Wi-Fi

1Session at a glance

2Open (0:00, 4 minutes)

• Say what the hour is for: wireless is the one part of the network where the medium itself is shared, and almost every complaint comes from that one fact.
• Flag where this sits: last session of the course, and the close pulls the seven together.

3Concept (0:04, 22 minutes)

The shared medium (4 min)

• Contrast with the cable first, because it is the thing they already understand: a cable is private and both ends can send at once.
• Then radio: one channel, one transmitter at a time, everybody else waiting. A device listens, sends if the air is quiet, backs off if it is not.
• Derive the three consequences out loud rather than listing them. Capacity is shared, so what one device gets depends on what the others are doing. A weak client sends at a lower rate, so it holds the air longer for the same data, and that time is gone for everyone. Interference is not about your network at all, because anything else on the channel takes turns with you.
• Land the measure: airtime, not signal strength, is what predicts performance. Say it here and use it again in the fourth diagnostic question.

The bands (3 min)

• Walk the table in section 2 of the Learner Guide. Say the trade as one sentence and repeat it: lower frequency travels further and carries less, higher frequency carries more and travels less far.
• Then the consequence that produces tickets: a device at the edge of coverage falls back to 2.4 GHz because it is the only band still audible, so it connects and is slow rather than disconnecting. The user reports slow, and the answer is coverage.

Channels and overlap (4 min)

• Two access points on one channel share the air. Two on different channels do not. That is the whole reason channel planning exists.
• Then 2.4 GHz specifically: the channels overlap, so channel 2 interferes with 1 and 3, and only 1, 6 and 11 do not overlap at all. Any site on other numbers has access points interfering with each other.
• 5 GHz has many more channels and they do not overlap the same way, which is most of why it performs better.
• Then the radar rule, because it explains a fault with no other explanation: some 5 GHz channels are shared with weather and military radar, and an access point using one must listen and move off immediately if it hears any. The symptom is a whole site dropping at once with nothing in the console, near an airport or a weather station.
• Finish on transmit power, because somebody always raises it. Power changes what the access point shouts and nothing about what the phone can answer with. Turn it up and devices hold onto distant access points instead of moving, and your own access points interfere across a wider area. Coverage comes from more access points at lower power.

Channel width (3 min)

• Define it as the trade it is: wider carries more data for one device and consumes more of the air for everybody.
• Give the vendor's stated guidance for its own equipment: 20 MHz on 2.4 GHz, 80 MHz on 5 GHz, 160 or 320 MHz on 6 GHz.
• Then do the arithmetic for why 2.4 stays at 20, because the number is the argument. Three non-overlapping channels at that width. Double to 40 and one network covers two of the three, leaving one usable channel for the rest of the building and the building next door.
• Add the vendor's own instruction, which is a good habit to model: if a widened channel produces unreliable connections or higher latency, revert it, because the defaults favor reliable connections over peak speed.

Signal and noise (4 min)

• Explain the scale before any number: dBm, decibel-milliwatts, always negative for received signal, and closer to zero is stronger. -65 is stronger than -80.
• Say it is logarithmic and give the step: roughly every 3 dBm is a halving or doubling of power, so -68 is about half of -65. This is why differences that look small are not.
• Then noise, and warn them the sign reads the other way: more negative noise is better, because it means less of it.
• Work the comparison on the board, because it is the block's whole payload. A client at -65 with noise at -95 has a 30 dB gap and works. The same client at -65 with noise at -70 has a 5 dB gap and barely works. Identical signal readings, opposite outcomes.
• Say the rule that follows: quoting a signal figure without a noise figure explains nothing.

Roaming (2 min)

• The client decides, not the network. Access points cannot push a device off, and clients are reluctant.
• Give the symptom in the words a user would use: fine at the desk, unusable for the first few minutes after arriving in the meeting room, then fine again once it finally reconnects.
• Name the two settings that nudge it, a minimum signal level and advertising neighboring access points, and be clear that both influence the client's decision rather than commanding it. The real fix is usually coverage.

Networks and security (2 min)

• Each wireless network is mapped to a VLAN in the console, a virtual local area network. Restate what a VLAN is in one clause, since it has been four sessions.
• Then the fault that matters, which is the fifth check question: the switch port feeding the access point has to carry every VLAN it broadcasts. Miss one and clients connect to that network and get no address, while the others on the same access point are fine.
• Security in three lines: WPA2 Personal is a shared passphrase, WPA3 is current and the default pairs them so older devices still connect, and protected management frames stop a device being knocked off by a forged disconnect.
• End on the decision that actually matters at a client site: who holds the staff passphrase. One that every visitor has had for three years is a guest network with a misleading name.

4Show (0:26, 10 minutes)

1. The wireless networks. Open the site's list and read each network with the VLAN it is attached to. Same pairing as the wired networks, from the wireless side.
2. One access point's radios. Show which bands it is serving and the channel and width on each. Compare the widths to the vendor guidance out loud and say whether this site follows it.
3. The 2.4 GHz plan across the site. Read the channel each access point is using. If they are all on 1, 6 and 11, say so and say why that is not an accident. If they are not, leave it visible and make it a ticket rather than moving past it.
4. The client list. Find a client with a good signal and one with a poor one. Read both figures, then find the noise figure for that channel and do the subtraction out loud for each.
5. The airtime figure. Show the utilization on one access point and ask the room what it tells them about how much room is left. This is the number the session is trying to make them reach for.

