<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/"><channel><title>Wireless 101: how Wi-Fi signal really travels · Network Doodles</title><link>https://networkdoodles.com/lessons/wireless-101/</link><description>Start from zero. Radio waves, dB and dBm, the 2.4, 5 and 6 GHz bands, antenna patterns in 3D, what walls and distance do to a signal, where to put access points, and how a laptop picks one and roams. Interactive 3D views you can turn and point at, and nothing to install.</description><language>en-us</language><managingEditor>David Henderson</managingEditor><lastBuildDate>Mon, 16 Nov 2026 09:00:00 -0500</lastBuildDate><atom:link href="https://networkdoodles.com/lessons/wireless-101/index.xml" rel="self" type="application/rss+xml"/><item><title>Wi-Fi security: WPA2, WPA3 and 802.1X</title><link>https://networkdoodles.com/lessons/wireless-101/task-7-1/</link><pubDate>Mon, 16 Nov 2026 09:00:00 -0500</pubDate><dc:creator>David Henderson</dc:creator><guid>https://networkdoodles.com/lessons/wireless-101/task-7-1/</guid><description>Open, OWE, WPA2 and WPA3 Personal and Enterprise compared, the four-way handshake step by step, and 802.1X with RADIUS and EAP-TLS. Beginner lesson.</description><content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[
<section class="lab-stage lab-stage-explain" id="explain">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">1</span>
    <span class="lab-stage-kind mono">explain</span><h2>How does Wi-Fi keep traffic private?</h2>
  </header>
  <div class="lab-stage-body">
    <h3 id="a-secret-knock">A secret knock</h3>
<p>A club has a secret knock. Anyone who knows it gets in, and once inside, members talk in a code that changes every
night. A cable network has walls round it; Wi-Fi doesn&rsquo;t. Anyone nearby with a radio can hear every frame, so
Wi-Fi security has two jobs:</p>
<ol>
<li><strong>Who gets in</strong>: proving the device is allowed (authentication).</li>
<li><strong>Privacy</strong>: encrypting every frame so eavesdroppers hear only noise, with keys fresh for each connection.</li>
</ol>
<p class="lab-recap"><span class="mono">in one sentence</span> Anyone nearby can hear Wi-Fi, so it has to check who&rsquo;s joining and encrypt everything.</p>
<h3 id="the-options">The options</h3>
<table>
  <thead>
      <tr>
          <th>Security</th>
          <th>Who gets in</th>
          <th>Privacy</th>
          <th>Use it for</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td><strong>Open</strong></td>
          <td>anyone</td>
          <td><strong>none</strong>: frames go unencrypted</td>
          <td>nothing new; only behind a captive portal you trust</td>
      </tr>
      <tr>
          <td><strong>Enhanced Open</strong> (OWE)</td>
          <td>anyone, no password</td>
          <td>yes: each device gets its own keys</td>
          <td>cafés, guest networks</td>
      </tr>
      <tr>
          <td>WEP, WPA (TKIP)</td>
          <td>—</td>
          <td>broken</td>
          <td>never: replace any device that needs them</td>
      </tr>
      <tr>
          <td><strong>WPA2-Personal</strong></td>
          <td>anyone with the passphrase</td>
          <td>yes</td>
          <td>homes, small offices, older devices</td>
      </tr>
      <tr>
          <td><strong>WPA3-Personal</strong> (SAE)</td>
          <td>anyone with the passphrase</td>
          <td>yes, and a captured handshake can&rsquo;t be cracked offline</td>
          <td>homes and small offices, if devices support it</td>
      </tr>
      <tr>
          <td><strong>WPA2/WPA3-Enterprise</strong> (802.1X)</td>
          <td>each person or device, checked by a server</td>
          <td>yes, with keys unique to each user</td>
          <td>offices, schools, anywhere people come and go</td>
      </tr>
  </tbody>
</table>
<p><strong>Personal</strong> means one shared passphrase. <strong>Enterprise</strong> means each user or device has its own credentials,
checked by a <strong>RADIUS</strong> server.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Personal: one shared passphrase. Enterprise: everyone their own credentials, checked by RADIUS.</p>
<h3 id="the-four-way-handshake">The four-way handshake</h3>
<p>Every WPA2 and WPA3 connection ends with the same four frames, the <strong>four-way handshake</strong> from lesson 6. Before it
starts, both the laptop and the access point already share a secret, the <strong>PMK</strong> (<em>pairwise master key</em>):</p>
<ul>
<li>on <strong>WPA2-Personal</strong>, it&rsquo;s worked out from the passphrase and the network name;</li>
<li>on <strong>WPA3-Personal</strong>, it comes from an exchange called <strong>SAE</strong> (<em>simultaneous authentication of equals</em>), fresh for
each connection;</li>
<li>on <strong>Enterprise</strong>, the RADIUS server hands it to the access point after checking the user.</li>
</ul>
<p>The handshake never sends the PMK. Instead:</p>
<ol>
<li>The access point sends a random number, the <strong>ANonce</strong> (<em>nonce</em> means <em>number used once</em>).</li>
<li>The laptop sends its own random number, the <strong>SNonce</strong>, plus a <strong>MIC</strong> (<em>message integrity check</em>): a stamp only
someone who knows the PMK could make.</li>
<li>The access point checks the MIC, and sends the <strong>group key</strong> for broadcasts, encrypted, with its own MIC.</li>
<li>The laptop confirms.</li>
</ol>
<p>Both sides mix the PMK, both nonces and both MAC addresses to get the <strong>PTK</strong> (<em>pairwise transient key</em>), the key that
encrypts this one connection. New nonces next time means new keys next time.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Two random numbers plus the shared secret make fresh keys, and the secret itself never crosses the air.</p>
<h3 id="why-wpa3-matters-for-passphrases">Why WPA3 matters for passphrases</h3>
<p>On WPA2-Personal, someone who records a handshake has everything they need to test passphrase guesses on their
own computer, billions of them, without ever touching the network again. A short or common passphrase falls fast.
WPA3&rsquo;s <strong>SAE</strong> exchange is built so a recording doesn&rsquo;t help: each guess needs a live attempt against the access
point. WPA3 also requires <strong>protected management frames</strong>, so an attacker can&rsquo;t forge the &ldquo;disconnect&rdquo; frames used to
kick people off.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> WPA3 stops offline guessing of the passphrase, and protects management frames.</p>
<h3 id="enterprise-8021x-and-radius">Enterprise: 802.1X and RADIUS</h3>
<p>In an office, a shared passphrase is a problem: when someone leaves, everyone&rsquo;s passphrase has to change. <strong>802.1X</strong>
gives each person or device their own credentials. Three roles:</p>
<ul>
<li>the <strong>supplicant</strong>: the laptop;</li>
<li>the <strong>authenticator</strong>: the access point, which relays and enforces;</li>
<li>the <strong>authentication server</strong>: a <strong>RADIUS</strong> server, which checks the credentials (RADIUS listens on UDP port 1812).</li>
</ul>
<p>They talk <strong>EAP</strong> (<em>extensible authentication protocol</em>), with a choice of methods:</p>
<ul>
<li><strong>EAP-TLS</strong>: certificates on both the laptop and the server. No passwords to guess or phish. The strongest common
choice.</li>
<li><strong>PEAP</strong>: the server shows a certificate, a protected tunnel is built, and the user&rsquo;s username and password go
inside it. Easy to roll out, but only safe if laptops are set to <strong>check the server&rsquo;s certificate</strong>.</li>
</ul>
<p>When the server says yes, it also sends the access point the PMK for that user, and can tell it which <strong>VLAN</strong> to put
them in, so staff and contractors can share one network name but land in different networks.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> 802.1X: the laptop proves itself to RADIUS through the access point, and gets keys of its own.</p>

  </div>
</section>


<section class="lab-stage lab-stage-show" id="show">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">2</span>
    <span class="lab-stage-kind mono">show</span><h2>Two exchanges to watch</h2>
  </header>
  <div class="lab-stage-body">
    <p>The first animation is the four-way handshake on a WPA2-Personal network called <em>home</em>: the same four frames you saw
at the end of joining in lesson 6, opened up.</p>
<p>The second is an Enterprise network called <em>office</em>: a laptop proving who it is to a RADIUS server with EAP-TLS,
through the access point, before the handshake. The real exchange has about a dozen frames; the animation shows the
parts that matter.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> One handshake on a home network; one 802.1X login, then the handshake, on an office network.</p>

