoint. Both are Objective 4.3.2 applications. Security mechanisms (WPA2/WPA3, 802.1X, captive portal crypto) are Domain 5; here the design is isolation, onboarding, and capacity.
Isolation. Put guests on a separate SSID and VLAN. Enable client isolation (peer-to-peer block) so two guests in the waiting room cannot attack each other. Filter guest traffic so it cannot reach EHR, voice, or management subnets. ACLs and firewall roles (Objective 4.5) enforce that segmentation on the wired side.
Captive portal awareness. Many guest networks intercept the first HTTP request or use a DNS walled garden, then present a browser page for terms of use, a voucher, or a sponsor approval. DHCP and DNS must work before the portal, or the phone never loads the page. Encrypted HTTPS “captive portal detection” on modern phones is fragile if the portal is HTTP-only or if DNS is blackholed too aggressively. Design the allow-list (DHCP, DNS, portal URL, maybe a CDN) explicitly.
Onboarding. Guests may get a rotating PSK, a voucher, SMS, or self-registration. Employees’ personal phones (BYOD) should not share the voice-badge PSK. Typical patterns are a dedicated BYOD SSID with 802.1X and an onboarding portal that installs a unique credential, or MDM/NAC posture at awareness level. Do not “just add personal phones” to the badge SSID because it is convenient.
Capacity and policy. Guest load is bursty (200 people after a lecture). Rate-limit the guest SSID so it cannot starve EHR. Put guest Internet on a path with enough WAN capacity (Objective 4.5.7). Log and expire accounts. Legal acceptable-use language belongs on the portal; CWNA cares that the network path matches the policy: isolated, onboarded, and rate-limited.
| Design item | Guest | Corporate BYOD | Corporate managed |
|---|---|---|---|
| VLAN | Isolated guest | Often separate BYOD | Employee / voice / IoT |
| Client isolation | On | Often on | Usually off (peer apps) |
| Auth / onboarding | Portal, voucher, PSK rotation | 802.1X / onboarding profile | 802.1X machine or user |
| Internal access | Denied | App-specific | Full as policy allows |
| Rate limit | Common | Sometimes | Rare |
Trap: bridging the guest SSID onto the employee VLAN “temporarily.” Another trap: a portal that requires DNS while the ACL blocks UDP/53, so every phone shows “connected, no Internet” and the ticket is filed as an RF outage.
On the exam
- High density = more APs, lower power, smaller cells, 5/6 GHz, directional antennas.
- Do not raise power to fix a packed room.
- Location = more APs + attenuation/calibration map + fingerprinting or AoA, not a coverage heatmap alone.
- Guest/BYOD = isolation, captive portal awareness, onboarding, not a shared employee PSK.
Key Takeaways
- Density is a reuse problem; power is the wrong first knob.
- RTLS geometry is different from coverage geometry.
- Guest convenience is not a reason to skip a VLAN.
Grand rounds put about 200 phones in Northbridge’s lecture hall. Which WLAN design direction matches Objective 4.3.2 high-density practice?
After the hall feels slow, a technician doubles AP transmit power. What is the likely result, and what is the better density response?
Biomed wants room-level infusion-pump tracking on the same floor as guest Wi-Fi. Which location and guest statements belong together for Objective 4.3.2?