5Do (0:36, 12 minutes)

1. Find yourself. Everyone finds their own device in the client list: signal, band, and which access point it is on.
2. Walk and look again. Walk to the far end of the building, wait, and check whether the device moved to a different access point or held on. Report what it did.
3. Check the plan. List the 2.4 GHz channels in use across the site and confirm no two access points within range of each other share one.
4. Trace one network to a port. Take one wireless network, follow it to the access point, then to the switch port feeding it, and confirm that port allows the VLAN it needs.

6Check (0:48, 7 minutes)

1. Why does one device with a weak signal slow down other devices that have a good signal on the same access point?

   Answer

   A weak client sends at a lower rate, so it occupies the air for longer to move the same amount of data. Only one device can transmit at a time, so the extra time it spends is time nobody else can use. The cost is airtime, not bandwidth, which is why the other devices' own signal strength does not protect them.

2. Why are only 1, 6 and 11 used on 2.4 GHz, and what happens to a site whose access points are on 3, 7 and 9?

   Answer

   The 2.4 GHz channels overlap each other, and 1, 6 and 11 are the only three that do not overlap at all. On 3, 7 and 9 every access point is partly on top of its neighbors, so they interfere with each other, and adding another access point makes the site worse rather than better.

3. A client shows a signal of -65 dBm at one site and -65 dBm at another, and the first works well while the second is unusable. Give the most likely explanation and the number you would ask for.

   Answer

   Noise. Ask for the noise figure on that channel. A client at -65 with noise at -95 has a 30 dB gap and works; the same client at -65 with noise at -70 has a 5 dB gap and barely works.

   The obvious answer is that two identical signal readings should behave identically, which is exactly why a signal figure quoted on its own is worth nothing. It describes one of the two numbers that decide the outcome.

4. A user's laptop is fine at their desk and unusable for the first few minutes in a meeting room at the other end of the floor. What is happening, and why is turning the transmit power up the wrong response?

   Answer

   The laptop is still attached to the access point by the desk and has not decided to move. The client makes that decision and clients are reluctant to make it.

   Turning the power up is the obvious response and it makes this worse. It changes only what the access point shouts, not what the laptop can answer with, so the distant access point becomes audible from further away and the laptop holds onto it longer. It also widens the area over which your own access points interfere with each other. The fix is coverage, and the settings available only nudge the client's own decision.

5. An access point broadcasts a staff network and a guest network. Guests connect to the right name, get prompted correctly, and then receive no address at all. Staff are fine on the same access point. Where is the fault?

   Answer

   The switch port feeding that access point. The guest VLAN is not allowed across it, so the requests for an address have nowhere to go. Wireless is working: the clients associated, which is why they connected and were prompted. One network failing while another on the same access point works is a port setting, not a radio problem.

6. The whole wireless network at a site near an airport drops for everyone at once, occasionally, with nothing in the console to explain it and no pattern to the timing. What would you suspect?

   Answer

   An access point on a 5 GHz channel shared with radar. It is required to listen for radar and move off immediately if it hears any, which takes every client on that radio with it. Near an airport or a weather station, the timing follows the radar rather than anything on the network, which is why there is no pattern visible from the console.

7. A user reports the Wi-Fi is slow. Give the first thing you would do, and say what its outcome rules out either way.

   Answer

   Put the same machine on a cable and repeat the same task. Still slow means the air is not the cause and the fault is somewhere the whole site shares, the wired network, a server or the internet connection. Fast on the cable means it is wireless, and only then is it worth reading signal, noise and airtime.

7Close (0:55, 5 minutes)

Work after this session

• The practice steps in section 9 of the Learner Guide, on a real site.
• The Network+ companion, chapter 3, wireless standards and site surveys, and chapter 11, wireless troubleshooting.

Close the course

Say this rather than summarizing the session, because it is the last one and the through-line is the point. The same habit has been applied at every layer for seven sessions: test in a fixed order, stop at the first thing that fails, and let what failed choose the next tool. The layers gave the order. Addressing decided who counts as a neighbor. The switch and the VLAN decided who hears whom. The router and its table decided the path. The two services decided whether a device could take part at all. The firewall decided what was permitted. The air decided how much of it was available. A fault lives at one of those, and naming which one is most of the work.

Open items to settle

• Any site found during the session to be on 2.4 GHz channels other than 1, 6 and 11, raised as a ticket.
• Which sites have a staff passphrase that has been handed to visitors, and whether those get changed or get a real guest network.
• Whether a short written triage sequence comes out of this course, so that the order the sessions taught is the order on the page when somebody is working a ticket.

8Sources

• IEEE, 802.11, Wireless LAN Medium Access Control and Physical Layer Specifications, for the shared medium, taking turns, and the bands and channels.
• Ubiquiti, Maximizing Wireless Speeds, for the stated channel width guidance of 20 MHz on 2.4 GHz, 80 MHz on 5 GHz and 160 or 320 MHz on 6 GHz, for preferring channels with less interference, and for reverting to defaults when connection quality degrades.
• Ubiquiti, Creating UniFi WiFi SSIDs, for attaching a wireless network to a VLAN, the requirement that upstream switch ports allow that VLAN, and the default security of WPA2 and WPA3 with optional protected management frames.
• Ubiquiti, WiFi Troubleshooting Guide.
• Ubiquiti, Optimizing WiFi Connectivity and Reducing Latency.
• Wi-Fi Alliance, Wi-Fi security, for WPA2, WPA3 and protected management frames.
• Kodi A. Cochran, CompTIA Network+ (N10-009) Certification Companion (Apress, 2026), chapter 3, wireless standards and site surveys, and chapter 11, wireless troubleshooting.

9After the session