  </div>
</section>


<section class="lab-stage lab-stage-follow" id="follow">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">3</span>
    <span class="lab-stage-kind mono">follow along</span><h2>Watch the keys being made</h2>
  </header>
  <div class="lab-stage-body">
    <figure class="netsim" data-netsim tabindex="0" aria-label="The WPA2 four-way handshake between a laptop and an access point: random numbers each way, a message integrity check proving both know the shared key, the group key delivered, and encrypted data after. Use the buttons, or the left and right arrow keys, to step through it.">
  <svg class="ns-svg" role="img" aria-label="The WPA2 four-way handshake between a laptop and an access point: random numbers each way, a message integrity check proving both know the shared key, the group key delivered, and encrypted data after"></svg>
  <div class="ns-ctl">
    <button class="icon-btn" type="button" data-ns="restart" aria-label="Back to the first step" title="Start again"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M6 5v14M19 5l-9 7 9 7z"/></svg>
</button>
    <button class="icon-btn" type="button" data-ns="prev" aria-label="Previous step" title="Previous step"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M19 12H5M11 6l-6 6 6 6"/></svg>
</button>
    <button class="btn sm primary ns-play" type="button" data-ns="play" aria-pressed="false"><span data-ns-play>Play</span></button>
    <button class="icon-btn" type="button" data-ns="next" aria-label="Next step" title="Next step"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M5 12h14M13 6l6 6-6 6"/></svg>
</button>
    <button class="icon-btn" type="button" data-ns="again" aria-label="Play this step again" title="Play this step again"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M4 12a8 8 0 1 0 2.5-5.8M4 4v4h4"/></svg>
</button>
    <span class="ns-count mono" data-ns-count></span>
    <button class="btn sm ns-speed" type="button" data-ns="speed" aria-pressed="false">Slow: off</button>
  </div>
  <div class="ns-dots" data-ns-dots role="group" aria-label="Jump to a step"></div>
  <figcaption class="ns-caption" data-ns-caption aria-live="polite"></figcaption>
  <div class="ns-panels">
    <div class="ns-codebox" data-ns-code hidden></div>
    <div class="ns-frame" data-ns-frame hidden></div>
    <div class="ns-tablebox" data-ns-table hidden></div>
  </div>
  <script type="application/json">{"h":300,"links":[["pc","ap",{"wifi":true}]],"narrow":{"at":{"ap":[180,90],"pc":[180,330]},"h":420,"w":360},"nodes":{"ap":{"kind":"wifi","label":"Access point","sub":"network: home","x":620,"y":150},"pc":{"kind":"laptop","label":"Laptop","sub":"knows the passphrase","x":140,"y":150}},"steps":[{"hot":["pc","ap"],"say":{"ap":"same PMK","pc":"PMK from\npassphrase + 'home'"},"table":{"cols":["","laptop","access point"],"rows":[["PMK","yes","yes"],["ANonce","","yes"],["SNonce","yes",""],["PTK","",""]],"title":"What each side knows"},"text":"On WPA2-Personal, both sides work out the **PMK** from the passphrase and the network name. It never crosses the air. The laptop has already associated (lesson 6), but it can't send data yet.","title":"Both start with the same secret."},{"frame":{"layers":[{"fields":[["to","laptop"],["from","access point"]],"l":2,"name":"802.11 data"},{"fields":[["message","1 of 4",true],["ANonce","random, 32 bytes",true],["MIC","none yet"]],"l":3,"name":"EAPOL-Key"}],"title":"Inside **message 1**"},"hot":["ap"],"links":[["pc","ap"]],"packets":[{"c":"violet","label":"1/4 ANonce","path":["ap","pc"]}],"table":{"cols":["","laptop","access point"],"hl":3,"rows":[["PMK","yes","yes"],["ANonce","yes","yes"],["SNonce","yes",""],["PTK","yes: worked out now",""]],"title":"What each side knows"},"text":"The access point sends a random number, the **ANonce**. Anyone can hear it; it's only useful together with the PMK.","title":"1 of 4: the ANonce."},{"frame":{"layers":[{"fields":[["message","2 of 4",true],["SNonce","random, 32 bytes",true],["MIC","made with the new key",true]],"l":3,"name":"EAPOL-Key"}],"title":"Inside **message 2**"},"hot":["pc"],"links":[["pc","ap"]],"packets":[{"c":"violet","label":"2/4 SNonce + MIC","path":["pc","ap"]}],"table":{"cols":["","laptop","access point"],"hl":3,"rows":[["PMK","yes","yes"],["ANonce","yes","yes"],["SNonce","yes","yes"],["PTK","yes","yes: worked out now"]],"title":"What each side knows"},"text":"The laptop mixes the PMK, both nonces and both MAC addresses into the **PTK**, its key for this connection. It sends its **SNonce** and a **MIC**: a stamp made with the new key, which only something that knows the PMK could make.","title":"2 of 4: the SNonce and a proof."},{"frame":{"layers":[{"fields":[["message","3 of 4",true],["group key","encrypted with the PTK",true],["MIC","made with the PTK",true],["install","yes"]],"l":3,"name":"EAPOL-Key"}],"title":"Inside **message 3**"},"hot":["ap"],"links":[["pc","ap"]],"packets":[{"c":"violet","label":"3/4 group key + MIC","path":["ap","pc"]}],"say":{"ap":"MIC checks out"},"text":"The access point works out the same PTK and checks the MIC. It matches, so the laptop knew the secret. It sends the **group key** (for broadcasts to everyone on the access point), encrypted with the PTK, and its own MIC.","title":"3 of 4: checked, and the group key."},{"frame":{"layers":[{"fields":[["message","4 of 4",true],["MIC","made with the PTK"],["meaning","keys installed, let's go",true]],"l":3,"name":"EAPOL-Key"}],"title":"Inside **message 4**"},"good":["pc","ap"],"hot":["pc","ap"],"links":[["pc","ap"]],"packets":[{"c":"violet","label":"4/4 ACK","path":["pc","ap"]}],"set":{"pc":"keys installed"},"text":"The laptop confirms. Both install their keys. From now on every data frame is encrypted (with AES, in WPA2 and WPA3).","title":"4 of 4: done."},{"frame":{"layers":[{"fields":[["to","access point"],["from","laptop"],["protected","yes",true]],"l":2,"name":"802.11 data"},{"fields":[["contents","AES-encrypted with the PTK",true]],"l":3,"name":"payload"}],"title":"Inside a **data frame**"},"hot":["pc","ap"],"links":[["pc","ap"]],"packets":[{"c":"green","label":"DATA (encrypted)","path":["pc","ap"]},{"at":1,"c":"green","label":"DATA (encrypted)","path":["ap","pc"]}],"text":"Now DHCP and everything else flows, encrypted. Someone nearby hears the frames but can't read them. With WPA2-Personal, though, if they **know the passphrase** and recorded these four messages, they could work out this PTK too: one more reason to prefer WPA3.","title":"Encrypted data."}],"w":760}</script>
</figure>
<script src="/js/netsim.c40a98d103487ae5866edfa12abb4afc1d3aa7ad2d12fed4fe925fe00b588106.js" integrity="sha256-xAqY0QNIeuWGbt&#43;hKrtK/B06p60tEv7U/pJf4AtYgQY=" defer></script>
<figure class="netsim" data-netsim tabindex="0" aria-label="A laptop logs in to an office network with 802.1X and EAP-TLS: the access point relays EAP between the laptop and a RADIUS server, the server checks the laptop&#39;s certificate, sends Access-Accept with the keys and a VLAN, and the four-way handshake follows. Use the buttons, or the left and right arrow keys, to step through it.">
  <svg class="ns-svg" role="img" aria-label="A laptop logs in to an office network with 802.1X and EAP-TLS: the access point relays EAP between the laptop and a RADIUS server, the server checks the laptop&#39;s certificate, sends Access-Accept with the keys and a VLAN, and the four-way handshake follows"></svg>
  <div class="ns-ctl">
    <button class="icon-btn" type="button" data-ns="restart" aria-label="Back to the first step" title="Start again"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M6 5v14M19 5l-9 7 9 7z"/></svg>
</button>
    <button class="icon-btn" type="button" data-ns="prev" aria-label="Previous step" title="Previous step"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M19 12H5M11 6l-6 6 6 6"/></svg>
</button>
    <button class="btn sm primary ns-play" type="button" data-ns="play" aria-pressed="false"><span data-ns-play>Play</span></button>
    <button class="icon-btn" type="button" data-ns="next" aria-label="Next step" title="Next step"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M5 12h14M13 6l6 6-6 6"/></svg>
</button>
    <button class="icon-btn" type="button" data-ns="again" aria-label="Play this step again" title="Play this step again"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M4 12a8 8 0 1 0 2.5-5.8M4 4v4h4"/></svg>
</button>
    <span class="ns-count mono" data-ns-count></span>
    <button class="btn sm ns-speed" type="button" data-ns="speed" aria-pressed="false">Slow: off</button>
  </div>
  <div class="ns-dots" data-ns-dots role="group" aria-label="Jump to a step"></div>
  <figcaption class="ns-caption" data-ns-caption aria-live="polite"></figcaption>
  <div class="ns-panels">
    <div class="ns-codebox" data-ns-code hidden></div>
    <div class="ns-frame" data-ns-frame hidden></div>
    <div class="ns-tablebox" data-ns-table hidden></div>
  </div>
  <script type="application/json">{"h":300,"links":[["pc","ap",{"wifi":true}],["ap","rad"]],"narrow":{"at":{"ap":[180,280],"pc":[180,480],"rad":[180,80]},"h":560,"w":360},"nodes":{"ap":{"kind":"wifi","label":"Access point","sub":"authenticator","x":390,"y":160},"pc":{"kind":"laptop","label":"Laptop","sub":"supplicant","sub2":"has a certificate","x":110,"y":160},"rad":{"kind":"server","label":"RADIUS server","sub":"authentication server","x":670,"y":160}},"steps":[{"hot":["ap"],"links":[["pc","ap"]],"say":{"ap":"EAP only,\nuntil RADIUS says yes"},"text":"The laptop has joined the access point (lesson 6), but on an 802.1X network the access point lets nothing through except the login conversation: **EAP**.","title":"Associated, but locked."},{"frame":{"layers":[{"fields":[["dst port","1812 (RADIUS)",true]],"l":4,"name":"UDP"},{"fields":[["code","Access-Request",true],["user name","laptop-042@office",true],["EAP message","Response / Identity"],["calling station","the laptop's MAC"]],"l":7,"name":"RADIUS"}],"title":"Inside the **Access-Request**"},"hot":["pc","ap","rad"],"links":[["pc","ap"],["ap","rad"]],"packets":[{"c":"amber","label":"EAP identity?","path":["ap","pc"]},{"at":1,"c":"amber","label":"identity","path":["pc","ap","rad"]}],"text":"The access point asks for an identity. The laptop answers with its name, say **laptop-042@office**. The access point wraps it in a RADIUS **Access-Request** and sends it to the server on UDP port 1812.","title":"Who are you?"},{"frame":{"layers":[{"fields":[["server shows","its certificate",true],["laptop checks","issued by the office's own CA"],["laptop shows","its certificate",true],["server checks","valid, not revoked"]],"l":7,"name":"EAP-TLS"}],"title":"Inside the **EAP-TLS** exchange"},"hot":["pc","rad"],"links":[["pc","ap"],["ap","rad"]],"packets":[{"c":"cyan","label":"server cert","path":["rad","ap","pc"]},{"at":2,"c":"cyan","label":"laptop cert","path":["pc","ap","rad"]}],"say":{"pc":"Is that really\nthe office server?","rad":"Is that a laptop\nwe issued?"},"text":"With **EAP-TLS**, the server shows its certificate and the laptop checks it is talking to the real server. Then the laptop shows its own certificate. The access point just relays: it can't read or change any of it.","title":"Certificates, both ways."},{"frame":{"layers":[{"fields":[["code","Access-Accept",true],["key material","for this laptop's PMK (encrypted)",true],["VLAN","staff (20)",true],["session timeout","8 hours"]],"l":7,"name":"RADIUS"}],"title":"Inside the **Access-Accept**"},"hot":["rad","ap"],"links":[["ap","rad"],["pc","ap"]],"packets":[{"c":"green","label":"Access-Accept","path":["rad","ap"]},{"at":1,"c":"green","label":"EAP-Success","path":["ap","pc"]}],"text":"The certificate is valid and not revoked, so the server sends **Access-Accept**. Inside it: key material the access point turns into this laptop's own **PMK**, and a **VLAN** to put it in. The laptop gets **EAP-Success**, and works out the same PMK itself from the TLS exchange.","title":"Access-Accept, with keys."},{"frame":null,"good":["pc"],"hot":["pc","ap"],"links":[["pc","ap"]],"packets":[{"c":"violet","label":"1/4","path":["ap","pc"]},{"at":1,"c":"violet","label":"2/4","path":["pc","ap"]},{"at":2,"c":"violet","label":"3/4","path":["ap","pc"]},{"at":3,"c":"violet","label":"4/4","path":["pc","ap"]}],"set":{"pc":"staff VLAN · own keys"},"text":"Same four frames as before, using this laptop's own PMK. Every user ends up with different keys: one laptop leaving the company means revoking **one** certificate, not changing a passphrase for everyone.","title":"The four-way handshake."}],"w":780}</script>
</figure>
<script src="/js/netsim.c40a98d103487ae5866edfa12abb4afc1d3aa7ad2d12fed4fe925fe00b588106.js" integrity="sha256-xAqY0QNIeuWGbt&#43;hKrtK/B06p60tEv7U/pJf4AtYgQY=" defer></script>
<h3 id="try-it-on-your-own-computer-optional">Try it on your own computer (optional)</h3>
<ul>
<li><strong>Mac</strong>: hold <kbd>Option</kbd> and click the Wi-Fi icon. <strong>Security</strong> shows what your network uses, such as
<em>WPA2 Personal</em> or <em>WPA3 Personal</em>.</li>
<li><strong>Windows</strong>: <code>netsh wlan show interfaces</code> in <em>Command Prompt</em>. Look at <strong>Authentication</strong> (for example
<em>WPA2-Personal</em>) and <strong>Cipher</strong> (<em>CCMP</em> is AES).</li>
<li><strong>Linux</strong>: <code>nmcli -f SSID,SECURITY device wifi list</code> lists nearby networks and their security.</li>
</ul>
<p class="lab-recap"><span class="mono">in one sentence</span> The handshake makes fresh keys from a shared secret; 802.1X gives every user their own secret first.</p>

  </div>
</section>


<section class="lab-stage lab-stage-challenge" id="challenge">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">4</span>
    <span class="lab-stage-kind mono">challenge</span><h2>Pick the security</h2>
  </header>
  <div class="lab-stage-body">
    <p>Choose security for each, and say why:</p>
<ol>
<li>A family home. Everyone&rsquo;s phones and laptops are recent; one old smart plug only does WPA2.</li>
<li>A café that wants free Wi-Fi with no password to hand out.</li>
<li>A 200-person office, staff laptops managed by IT.</li>
<li>The same office&rsquo;s visitors.</li>
<li>The office&rsquo;s 40 printers and badge readers, which don&rsquo;t support 802.1X.</li>
</ol>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>Ask, for each: who should get in, can each person have their own credentials, and what can the devices actually do?</p>
</div>
</details>
<details class="lab-solution">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg>
 <span>show the solution</span></summary>
  <div class="lab-solution-body">
    <ol>
<li><strong>WPA3-Personal in transition mode</strong> (WPA2 and WPA3 on one network name), with a long passphrase. Recent devices
use WPA3; the smart plug falls back to WPA2. Or put the plug on its own WPA2 network.</li>
<li><strong>Enhanced Open (OWE)</strong>, ideally with an open fallback for the few old devices that can&rsquo;t do it: no password, but
customers can&rsquo;t read each other&rsquo;s traffic.</li>
<li><strong>WPA3-Enterprise with EAP-TLS</strong>: certificates IT already manages, no passwords to phish, and one laptop&rsquo;s
certificate can be revoked on its own. PEAP is the easier fallback, as long as laptops check the server
certificate.</li>
<li>A separate <strong>guest network</strong> in its own VLAN, with OWE or a passphrase that changes regularly, often behind a
captive portal (a sign-in or terms page), blocked from the staff network.</li>
<li>Their own network name and VLAN, <strong>WPA2- or WPA3-Personal</strong> with a long passphrase (some systems can give each
device its own passphrase), and firewall rules so they reach only the servers they need.</li>
</ol>
  </div>
</details>

  </div>
</section>

<p>That&rsquo;s the last lesson. Back to <a href="/lessons/wireless-101/task-6-1/">lesson 6</a> for joining and roaming, or to the
<a href="/lessons/wireless-101/">start of the course</a>. The hands-on part of the course will configure a wireless
controller and access points, join laptops to them, and capture these same frames in Wireshark; until then,
<a href="/lessons/networking-101/">Networking 101</a> covers the wired side.</p>
]]></content:encoded></item><item><title>Joining and roaming: how a laptop picks an access point</title><link>https://networkdoodles.com/lessons/wireless-101/task-6-1/</link><pubDate>Mon, 16 Nov 2026 09:00:00 -0500</pubDate><dc:creator>David Henderson</dc:creator><guid>https://networkdoodles.com/lessons/wireless-101/task-6-1/</guid><description>Beacons, probes, association and the four-way handshake, then roaming: thresholds, margins, sticky clients and 802.11k, v and r, with a laptop you walk in 3D.</description><content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[
<section class="lab-stage lab-stage-explain" id="explain">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">1</span>
    <span class="lab-stage-kind mono">explain</span><h2>How does a laptop join, and when does it move?</h2>
  </header>
  <div class="lab-stage-body">
    <h3 id="shop-signs">Shop signs</h3>
<p>Walking down a street you see shop signs without asking anyone: each shop puts its name up for passers-by. You
can also call out <em>&ldquo;Anyone sell batteries?&rdquo;</em> and the shops that do answer. Wi-Fi works both ways:</p>
<ul>
<li><strong>Beacons</strong>: each access point broadcasts a small frame roughly <strong>ten times a second</strong> (every 102.4 ms), for
each network it offers. A beacon carries the network name (the <strong>SSID</strong>, <em>service set identifier</em>), the
channel, the security it needs and the speeds it supports.</li>
<li><strong>Probes</strong>: a laptop can send a <strong>probe request</strong>, <em>&ldquo;Is network X here?&rdquo;</em>, and every access point that has it
sends a <strong>probe response</strong>, which is a beacon sent just to that laptop.</li>
</ul>
<p>A laptop builds its list of networks, and which access points offer each one, from beacons and probe responses.
Each access point radio offering a network has its own address, the <strong>BSSID</strong> (it looks like a MAC address).</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Access points announce themselves with beacons; laptops can also ask with probes.</p>
<h3 id="joining-four-exchanges">Joining: four exchanges</h3>
<p>Once the laptop has picked an access point, usually the strongest one for the network it wants, joining takes
four exchanges:</p>
<ol>
<li><strong>Authentication.</strong> A leftover from early Wi-Fi: in nearly every network today it&rsquo;s a formality, two frames
that say <em>hello</em> and <em>OK</em> (<em>open system</em> authentication). The real security comes later. (With WPA3,
lesson 7, this step carries a real exchange.)</li>
<li><strong>Association.</strong> The laptop asks to join (<strong>association request</strong>, with its capabilities) and the access
point says yes (<strong>association response</strong>), giving it an <strong>association ID</strong>.</li>
<li><strong>The four-way handshake.</strong> On a secure network, four frames prove both sides know the network&rsquo;s secret and
agree fresh encryption keys for this connection. Lesson 7 opens it up.</li>
<li><strong>DHCP.</strong> Now the laptop is on the network like any wired host, and gets its IP address exactly as in
<a href="/lessons/networking-101/task-4-1/">Networking 101&rsquo;s DHCP lesson</a>.</li>
</ol>
<p class="lab-recap"><span class="mono">in one sentence</span> Authenticate (a formality), associate, four-way handshake, then DHCP.</p>
<h3 id="roaming-the-laptop-decides">Roaming: the laptop decides</h3>
<p>As you walk away from an access point, its signal fades. At some point the laptop should move to a better one:
that&rsquo;s a <strong>roam</strong>. The surprise for most people is <strong>who decides: the laptop</strong>, not the access point or the
controller. Each laptop, phone and scanner runs its own logic, and the common shape is:</p>
<ul>
<li>Stay put while the current access point is above the <strong>roam threshold</strong> (often somewhere around −70 to
−75 dBm, and different on every device).</li>
<li>Below it, start looking: listen for beacons, send probes.</li>
<li>Move only to an access point that&rsquo;s better by at least a <strong>margin</strong> (also called <strong>hysteresis</strong>), often 5 to
10 dB, so it doesn&rsquo;t flip back and forth.</li>
</ul>
<p>A roam is a <strong>reassociation</strong> to the new access point, then the handshake again. Done well, it takes a few tens
of milliseconds and a call doesn&rsquo;t notice.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> The laptop roams when its signal drops below its threshold and another AP is better by its margin.</p>
<h3 id="sticky-clients">Sticky clients</h3>
<p>If a device&rsquo;s threshold is set too low, say −80 dBm, it clings to an access point long after a much better one
is right next to it. It still has a connection, at a terrible speed, and because it&rsquo;s slow it hogs the access
point&rsquo;s airtime too. That&rsquo;s a <strong>sticky client</strong>, one of the most common Wi-Fi complaints.</p>
<p>You can&rsquo;t reach into every device to fix its settings. Networks help in three ways, each a standard:</p>
<table>
  <thead>
      <tr>
          <th>Standard</th>
          <th>Nickname</th>
          <th>What it does</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td><strong>802.11k</strong></td>
          <td>neighbor reports</td>
          <td>the access point tells the laptop which neighbors to check, so it scans faster</td>
      </tr>
      <tr>
          <td><strong>802.11v</strong></td>
          <td>BSS transition management</td>
          <td>the access point can <em>suggest</em> the laptop move (it can still say no)</td>
      </tr>
      <tr>
          <td><strong>802.11r</strong></td>
          <td>fast transition</td>
          <td>the keys for the next access point are worked out ahead, so the roam skips most of the handshake</td>
      </tr>
  </tbody>
</table>
<p>Some networks also disconnect a device whose signal falls very low, to force it to find a better access point.
That&rsquo;s a blunt tool, and it can make things worse for devices that roam badly anyway.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Sticky clients cling too long. 802.11k helps them look, 802.11v nudges them, 802.11r makes the roam quick.</p>

  </div>
</section>


<section class="lab-stage lab-stage-show" id="show">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">2</span>
    <span class="lab-stage-kind mono">show</span><h2>The frames, one at a time</h2>
  </header>
  <div class="lab-stage-body">
    <p>A laptop joins <strong>AP 1</strong> on a WPA2 network called <em>office</em>, then walks toward <strong>AP 2</strong> and roams. Both access
points offer the same network on different channels (lesson 2). The <strong>inside the frame</strong> box shows the important
fields in each frame.</p>
<p>The frames to watch for:</p>
<table>
  <thead>
      <tr>
          <th>Frame</th>
          <th>From</th>
          <th>What it says</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td>Beacon</td>
          <td>AP</td>
          <td><em>&ldquo;I&rsquo;m network office, on channel 36, WPA2.&rdquo;</em></td>
      </tr>
      <tr>
          <td>Probe request and response</td>
          <td>laptop, then AP</td>
          <td><em>&ldquo;Is office here?&rdquo;</em> <em>&ldquo;Yes.&rdquo;</em></td>
      </tr>
      <tr>
          <td>Authentication</td>
          <td>both</td>
          <td><em>&ldquo;Hello.&rdquo; &ldquo;OK.&rdquo;</em></td>
      </tr>
      <tr>
          <td>Association request and response</td>
          <td>laptop, then AP</td>
          <td><em>&ldquo;May I join?&rdquo;</em> <em>&ldquo;Yes, you&rsquo;re number 1.&rdquo;</em></td>
      </tr>
      <tr>
          <td>Four-way handshake</td>
          <td>both</td>
          <td>four EAPOL-Key frames that set up encryption</td>
      </tr>
      <tr>
          <td>Reassociation</td>
          <td>laptop, then new AP</td>
          <td><em>&ldquo;I&rsquo;m moving to you from AP 1.&rdquo;</em></td>
      </tr>
  </tbody>
</table>
<p class="lab-recap"><span class="mono">in one sentence</span> Beacon, probe, authenticate, associate, handshake: and on a roam, reassociate and handshake again.</p>

  </div>
</section>


<section class="lab-stage lab-stage-follow" id="follow">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">3</span>
    <span class="lab-stage-kind mono">follow along</span><h2>Watch a laptop join, then roam</h2>
  </header>
  <div class="lab-stage-body">
    <figure class="netsim" data-netsim tabindex="0" aria-label="A laptop joins access point AP 1 with beacon, probe, authentication, association and the four-way handshake, then walks toward AP 2, its signal falls below its roam threshold, and it reassociates to AP 2. Use the buttons, or the left and right arrow keys, to step through it.">
  <svg class="ns-svg" role="img" aria-label="A laptop joins access point AP 1 with beacon, probe, authentication, association and the four-way handshake, then walks toward AP 2, its signal falls below its roam threshold, and it reassociates to AP 2"></svg>
  <div class="ns-ctl">
    <button class="icon-btn" type="button" data-ns="restart" aria-label="Back to the first step" title="Start again"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M6 5v14M19 5l-9 7 9 7z"/></svg>
</button>
    <button class="icon-btn" type="button" data-ns="prev" aria-label="Previous step" title="Previous step"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M19 12H5M11 6l-6 6 6 6"/></svg>
</button>
    <button class="btn sm primary ns-play" type="button" data-ns="play" aria-pressed="false"><span data-ns-play>Play</span></button>
    <button class="icon-btn" type="button" data-ns="next" aria-label="Next step" title="Next step"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M5 12h14M13 6l6 6-6 6"/></svg>
</button>
    <button class="icon-btn" type="button" data-ns="again" aria-label="Play this step again" title="Play this step again"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M4 12a8 8 0 1 0 2.5-5.8M4 4v4h4"/></svg>
</button>
    <span class="ns-count mono" data-ns-count></span>
    <button class="btn sm ns-speed" type="button" data-ns="speed" aria-pressed="false">Slow: off</button>
  </div>
  <div class="ns-dots" data-ns-dots role="group" aria-label="Jump to a step"></div>
  <figcaption class="ns-caption" data-ns-caption aria-live="polite"></figcaption>
  <div class="ns-panels">
    <div class="ns-codebox" data-ns-code hidden></div>
    <div class="ns-frame" data-ns-frame hidden></div>
    <div class="ns-tablebox" data-ns-table hidden></div>
  </div>
  <script type="application/json">{"h":360,"links":[["pc","ap1",{"wifi":true}],["pc","ap2",{"wifi":true}]],"narrow":{"at":{"ap1":[90,90],"ap2":[270,90],"pc":[110,420]},"h":560,"w":360},"nodes":{"ap1":{"kind":"wifi","label":"AP 1","sub":"office · channel 36","x":150,"y":90},"ap2":{"kind":"wifi","label":"AP 2","sub":"office · channel 149","x":630,"y":90},"pc":{"kind":"laptop","label":"Laptop","sub":"not connected","sub2":"roams below −70 dBm","x":200,"y":270}},"steps":[{"frame":{"layers":[{"fields":[["type","beacon",true],["to","everyone (broadcast)"],["BSSID","AP 1's radio"],["SSID","office",true],["channel","36"],["security","WPA2, AES",true],["interval","102.4 ms"]],"l":2,"name":"802.11 management"}],"title":"Inside a **BEACON** from AP 1"},"hot":["ap1","ap2"],"links":[["pc","ap1"],["pc","ap2"]],"packets":[{"c":"cyan","label":"BEACON","path":["ap1","pc"]},{"c":"cyan","label":"BEACON","path":["ap2","pc"]}],"table":{"cols":["AP","signal"],"hl":0,"rows":[["AP 1","−52 dBm"],["AP 2","−78 dBm"]],"title":"What the laptop hears"},"text":"Both access points broadcast beacons about ten times a second. The laptop hears both: AP 1 at **−52 dBm**, AP 2 far away at **−78 dBm**.","title":"Beacons."},{"frame":{"layers":[{"fields":[["type","probe request",true],["to","everyone"],["SSID wanted","office",true],["speeds I support","Wi-Fi 6, 2 streams"]],"l":2,"name":"802.11 management"}],"title":"Inside the **PROBE REQUEST**"},"hot":["pc","ap1"],"links":[["pc","ap1"]],"packets":[{"c":"amber","label":"PROBE REQ","path":["pc","ap1"]},{"at":1,"c":"green","label":"PROBE RESP","path":["ap1","pc"]}],"say":{"pc":"Is office here?"},"text":"The laptop asks out loud: *is **office** here?* AP 1 answers with a **probe response**, a beacon sent just to the laptop.","title":"Probe."},{"frame":{"layers":[{"fields":[["type","authentication",true],["algorithm","open system",true],["sequence","1, then 2"],["status","success"]],"l":2,"name":"802.11 management"}],"title":"Inside the **AUTHENTICATION**"},"hot":["pc","ap1"],"links":[["pc","ap1"]],"packets":[{"c":"amber","label":"AUTH","path":["pc","ap1"]},{"at":1,"c":"green","label":"AUTH OK","path":["ap1","pc"]}],"text":"Two frames: *hello* and *OK*. On a WPA2 network this is **open system** authentication, a formality kept from early Wi-Fi. The real security comes two steps later.","title":"Authentication."},{"frame":{"layers":[{"fields":[["type","association response",true],["status","success"],["association ID","1",true],["speeds agreed","Wi-Fi 6, 2 streams"]],"l":2,"name":"802.11 management"}],"title":"Inside the **ASSOCIATION RESPONSE**"},"hot":["pc","ap1"],"links":[["pc","ap1"]],"packets":[{"c":"amber","label":"ASSOC REQ","path":["pc","ap1"]},{"at":1,"c":"green","label":"ASSOC RESP","path":["ap1","pc"]}],"set":{"pc":"associated to AP 1"},"text":"The laptop asks to join, listing what it can do. AP 1 says yes and gives it an **association ID**. The laptop is on the access point, but it can't send data yet: no keys.","title":"Association."},{"frame":{"layers":[{"fields":[["to","AP 1"],["from","laptop"]],"l":2,"name":"802.11 data"},{"fields":[["message","2 of 4",true],["nonce","random, from the laptop"],["MIC","proves it knows the secret",true]],"l":3,"name":"EAPOL-Key"}],"title":"Inside **message 2 of 4**"},"good":["pc"],"hot":["pc","ap1"],"links":[["pc","ap1"]],"packets":[{"c":"violet","label":"1/4","path":["ap1","pc"]},{"at":1,"c":"violet","label":"2/4","path":["pc","ap1"]},{"at":2,"c":"violet","label":"3/4","path":["ap1","pc"]},{"at":3,"c":"violet","label":"4/4","path":["pc","ap1"]}],"set":{"pc":"connected to AP 1 · −52 dBm"},"text":"Four **EAPOL-Key** frames prove both sides know the network's secret and agree fresh keys for this one connection. Lesson 7 opens them up. Then DHCP, and the laptop is online.","title":"The four-way handshake."},{"frame":{"layers":[{"fields":[["type","probe request",true],["SSID wanted","office"],["why","signal below the roam threshold",true]],"l":2,"name":"802.11 management"}],"title":"Inside the **PROBE REQUEST**, now to AP 2's channel"},"hot":["pc"],"links":[["pc","ap1"],["pc","ap2"]],"move":{"pc":{"narrow":[250,420],"wide":[560,270]}},"packets":[{"c":"amber","label":"PROBE REQ","path":["pc","ap2"]},{"at":1,"c":"green","label":"PROBE RESP","path":["ap2","pc"]}],"say":{"pc":"Below −70.\nWho else is here?"},"set":{"pc":"AP 1 · −72 dBm"},"table":{"cols":["AP","signal"],"hl":1,"rows":[["AP 1","−72 dBm"],["AP 2","−55 dBm"]],"title":"What the laptop hears"},"text":"The laptop's owner walks toward AP 2. AP 1 fades: **−66 dBm**, then **−72 dBm**. While it was above **−70 dBm**, the laptop didn't even look. Now it's below its threshold, so it starts listening for beacons and sending probes.","title":"Walking away."},{"cut":[["pc","ap1"]],"frame":{"layers":[{"fields":[["type","reassociation request",true],["to","AP 2"],["current AP","AP 1",true],["SSID","office"]],"l":2,"name":"802.11 management"}],"title":"Inside the **REASSOCIATION REQUEST**"},"good":["pc"],"hot":["pc","ap2"],"links":[["pc","ap2"]],"packets":[{"c":"amber","label":"REASSOC REQ","path":["pc","ap2"]},{"at":1,"c":"green","label":"REASSOC RESP","path":["ap2","pc"]},{"at":2,"c":"violet","label":"1/4 … 4/4","path":["ap2","pc"]}],"set":{"pc":"connected to AP 2 · −55 dBm"},"text":"AP 2 is **17 dB** better, more than the laptop's 5 dB margin. It sends a **reassociation request** naming AP 1 as where it came from, AP 2 accepts, and the four-way handshake runs again with AP 2. With **802.11r** most of that handshake would be skipped.","title":"Reassociation to AP 2."}],"w":780}</script>
</figure>
<script src="/js/netsim.c40a98d103487ae5866edfa12abb4afc1d3aa7ad2d12fed4fe925fe00b588106.js" integrity="sha256-xAqY0QNIeuWGbt&#43;hKrtK/B06p60tEv7U/pJf4AtYgQY=" defer></script>
<p>Now walk a laptop through the office from the earlier lessons. Three access points at 5 dBm, a laptop at the
start of a walk, and a graph of every access point&rsquo;s signal along the way.</p>
<link rel="stylesheet" href="/css/rflab.min.6470f6e7f26d78ca57cbe2d4ea3c5e64bc6a1cf89d42def3f2f99e04198472e4.css" integrity="sha256-ZHD25/JteMpXy&#43;LU6jxeZLxqHPidQt7z8vmeBBmEcuQ=">
<script src="/js/rflab-loader.d2c182aadbcb506f3c99779fa2c245f6721a9919c45d140ceb6c595cdea20b9c.js" integrity="sha256-0sGCqtvLUG88mXefosJF9nIamRnEXRQM62xZXN6iC5w=" defer></script>
<figure class="rflab" data-rflab data-mode="floor" data-src="/js/rflab/main.1374b7b3a47b299b16ed801f63ad98980a6452563b2455f12aef6f7fd65a3a0d.js" data-integrity="sha256-E3S3s6R7KZsW7YAfY62YmApkUlY7JFXxKu9vf9ZaOg0=" data-start="click" aria-label="Guided tasks: walk a laptop through the office with three access points; make it a sticky client, then tune its roaming">
  <div class="rf-poster">
    <img src="/images/series/wireless-101/roaming_hu_dd3677b2cbb6578a.webp" width="1020" height="682" alt="Guided tasks: walk a laptop through the office with three access points; make it a sticky client, then tune its roaming" loading="lazy" decoding="async">
    <button class="btn primary rf-start" type="button" data-rflab-start>Start the 3D lab</button>
  </div>
  <div data-rflab-app></div>
  <script type="application/json" data-rflab-config>{"antenna":"dipole","mode":"floor","scenario":"office","tasks":["sticky","roam"]}</script><figcaption><p>Two tasks, each checked by the lab:</p>
<ol>
<li><strong>Make a sticky client.</strong> Change the laptop&rsquo;s roam threshold so it clings to AP 1 until the break room door.</li>
<li><strong>Tune the roaming.</strong> Starting from a sticky −82 dBm, keep the signal above −72 dBm the whole walk, without
flapping back to an access point it just left.</li>
</ol>
<p>Press <strong>Walk the path</strong> to walk. The trail on the floor and the thick line on the graph show which access point the
laptop is using; dotted lines mark each roam. You can also drag the laptop yourself and read why it stayed or moved.</p>
</figcaption>
</figure>
<p class="lab-recap"><span class="mono">in one sentence</span> The laptop joins in four steps, and roams when its signal drops below its threshold and a better AP is past its margin.</p>

  </div>
</section>


<section class="lab-stage lab-stage-challenge" id="challenge">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">4</span>
    <span class="lab-stage-kind mono">challenge</span><h2>Diagnose the complaint</h2>
  </header>
  <div class="lab-stage-body">
    <p>A user says: <em>&ldquo;In the meeting room my laptop shows full bars, but video calls stutter. Everyone else&rsquo;s are fine.&rdquo;</em>
The meeting room has its own access point. A quick look at the user&rsquo;s laptop shows it connected at <strong>−79 dBm</strong>,
to an access point at the other end of the floor.</p>
<ol>
<li>What&rsquo;s going on?</li>
<li>Why does the laptop show <em>full bars</em>?</li>
<li>Give two fixes, one on the laptop and one on the network.</li>
</ol>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>Which access point is it on, and when would it have last chosen one?</p>
</div>
</details>
<details class="lab-solution">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg>
 <span>show the solution</span></summary>
  <div class="lab-solution-body">
    <ol>
<li>A <strong>sticky client</strong>. The laptop joined the far access point earlier, maybe at a desk near it, and its roam
threshold is lower than −79 dBm, so it has never gone looking, even though the meeting room&rsquo;s own access
point would give it −50 dBm.</li>
<li>Bars are each maker&rsquo;s own scale, and some show <em>connected</em> as full bars. The −79 dBm number tells the truth.</li>
<li>On the laptop: many Wi-Fi drivers have a <strong>roaming aggressiveness</strong> setting; raise it, or update the driver.
Simply turning Wi-Fi off and on makes it pick the strongest access point again. On the network: turn on
<strong>802.11k and 802.11v</strong> so access points can point it at better neighbors, and check whether the far access
point&rsquo;s power is higher than it needs to be, since a loud distant access point attracts sticky clients.</li>
</ol>
  </div>
</details>

  </div>
</section>

<p>Back to <a href="/lessons/wireless-101/task-5-1/">lesson 5</a> for placing access points, or on to
<a href="/lessons/wireless-101/task-7-1/">lesson 7</a>, which opens up the four-way handshake.</p>
]]></content:encoded></item><item><title>Coverage and AP placement: height, power and overlap</title><link>https://networkdoodles.com/lessons/wireless-101/task-5-1/</link><pubDate>Mon, 16 Nov 2026 09:00:00 -0500</pubDate><dc:creator>David Henderson</dc:creator><guid>https://networkdoodles.com/lessons/wireless-101/task-5-1/</guid><description>Design for −67 dBm and 25 dB SNR: mounting height and the null under an AP, why full power backfires, overlap, and covering an office in an interactive 3D lab.</description><content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[
<section class="lab-stage lab-stage-explain" id="explain">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">1</span>
    <span class="lab-stage-kind mono">explain</span><h2>What makes a good Wi-Fi design?</h2>
  </header>
  <div class="lab-stage-body">
    <h3 id="coverage-is-only-half-of-it">Coverage is only half of it</h3>
<p>Lighting a room is easy: put up enough bulbs and it&rsquo;s bright. Wi-Fi is harder, because it&rsquo;s a <strong>conversation</strong>.
Every access point has to hear every device as well as the device hears it, and every access point on a
channel shares that channel&rsquo;s airtime with every other one it can hear. So a good design is about three
things at once:</p>
<ol>
<li><strong>Coverage</strong>: enough signal everywhere people work.</li>
<li><strong>Capacity</strong>: enough airtime for everyone, which means not too many devices per access point, and not too
many access points on the same channel within earshot of each other.</li>
<li><strong>Roaming</strong>: as people walk, there&rsquo;s always a good next access point (lesson 6).</li>
</ol>
<p class="lab-recap"><span class="mono">in one sentence</span> Good Wi-Fi is coverage, capacity and roaming, not just a strong signal.</p>
<h3 id="design-targets">Design targets</h3>
<p>The common targets:</p>
<table>
  <thead>
      <tr>
          <th>Use</th>
          <th>Signal (RSSI)</th>
          <th>SNR</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td>Voice and video calls</td>
          <td><strong>−67 dBm</strong> or better</td>
          <td><strong>25 dB</strong> or more</td>
      </tr>
      <tr>
          <td>Web, email, general data</td>
          <td>about −70 dBm</td>
          <td>about 20 dB</td>
      </tr>
      <tr>
          <td>Barely working</td>
          <td>−80 dBm</td>
          <td>10 to 15 dB</td>
      </tr>
  </tbody>
</table>
<p>Designers aim for the target <strong>everywhere people use Wi-Fi</strong>, not just on average, and they leave a margin: a
body between you and the access point costs several dB, and fading (lesson 4) swings the signal by several more.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Design for −67 dBm and 25 dB SNR where people work, with a margin to spare.</p>
<h3 id="mounting-height-and-the-antenna">Mounting height and the antenna</h3>
<p>Most office access points hang from the ceiling, 8 to 12 ft up; warehouses and halls can be 20 ft or more. Two
things change with height:</p>
<ul>
<li><strong>Distance.</strong> Higher means farther from everyone, everywhere: a little less signal across the board.</li>
<li><strong>The antenna&rsquo;s shape.</strong> From lesson 3, a plain upright dipole has a hole straight down. Hang one 20 ft up and
the spot directly underneath gets the hole. That&rsquo;s why ceiling access points use antennas shaped to send
<strong>down</strong>, strongest about 45° below the horizon, with only a shallow dip straight down.</li>
</ul>
<p>In high ceilings, designers often use access points with directional antennas aimed at the floor, or mount them
lower on walls and columns.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Higher is farther from everyone, and the antenna must be shaped to send down.</p>
<h3 id="transmit-power-the-conversation-has-two-sides">Transmit power: the conversation has two sides</h3>
<p>It&rsquo;s tempting to turn every access point up to the maximum. It backfires twice:</p>
<ul>
<li><strong>The device can&rsquo;t shout back.</strong> An access point might send 20 dBm; a phone often sends 12 to 15 dBm with a
worse antenna. If the access point is loud enough to reach a phone far away, the phone&rsquo;s answer may not make
it back. The link is lopsided: full bars, poor performance.</li>
<li><strong>Bigger cells overlap more.</strong> A Wi-Fi radio waits its turn whenever it can hear another one on its channel,
down to very faint signals (about <strong>−82 dBm</strong> for a 20 MHz channel). Turn every access point up and each one
hears more of the others, so they share airtime with more of them. That&rsquo;s co-channel interference from
lesson 2, made worse.</li>
</ul>
<p>So designers usually run <strong>more access points at moderate power</strong> (often 11 to 17 dBm on 5 GHz) rather than a few
at full blast.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Loud access points make lopsided links and share airtime with more neighbors. More, quieter APs win.</p>
<h3 id="spacing-and-overlap">Spacing and overlap</h3>
<p>Neighboring access points&rsquo; −67 dBm areas should <strong>overlap a little</strong>, so a device walking between them always
has a good one to move to. A common starting
point in offices on 5 GHz is roughly one access point per <strong>2,500 to 4,000 sq ft</strong>, closer where walls are heavy
or many people gather. Rules of thumb only start a design: the floor plan, the walls and the people decide it.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Overlap neighbors a little at −67 dBm so the walk between them never hits a gap.</p>
<h3 id="directional-antennas-for-odd-shapes">Directional antennas for odd shapes</h3>
<p>Long, narrow spaces, such as hallways, warehouse aisles and stadium seating, suit <strong>directional</strong> antennas from
lesson 3. A beam aimed down an aisle puts the signal where the people are, and the racks or walls either side
keep it out of the neighbors&rsquo; space.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Long, narrow spaces get beams aimed down them.</p>

  </div>
</section>


<section class="lab-stage lab-stage-show" id="show">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">2</span>
    <span class="lab-stage-kind mono">show</span><h2>How designers work</h2>
  </header>
  <div class="lab-stage-body">
    <p>A wireless design usually goes in four steps:</p>
<ol>
<li><strong>Requirements</strong>: what people do where (calls, video, scanners, cameras), how many devices, which bands.</li>
<li><strong>Predictive design</strong>: a model like this course&rsquo;s lab, with the real floor plan and wall types, to place
access points and pick power and channels before anything is installed.</li>
<li><strong>Install, then survey</strong>: walk the building with a laptop measuring the real signal (a <strong>site survey</strong>), because
real walls, furniture and people never match the model exactly.</li>
<li><strong>Adjust and recheck</strong>: move or add access points, change power and channels, and survey again.</li>
</ol>
<p>The lab below is step 2. It shows the share of the floor at −67 dBm or better as you work, and it can export
your design as a file (in its panel, under <em>Design</em>), the same shape the hands-on part of the course reads.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Requirements, predict, survey, adjust: models start a design, surveys finish it.</p>

  </div>
</section>


<section class="lab-stage lab-stage-follow" id="follow">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">3</span>
    <span class="lab-stage-kind mono">follow along</span><h2>Height, antennas and covering the office</h2>
  </header>
  <div class="lab-stage-body">
    <link rel="stylesheet" href="/css/rflab.min.6470f6e7f26d78ca57cbe2d4ea3c5e64bc6a1cf89d42def3f2f99e04198472e4.css" integrity="sha256-ZHD25/JteMpXy&#43;LU6jxeZLxqHPidQt7z8vmeBBmEcuQ=">
<script src="/js/rflab-loader.d2c182aadbcb506f3c99779fa2c245f6721a9919c45d140ceb6c595cdea20b9c.js" integrity="sha256-0sGCqtvLUG88mXefosJF9nIamRnEXRQM62xZXN6iC5w=" defer></script>
<figure class="rflab" data-rflab data-mode="floor" data-src="/js/rflab/main.1374b7b3a47b299b16ed801f63ad98980a6452563b2455f12aef6f7fd65a3a0d.js" data-integrity="sha256-E3S3s6R7KZsW7YAfY62YmApkUlY7JFXxKu9vf9ZaOg0=" data-start="click" aria-label="Guided tasks on the office floor: fill in the weak spot under a 20 ft dipole, then cover 95% of the office at −67 dBm with three access points">
  <div class="rf-poster">
    <img src="/images/series/wireless-101/signal-lab_hu_40697ca6ad62c097.webp" width="1020" height="682" alt="Guided tasks on the office floor: fill in the weak spot under a 20 ft dipole, then cover 95% of the office at −67 dBm with three access points" loading="lazy" decoding="async">
    <button class="btn primary rf-start" type="button" data-rflab-start>Start the 3D lab</button>
  </div>
  <div data-rflab-app></div>
  <script type="application/json" data-rflab-config>{"antenna":"dipole","mode":"floor","scenario":"office","tasks":["fill","cover"]}</script><figcaption><p>Two tasks, each checked by the lab:</p>
<ol>
<li><strong>Fill in the hole.</strong> A dipole hangs 20 ft up. Get −67 dBm straight under it without moving it sideways. Try
the height and the antenna.</li>
<li><strong>Cover the office.</strong> Get 95% of the floor to −67 dBm on 5 GHz with three access points or fewer. Watch the
coverage figure as you drag them, and remember the coated glass.</li>
</ol>
<p>Things to try once they&rsquo;re done: turn one access point up to 23 dBm and watch how far its cell spreads; switch
the whole floor to 2.4 GHz and see the cells grow; switch to 6 GHz and see them shrink.</p>
</figcaption>
</figure>
<p class="lab-recap"><span class="mono">in one sentence</span> The antenna fixes the hole; spreading three access points out covers the floor.</p>

  </div>
</section>


<section class="lab-stage lab-stage-challenge" id="challenge">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">4</span>
    <span class="lab-stage-kind mono">challenge</span><h2>The warehouse</h2>
  </header>
  <div class="lab-stage-body">
    <p>A warehouse is 200 ft long. Six aisles run the full length, separated by metal racks floor to ceiling. The ceiling
is 30 ft. Staff use handheld scanners in the aisles.</p>
<ol>
<li>Why is one powerful omni on the ceiling in the middle a poor idea?</li>
<li>Sketch a better design.</li>
<li>What would you check after it&rsquo;s installed?</li>
</ol>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>Think about what the metal racks do to a signal crossing them (lesson 4), and what a dipole does straight below it
from 30 ft up (lesson 3).</p>
</div>
</details>
<details class="lab-solution">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg>
 <span>show the solution</span></summary>
  <div class="lab-solution-body">
    <ol>
<li>Three problems at once: metal racks full of stock cost a lot each (often 20 dB or more), so the signal
crossing several of them dies;
from 30 ft up an omni&rsquo;s weakest direction points at the floor; and a scanner&rsquo;s small radio can&rsquo;t shout back
across 100 ft and several racks.</li>
<li>Put the access points <strong>in the aisles</strong>, on the end walls or columns, with <strong>directional antennas aimed down
each aisle</strong>, one or two per aisle depending on length, or ceiling access points with down-facing antennas
lower over each aisle. Moderate power, and channels alternated so neighboring aisles differ.</li>
<li>A <strong>site survey</strong> walking each aisle with a scanner-like device: signal and SNR along the whole length, roaming
between access points as you walk (lesson 6), and a check that full racks of stock haven&rsquo;t changed the picture
since the empty-building survey.</li>
</ol>
  </div>
</details>

  </div>
</section>

<p>Back to <a href="/lessons/wireless-101/task-4-1/">lesson 4</a> for what walls cost, or on to
<a href="/lessons/wireless-101/task-6-1/">lesson 6</a>, where a laptop walks between the access points you placed. Channels
for all these access points come from <a href="/lessons/wireless-101/task-2-1/">lesson 2</a>.</p>
]]></content:encoded></item><item><title>How signal travels: distance, walls and bounces</title><link>https://networkdoodles.com/lessons/wireless-101/task-4-1/</link><pubDate>Mon, 16 Nov 2026 09:00:00 -0500</pubDate><dc:creator>David Henderson</dc:creator><guid>https://networkdoodles.com/lessons/wireless-101/task-4-1/</guid><description>Free-space loss, what drywall, glass, coated glass, concrete and metal cost on 2.4 and 5 GHz, and reflection and multipath, on an interactive 3D office floor.</description><content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[
<section class="lab-stage lab-stage-explain" id="explain">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">1</span>
    <span class="lab-stage-kind mono">explain</span><h2>What happens to a signal on its way?</h2>
  </header>
  <div class="lab-stage-body">
    <h3 id="spreading-out-free-space-loss">Spreading out: free-space loss</h3>
<p>Back to the pond from <a href="/lessons/wireless-101/task-1-1/">lesson 1</a>. A ripple gets weaker as it spreads, because
the same energy is shared over a bigger ring. Radio spreads over a bigger <em>sphere</em>, so the loss grows with the
square of the distance. Engineers call it <strong>free-space path loss</strong> (<strong>FSPL</strong>):</p>
<div class="highlight"><pre tabindex="0" class="chroma"><code class="language-text" data-lang="text"><span class="line"><span class="cl">FSPL (dB) = 20 × log10(distance in m) + 20 × log10(frequency in MHz) − 27.55
</span></span></code></pre></div><p>You don&rsquo;t need the formula; you need what it means:</p>
<ul>
<li><strong>Twice the distance costs about 6 dB</strong>, a quarter of the power. Ten times the distance costs 20 dB.</li>
<li>At 1 m (3 ft), the air already costs <strong>40 dB</strong> on 2.4 GHz and <strong>47 dB</strong> on 5 GHz.</li>
</ul>
<p class="lab-recap"><span class="mono">in one sentence</span> In open space, every doubling of distance costs about 6 dB.</p>
<h3 id="why-does-5-ghz-reach-less-far">Why does 5 GHz reach less far?</h3>
<p>You&rsquo;ll often hear that higher frequencies &ldquo;fade faster&rdquo;. In empty space they don&rsquo;t. The wave spreads exactly the
same way. What changes is the <strong>receiving antenna</strong>: an antenna of the same type is smaller at a higher frequency
(about half a wavelength, from lesson 1), so it catches less of the passing wave. That&rsquo;s the
20 × log10(frequency) part of the formula, and it works out to about <strong>6.5 dB</strong> between 2.4 and 5 GHz, and
8 dB to 6 GHz.</p>
<p>Then the walls make it worse: most building materials absorb more at higher frequencies, as the table below
shows. Indoors, that&rsquo;s usually the bigger difference.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> 5 GHz loses about 6.5 dB more in the air because the antennas catch less, and more in most walls.</p>
<h3 id="absorption-walls">Absorption: walls</h3>
<p>Every wall takes a share of the signal. These are the numbers this course&rsquo;s lab uses, straight through a wall at
right angles:</p>
<table>
  <thead>
      <tr>
          <th>Material</th>
          <th>2.4 GHz</th>
          <th>5 GHz</th>
          <th>6 GHz</th>
          <th>Where the number comes from</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td>Clear glass</td>
          <td>2.5 dB</td>
          <td>3.0 dB</td>
          <td>3.2 dB</td>
          <td>3GPP TR 38.901</td>
      </tr>
      <tr>
          <td>Drywall (two sheets on studs)</td>
          <td>3 dB</td>
          <td>4 dB</td>
          <td>4 dB</td>
          <td>WLAN design rule of thumb</td>
      </tr>
      <tr>
          <td>Wooden door</td>
          <td>5.1 dB</td>
          <td>5.5 dB</td>
          <td>5.6 dB</td>
          <td>3GPP TR 38.901</td>
      </tr>
      <tr>
          <td>Brick</td>
          <td>6 dB</td>
          <td>10 dB</td>
          <td>11 dB</td>
          <td>rule of thumb (NIST IR 6055)</td>
      </tr>
      <tr>
          <td>Concrete block</td>
          <td>10 dB</td>
          <td>15 dB</td>
          <td>16 dB</td>
          <td>rule of thumb (NIST IR 6055)</td>
      </tr>
      <tr>
          <td>Poured concrete (a structural wall)</td>
          <td>14.7 dB</td>
          <td>25.7 dB</td>
          <td>29.5 dB</td>
          <td>3GPP TR 38.901</td>
      </tr>
      <tr>
          <td>Coated (low-E) glass</td>
          <td>23.7 dB</td>
          <td>24.6 dB</td>
          <td>24.8 dB</td>
          <td>3GPP TR 38.901</td>
      </tr>
      <tr>
          <td>Metal (elevator doors, fire doors)</td>
          <td>30 dB</td>
          <td>35 dB</td>
          <td>35 dB</td>
          <td>nearly opaque</td>
      </tr>
  </tbody>
</table>
<p>Three things to notice:</p>
<ol>
<li><strong>Concrete gets much worse with frequency</strong>: 15 dB on 2.4 GHz, 26 on 5, 30 on 6.</li>
<li><strong>Coated glass is the surprise.</strong> Modern office windows and glass walls often have a thin metal coating to keep
heat in or out. You can&rsquo;t see it, but radio can: it costs as much as a concrete wall.</li>
<li>A wall crossed <strong>at a slant</strong> costs more, because the signal goes through more of it.</li>
</ol>
<p>Real walls vary a lot: wet concrete, metal studs, foil-backed insulation and steel mesh all change the numbers.
Treat the table as a starting point, not a measurement.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Drywall and clear glass cost a few dB; concrete, coated glass and metal cost a lot, more so at 5 and 6 GHz.</p>
<h3 id="reflection-scatter-and-diffraction">Reflection, scatter and diffraction</h3>
<p>Walls don&rsquo;t only absorb. Some of the signal does other things:</p>
<ul>
<li><strong>Reflection</strong>: it bounces off, like light off a mirror. Metal and coated glass reflect most of what hits them.</li>
<li><strong>Scatter</strong>: it hits something rough or cluttered, shelves, pipes, people, and sprays off in lots of directions.</li>
<li><strong>Diffraction</strong>: it bends a little around corners and through doorways, the way you can hear someone talking
round a corner.</li>
</ul>
<p>Because of these, the area &ldquo;behind&rdquo; a solid wall is never completely dark: some signal bounces and bends into it.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Signal also bounces, scatters and bends, so shadows behind walls are never completely dark.</p>
<h3 id="multipath-and-fading">Multipath and fading</h3>
<p>Indoors, the signal reaches a laptop by many routes at once: straight through the walls, off the ceiling, off a
filing cabinet. The copies arrive a few billionths of a second apart and <strong>add up or cancel</strong>, depending on exactly
where the laptop is. That&rsquo;s <strong>multipath</strong>. Because the waves are only a few inches long, moving the laptop an inch
or two can swing the signal by several dB. That swing is called <strong>fading</strong>.</p>
<p>Modern Wi-Fi turns multipath to its advantage (MIMO from lesson 3 needs it), and radios average their
measurements to smooth out fading. But it&rsquo;s why one reading from one spot never tells the whole story.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Copies of the signal arrive by different routes and add or cancel: the signal wobbles over inches.</p>

  </div>
</section>


<section class="lab-stage lab-stage-show" id="show">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">2</span>
    <span class="lab-stage-kind mono">show</span><h2>What the lab gets right, and what it leaves out</h2>
  </header>
  <div class="lab-stage-body">
    <p>The office below is a <strong>teaching model</strong>, not a survey tool. For every point on the floor it adds up exactly the
link budget from lesson 1: transmit power, the antenna&rsquo;s gain toward that point (lesson 3), free-space loss over
the straight-line distance, and the loss of every wall the straight line crosses.</p>
<p>What it <strong>leaves out</strong>, on purpose:</p>
<ul>
<li><strong>Reflections and diffraction.</strong> Shadows behind walls are darker than real ones, so the lab caps the walls on
any one path at 40 dB, a stand-in for signal finding its way around.</li>
<li><strong>Multipath and fading.</strong> The colors are smooth; a real floor is speckled.</li>
<li><strong>Real antennas.</strong> The patterns are ideal shapes, not a particular product&rsquo;s datasheet.</li>
<li><strong>People, furniture and clutter</strong>, which absorb a few dB each.</li>
<li><strong>Interference</strong> from neighbors and other devices.</li>
</ul>
<p>A real office loses more than free space suggests. In practice the signal often falls off closer to <strong>9 dB per
doubling</strong> of distance than 6 (the ITU&rsquo;s indoor model for offices). That&rsquo;s why professionals walk the building
with a laptop before and after a design: it&rsquo;s called a <strong>site survey</strong>.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> The lab adds up the straight path and its walls. Real buildings are messier, and surveys check them.</p>

  </div>
</section>


<section class="lab-stage lab-stage-follow" id="follow">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">3</span>
    <span class="lab-stage-kind mono">follow along</span><h2>Walls, bands and the conference room</h2>
  </header>
  <div class="lab-stage-body">
    <p>The office has drywall offices along the top with wooden doors, a concrete core with a metal elevator shaft, a
concrete-block storage room, a break room with a glass front, and a conference room behind coated glass. One
ceiling access point hangs in the middle at 10 ft.</p>
<link rel="stylesheet" href="/css/rflab.min.6470f6e7f26d78ca57cbe2d4ea3c5e64bc6a1cf89d42def3f2f99e04198472e4.css" integrity="sha256-ZHD25/JteMpXy&#43;LU6jxeZLxqHPidQt7z8vmeBBmEcuQ=">
<script src="/js/rflab-loader.d2c182aadbcb506f3c99779fa2c245f6721a9919c45d140ceb6c595cdea20b9c.js" integrity="sha256-0sGCqtvLUG88mXefosJF9nIamRnEXRQM62xZXN6iC5w=" defer></script>
<figure class="rflab" data-rflab data-mode="floor" data-src="/js/rflab/main.1374b7b3a47b299b16ed801f63ad98980a6452563b2455f12aef6f7fd65a3a0d.js" data-integrity="sha256-E3S3s6R7KZsW7YAfY62YmApkUlY7JFXxKu9vf9ZaOg0=" data-start="click" aria-label="Guided tasks on the office floor: rank wall losses on 2.4 and 5 GHz, find a spot covered on 2.4 but not on 5 GHz, and light up the conference room">
  <div class="rf-poster">
    <img src="/images/series/wireless-101/signal-lab_hu_40697ca6ad62c097.webp" width="1020" height="682" alt="Guided tasks on the office floor: rank wall losses on 2.4 and 5 GHz, find a spot covered on 2.4 but not on 5 GHz, and light up the conference room" loading="lazy" decoding="async">
    <button class="btn primary rf-start" type="button" data-rflab-start>Start the 3D lab</button>
  </div>
  <div data-rflab-app></div>
  <script type="application/json" data-rflab-config>{"antenna":"dipole","mode":"floor","scenario":"office","tasks":["walls","bands","glass"]}</script><figcaption><p>Three tasks, each checked by the lab:</p>
<ol>
<li><strong>Which walls cost the most?</strong> Click walls to read their loss, on both bands. The order isn&rsquo;t the same.</li>
<li><strong>Good on 2.4, not on 5.</strong> Find a spot one access point covers on 2.4 GHz but not on 5 GHz.</li>
<li><strong>Light up the conference room.</strong> Get −67 dBm in the middle of it without adding an access point.</li>
</ol>
<p>Click the floor to put the laptop down: the panel shows the whole link budget for that spot, wall by wall.</p>
</figcaption>
</figure>
<h3 id="try-it-on-your-own-computer-optional">Try it on your own computer (optional)</h3>
<p>Your own laptop will tell you its signal in dBm:</p>
<ul>
<li><strong>Mac</strong>: hold <kbd>Option</kbd> and click the Wi-Fi icon in the menu bar. Look for <strong>RSSI</strong> and <strong>Noise</strong>; the
difference is your SNR.</li>
<li><strong>Windows</strong>: in <em>Command Prompt</em>, type <code>netsh wlan show interfaces</code>. It shows <strong>Signal</strong> as a percentage;
recent versions also show <strong>Rssi</strong> in dBm.</li>
<li><strong>Linux</strong>: <code>iw dev</code> lists your Wi-Fi interface; then <code>iw dev wlan0 link</code> (with your interface&rsquo;s name) shows
<strong>signal</strong> in dBm.</li>
</ul>
<p>Walk around with it. Put a wall between you and the access point, then a door, then two walls. You&rsquo;ll see the
numbers from the table, give or take a lot: that&rsquo;s real buildings.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> The lab adds up walls exactly; your laptop shows what real ones do.</p>

  </div>
</section>


<section class="lab-stage lab-stage-challenge" id="challenge">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">4</span>
    <span class="lab-stage-kind mono">challenge</span><h2>Behind the concrete</h2>
  </header>
  <div class="lab-stage-body">
    <p>A laptop is <strong>30 ft</strong> from an access point, in open space, and gets <strong>−52 dBm</strong> on 5 GHz.</p>
<ol>
<li>You move it to <strong>60 ft</strong>, still in open space. Roughly what does it get?</li>
<li>Back at 30 ft, a <strong>poured concrete</strong> wall goes up between them. What now, on 5 GHz?</li>
<li>The same wall on <strong>2.4 GHz</strong> instead, if the access point is just as loud on both bands?</li>
<li>Which of those three spots is good enough for a video call?</li>
</ol>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>Twice the distance costs about 6 dB. Use the table for the wall. And remember 2.4 GHz also costs about 6.5 dB
less in the air.</p>
</div>
</details>
<details class="lab-solution">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg>
 <span>show the solution</span></summary>
  <div class="lab-solution-body">
    <ol>
<li>About <strong>−58 dBm</strong> (twice the distance, 6 dB less).</li>
<li>About <strong>−78 dBm</strong> (−52 − 25.7). Too weak for a call.</li>
<li>About <strong>−60 dBm</strong>: the air costs 6.5 dB less (−45.5 dBm without the wall) and the concrete about 14.7 dB:
−45.5 − 14.7 ≈ −60.</li>
<li>The first and the third. On 5 GHz, behind the concrete, you&rsquo;d want another access point on the other side of
the wall, which is lesson 5.</li>
</ol>
  </div>
</details>

  </div>
</section>

<p>Back to <a href="/lessons/wireless-101/task-3-1/">lesson 3</a> for antenna patterns, or on to
<a href="/lessons/wireless-101/task-5-1/">lesson 5</a>, where you put this to work and place access points.</p>
]]></content:encoded></item><item><title>Antennas and patterns: dBi, donuts and beams</title><link>https://networkdoodles.com/lessons/wireless-101/task-3-1/</link><pubDate>Mon, 16 Nov 2026 09:00:00 -0500</pubDate><dc:creator>David Henderson</dc:creator><guid>https://networkdoodles.com/lessons/wireless-101/task-3-1/</guid><description>What dBi means, why antenna gain is focus and not power, the omni donut and its null, beams for hallways, and how to read a datasheet, in a 3D view you can turn.</description><content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[
<section class="lab-stage lab-stage-explain" id="explain">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">1</span>
    <span class="lab-stage-kind mono">explain</span><h2>What does an antenna actually do?</h2>
  </header>
  <div class="lab-stage-body">
    <h3 id="a-lamp-and-a-flashlight">A lamp and a flashlight</h3>
<p>A bare light bulb lights the whole room a little. Put the same bulb in a flashlight and it lights one spot a lot,
and the rest of the room hardly at all. The bulb didn&rsquo;t get brighter; the reflector <strong>focused</strong> it.</p>
<p>An antenna is the reflector. The radio supplies the power; the antenna decides <strong>where it goes</strong>. That&rsquo;s the most
important idea in this lesson, so it&rsquo;s worth saying twice: <strong>antenna gain is focus, not extra power.</strong></p>
<p class="lab-recap"><span class="mono">in one sentence</span> The radio makes the power; the antenna aims it.</p>
<h3 id="dbi-compared-with-a-perfect-bulb">dBi: compared with a perfect bulb</h3>
<p>To say how much an antenna focuses, engineers compare it with an imaginary <strong>isotropic</strong> antenna that sends
exactly the same amount in every direction, like a perfect bare bulb floating in space. It can&rsquo;t be built, but it
makes a handy yardstick. <strong>dBi</strong> means <em>decibels compared with isotropic</em>:</p>
<ul>
<li><strong>0 dBi</strong>: the same as the imaginary bulb.</li>
<li><strong>+6 dBi</strong>: in its best direction, four times stronger than the bulb (+6 dB, from lesson 1). Somewhere else it
has to be weaker, to pay for it.</li>
</ul>
<p>So when a datasheet says <em>6 dBi</em>, it means <em>in its best direction</em>. Every gain in this course is in dBi.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> dBi is gain in the best direction, compared with an antenna that sends the same everywhere.</p>
<h3 id="the-radiation-pattern">The radiation pattern</h3>
<p>An antenna&rsquo;s <strong>radiation pattern</strong> is a 3D shape: how strongly it sends in each direction. The common ones:</p>
<table>
  <thead>
      <tr>
          <th>Antenna</th>
          <th>Shape</th>
          <th>Peak gain</th>
          <th>Good for</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td><strong>Dipole</strong> (the classic stick)</td>
          <td>a donut, with a hole straight up and down</td>
          <td>2.15 dBi</td>
          <td>general use, standing upright</td>
      </tr>
      <tr>
          <td><strong>High-gain omni</strong></td>
          <td>a flatter donut, a wider hole</td>
          <td>about 6 dBi</td>
          <td>reaching far along one floor</td>
      </tr>
      <tr>
          <td><strong>Ceiling AP antenna</strong></td>
          <td>aimed down, strongest 45° below</td>
          <td>about 4 dBi</td>
          <td>office ceilings</td>
      </tr>
      <tr>
          <td><strong>Patch</strong></td>
          <td>a wide beam one way</td>
          <td>about 10 dBi</td>
          <td>a hallway or a room from one wall</td>
      </tr>
      <tr>
          <td><strong>Yagi or sector</strong></td>
          <td>a narrow beam</td>
          <td>about 15 dBi</td>
          <td>a long warehouse aisle</td>
      </tr>
  </tbody>
</table>
<p><em>Omni</em> (short for <em>omnidirectional</em>) means it sends equally all the way around the horizon. It doesn&rsquo;t mean
equally in every direction: an omni&rsquo;s pattern is a donut, not a ball.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Omnis make donuts with holes above and below; directional antennas make beams.</p>
<h3 id="beamwidth">Beamwidth</h3>
<p>A beam doesn&rsquo;t stop sharply. Its <strong>beamwidth</strong> is measured between the points where the signal is <strong>3 dB below
the peak</strong>, half the power. A 65° patch is still sending a little at 80° off center, just less. And the narrower the
beam, the higher the gain, because the same power is squeezed into less space.</p>
<p>Most directional antennas also send a little backwards. The difference between front and back is the
<strong>front-to-back ratio</strong>, often about 20 dB.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Beamwidth is measured at half power, and narrower beams have more gain.</p>
<h3 id="polarization-and-many-antennas-at-once">Polarization, and many antennas at once</h3>
<p>Radio waves wiggle in a direction too: an upright antenna sends <strong>vertically polarized</strong> waves, and receives
them best. A laptop&rsquo;s antennas point every which way, and walls and furniture scramble polarization indoors, so
indoors it rarely matters much. Outdoors, on a long point-to-point link, both ends must match.</p>
<p>Look inside a modern access point and you&rsquo;ll find two, four or eight antennas, not one. They work together
(<strong>MIMO</strong>, <em>multiple input, multiple output</em>) to send several streams of data at once. They still add up to
roughly the shapes in the table above.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Indoors, polarization is scrambled anyway; modern access points use several antennas together.</p>

  </div>
</section>


<section class="lab-stage lab-stage-show" id="show">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">2</span>
    <span class="lab-stage-kind mono">show</span><h2>Reading an antenna datasheet</h2>
  </header>
  <div class="lab-stage-body">
    <p>A 3D shape is hard to print, so antenna datasheets slice it twice and draw each slice as a round <strong>polar plot</strong>:</p>
<ul>
<li><strong>Azimuth</strong> (also called the <em>H-plane</em>): the pattern sliced flat, seen <strong>from above</strong>. For an upright omni
it&rsquo;s a circle: the same all the way around.</li>
<li><strong>Elevation</strong> (the <em>E-plane</em>): the pattern sliced upright, seen <strong>from the side</strong>. For an upright dipole it&rsquo;s a
figure 8 on its side: strong toward the horizon, nothing straight up or down.</li>
</ul>
<p>The rings are usually in dB below the peak: the outer ring is the peak, then −10, −20, −30 dB. A shape that
reaches the outer ring in some direction is at full strength there; one that shrinks to the middle sends almost
nothing.</p>
<p>The 3D view below draws both: the full shape, and underneath it the two slices, the way a datasheet prints them.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Azimuth is the pattern from above; elevation is from the side. Rings are dB below the peak.</p>

  </div>
</section>


<section class="lab-stage lab-stage-follow" id="follow">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">3</span>
    <span class="lab-stage-kind mono">follow along</span><h2>Turn the patterns around</h2>
  </header>
  <div class="lab-stage-body">
    <link rel="stylesheet" href="/css/rflab.min.6470f6e7f26d78ca57cbe2d4ea3c5e64bc6a1cf89d42def3f2f99e04198472e4.css" integrity="sha256-ZHD25/JteMpXy&#43;LU6jxeZLxqHPidQt7z8vmeBBmEcuQ=">
<script src="/js/rflab-loader.d2c182aadbcb506f3c99779fa2c245f6721a9919c45d140ceb6c595cdea20b9c.js" integrity="sha256-0sGCqtvLUG88mXefosJF9nIamRnEXRQM62xZXN6iC5w=" defer></script>
<figure class="rflab" data-rflab data-mode="antenna" data-src="/js/rflab/main.1374b7b3a47b299b16ed801f63ad98980a6452563b2455f12aef6f7fd65a3a0d.js" data-integrity="sha256-E3S3s6R7KZsW7YAfY62YmApkUlY7JFXxKu9vf9ZaOg0=" data-start="click" aria-label="3D radiation pattern of an upright dipole antenna: a donut around the antenna, with a hole straight above and below it, and its azimuth and elevation cuts">
  <div class="rf-poster">
    <img src="/images/series/wireless-101/antenna-viewer_hu_7cef30385b6ae424.webp" width="1020" height="766" alt="3D radiation pattern of an upright dipole antenna: a donut around the antenna, with a hole straight above and below it, and its azimuth and elevation cuts" loading="lazy" decoding="async">
    <button class="btn primary rf-start" type="button" data-rflab-start>Start the 3D lab</button>
  </div>
  <div data-rflab-app></div>
  <script type="application/json" data-rflab-config>{"antenna":"dipole","mode":"antenna"}</script><figcaption><p>Press <strong>Start the 3D lab</strong>, then try these:</p>
<ol>
<li><strong>Dipole.</strong> Drag to turn it. It&rsquo;s a donut, with a hole straight up and straight down. Point at the hole: about
25 dB weaker than the horizon. Its azimuth slice is a circle; its elevation slice is a figure 8.</li>
<li><strong>Isotropic.</strong> A perfect ball, the 0 dBi yardstick. The faint sphere around every other pattern is this ball.</li>
<li><strong>High-gain omni.</strong> A flatter donut that pokes out farther along the horizon (6 dBi) and pulls in above and
below. Same power, different shape.</li>
<li><strong>Ceiling AP.</strong> Tipped down, strongest about 45° below the horizon, with only a shallow dip straight down.
That&rsquo;s the shape you want hanging from an office ceiling.</li>
<li><strong>Patch</strong> and <strong>Narrow beam.</strong> Beams along the orange boresight line (the direction the antenna points).
Point at the edge of each beam: 3 dB down at 32° off for the patch, 15° for the narrow beam.</li>
</ol>
<p>The 3D shape shows <strong>reach</strong>: how far the signal goes in each direction before it fades to the same level. The
flat slices are in dB, like a datasheet.</p>
</figcaption>
</figure>
<p>Now see what a donut&rsquo;s hole does on a real floor. This task hangs a dipole 20 ft up in the office from lesson 4.</p>
<link rel="stylesheet" href="/css/rflab.min.6470f6e7f26d78ca57cbe2d4ea3c5e64bc6a1cf89d42def3f2f99e04198472e4.css" integrity="sha256-ZHD25/JteMpXy&#43;LU6jxeZLxqHPidQt7z8vmeBBmEcuQ=">
<script src="/js/rflab-loader.d2c182aadbcb506f3c99779fa2c245f6721a9919c45d140ceb6c595cdea20b9c.js" integrity="sha256-0sGCqtvLUG88mXefosJF9nIamRnEXRQM62xZXN6iC5w=" defer></script>
<figure class="rflab" data-rflab data-mode="floor" data-src="/js/rflab/main.1374b7b3a47b299b16ed801f63ad98980a6452563b2455f12aef6f7fd65a3a0d.js" data-integrity="sha256-E3S3s6R7KZsW7YAfY62YmApkUlY7JFXxKu9vf9ZaOg0=" data-start="click" aria-label="Guided task: a dipole hung 20 ft up in the office; find the weakest spot within 10 ft of it">
  <div class="rf-poster">
    <img src="/images/series/wireless-101/signal-lab_hu_40697ca6ad62c097.webp" width="1020" height="682" alt="Guided task: a dipole hung 20 ft up in the office; find the weakest spot within 10 ft of it" loading="lazy" decoding="async">
    <button class="btn primary rf-start" type="button" data-rflab-start>Start the 3D lab</button>
  </div>
  <div data-rflab-app></div>
  <script type="application/json" data-rflab-config>{"antenna":"dipole","mode":"floor","scenario":"office","tasks":["null"]}</script><figcaption><p><strong>Find the hole under the AP.</strong> Put the laptop on the weakest spot within 10 ft of the access point. Click the
floor to put the laptop down, then press <strong>Check</strong>.</p>
</figcaption>
</figure>
<p class="lab-recap"><span class="mono">in one sentence</span> A dipole&rsquo;s hole points at the floor when you hang it from a ceiling.</p>

  </div>
</section>


<section class="lab-stage lab-stage-challenge" id="challenge">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">4</span>
    <span class="lab-stage-kind mono">challenge</span><h2>Which antenna?</h2>
  </header>
  <div class="lab-stage-body">
    <p>Pick an antenna for each, and say why:</p>
<ol>
<li>A long, narrow warehouse aisle, with the access point on the wall at one end.</li>
<li>An open-plan office, access points on the ceiling.</li>
<li>A hallway 120 ft long, the access point at one end.</li>
<li>A café, the access point on a shelf in the middle of the room.</li>
</ol>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>Ask two things each time: where&rsquo;s the access point, and what shape is the space you need to fill?</p>
</div>
</details>
<details class="lab-solution">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg>
 <span>show the solution</span></summary>
  <div class="lab-solution-body">
    <ol>
<li><strong>Narrow beam</strong> (Yagi or sector). The aisle is long and thin, so a 30° beam aimed down it puts the signal where
the people are and keeps it out of the neighboring aisles. Lesson 5 tries this.</li>
<li><strong>Ceiling AP antennas</strong>, aimed down. A plain dipole would point its hole at the people underneath.</li>
<li>A <strong>patch</strong>: a 65° beam from one end fills a hallway nicely. A narrow beam would also reach, but more of the
hallway near the access point would fall outside it.</li>
<li>An <strong>omni</strong> standing upright, like a dipole. The room is all round it on one level, which is exactly the donut&rsquo;s
strong ring.</li>
</ol>
  </div>
</details>

  </div>
</section>

<p>Back to <a href="/lessons/wireless-101/task-2-1/">lesson 2</a> for channels, or on to
<a href="/lessons/wireless-101/task-4-1/">lesson 4</a>, where the signal meets walls. EIRP from
<a href="/lessons/wireless-101/task-1-1/">lesson 1</a> is transmit power, minus the cable, plus exactly this gain.</p>
]]></content:encoded></item><item><title>Bands and channels: 2.4, 5 and 6 GHz</title><link>https://networkdoodles.com/lessons/wireless-101/task-2-1/</link><pubDate>Mon, 16 Nov 2026 09:00:00 -0500</pubDate><dc:creator>David Henderson</dc:creator><guid>https://networkdoodles.com/lessons/wireless-101/task-2-1/</guid><description>Wi-Fi's three bands, why 2.4 GHz has only channels 1, 6 and 11 clear, channel width trade-offs and DFS, with a channel planner you can try. Beginner lesson.</description><content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[
<section class="lab-stage lab-stage-explain" id="explain">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">1</span>
    <span class="lab-stage-kind mono">explain</span><h2>Why does Wi-Fi have channels?</h2>
  </header>
  <div class="lab-stage-body">
    <h3 id="lanes-on-a-road">Lanes on a road</h3>
<p>A band is a road, and channels are its lanes. If two cars try to use the same lane at the same moment, one has
to wait. If a car straddles two lanes, it gets in everyone&rsquo;s way. Wi-Fi&rsquo;s whole channel story is about keeping
access points that are near each other in lanes of their own.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> A band is the road; channels are its lanes.</p>
<h3 id="three-bands">Three bands</h3>
<p>Wi-Fi uses three bands. Each has a trade-off:</p>
<table>
  <thead>
      <tr>
          <th>Band</th>
          <th>In the US</th>
          <th>Good at</th>
          <th>Not so good at</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td><strong>2.4 GHz</strong></td>
          <td>2,400 to 2,483.5 MHz</td>
          <td>reaching farther and through walls</td>
          <td>only 3 clear channels; shared with Bluetooth and microwave ovens</td>
      </tr>
      <tr>
          <td><strong>5 GHz</strong></td>
          <td>about 5,150 to 5,850 MHz</td>
          <td>lots of channels; much less crowded</td>
          <td>a little shorter reach; some channels need DFS</td>
      </tr>
      <tr>
          <td><strong>6 GHz</strong></td>
          <td>5,925 to 7,125 MHz</td>
          <td>huge and empty: 59 channels at 20 MHz</td>
          <td>shortest reach; only Wi-Fi 6E and Wi-Fi 7 devices can use it</td>
      </tr>
  </tbody>
</table>
<p>Which devices can use which band depends on their <strong>Wi-Fi generation</strong>:</p>
<table>
  <thead>
      <tr>
          <th>Name</th>
          <th>Standard</th>
          <th>Bands</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td>Wi-Fi 4</td>
          <td>802.11n</td>
          <td>2.4 and 5 GHz</td>
      </tr>
      <tr>
          <td>Wi-Fi 5</td>
          <td>802.11ac</td>
          <td>5 GHz</td>
      </tr>
      <tr>
          <td>Wi-Fi 6</td>
          <td>802.11ax</td>
          <td>2.4 and 5 GHz</td>
      </tr>
      <tr>
          <td>Wi-Fi 6E</td>
          <td>802.11ax</td>
          <td>adds 6 GHz</td>
      </tr>
      <tr>
          <td>Wi-Fi 7</td>
          <td>802.11be</td>
          <td>2.4, 5 and 6 GHz, and wider channels</td>
      </tr>
  </tbody>
</table>
<p><em>802.11</em> is the family of IEEE standards behind Wi-Fi; the letters after it are the version.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> 2.4 GHz reaches farthest, 5 GHz has the room, 6 GHz is wide open but only for newer devices.</p>
<h3 id="channels-numbers-5-mhz-apart">Channels: numbers 5 MHz apart</h3>
<p>A channel number names a spot in the band, and the numbers are only <strong>5 MHz apart</strong>. On 2.4 GHz, channel 1 is
centered on 2,412 MHz, channel 2 on 2,417, channel 6 on 2,437, and channel 11 on 2,462.</p>
<p>But a Wi-Fi signal is about <strong>20 MHz wide</strong>. So channel 1 spreads from about 2,402 to 2,422 MHz, and channel 3
(2,422 MHz in the middle) lands right on top of half of it. To get completely clear of channel 1 you have to jump
to channel 6, and then to 11. That&rsquo;s why you&rsquo;ll hear <strong>1, 6 and 11</strong> over and over: in the US, they&rsquo;re the only
three 2.4 GHz channels that don&rsquo;t overlap.</p>
<p>On 5 and 6 GHz the channel numbers go up in steps of 4 (36, 40, 44…), which is exactly 20 MHz, so neighboring
20 MHz channels sit side by side without overlapping.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Channel numbers are 5 MHz apart but signals are 20 MHz wide. On 2.4 GHz only 1, 6 and 11 stand clear.</p>
<h3 id="sharing-a-channel-versus-overlapping">Sharing a channel versus overlapping</h3>
<p>Two access points near each other can relate in three ways:</p>
<ul>
<li><strong>Different, clear channels</strong>: they don&rsquo;t affect each other at all. Best.</li>
<li><strong>The same channel</strong>: Wi-Fi radios <em>listen before they talk</em>. Each one hears the other and waits its turn, so
they <strong>share the airtime</strong>: each gets roughly half. Not great, but orderly. Engineers call it
<strong>co-channel interference</strong>.</li>
<li><strong>Partly overlapping channels</strong> (like 1 and 3): each hears the other as <strong>noise</strong> it can&rsquo;t decode, so it doesn&rsquo;t
wait. They talk over each other, frames get garbled and resent, and both slow down. This is the worst case:
<strong>adjacent-channel interference</strong>.</li>
</ul>
<p class="lab-recap"><span class="mono">in one sentence</span> Same channel: they take turns. Overlapping channels: they trample each other. Overlap is worse.</p>
<h3 id="channel-width-faster-but-fewer">Channel width: faster, but fewer</h3>
<p>A radio can bond neighboring 20 MHz channels into one wider channel, <strong>40, 80 or 160 MHz</strong> (and 320 MHz on 6 GHz
with Wi-Fi 7). A channel twice as wide carries about twice as much data. But:</p>
<ul>
<li>it uses up the band twice as fast, so there are half as many clear channels to go around;</li>
<li>it catches twice as much noise, which raises the noise floor by 3 dB.</li>
</ul>
<table>
  <thead>
      <tr>
          <th>Width</th>
          <th>2.4 GHz</th>
          <th>5 GHz (US)</th>
          <th>6 GHz (US)</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td>20 MHz</td>
          <td>3</td>
          <td>25</td>
          <td>59</td>
      </tr>
      <tr>
          <td>40 MHz</td>
          <td>1, really</td>
          <td>12</td>
          <td>29</td>
      </tr>
      <tr>
          <td>80 MHz</td>
          <td>none</td>
          <td>6</td>
          <td>14</td>
      </tr>
      <tr>
          <td>160 MHz</td>
          <td>none</td>
          <td>2</td>
          <td>7</td>
      </tr>
  </tbody>
</table>
<p>In an office full of access points, <strong>20 or 40 MHz on 5 GHz</strong> is the usual choice: enough clear channels that
neighbors don&rsquo;t have to share. A home with one access point can happily use 80 MHz.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Wider channels are faster but fewer, and a little noisier. Busy places use narrower channels.</p>
<h3 id="dfs-sharing-with-radar">DFS: sharing with radar</h3>
<p>On 5 GHz, channels <strong>52 to 144</strong> sit in bands also used by weather and military radar. An access point may use
them only if it does <strong>DFS</strong> (<em>dynamic frequency selection</em>): it listens for radar for a while before it starts,
and if it ever hears radar, it has to leave the channel and stay off it for half an hour. Clients on that access
point get moved without warning. DFS channels more than double the clear 5 GHz channels, which is why most
designs use them, but near an airport or a weather station they&rsquo;re worth avoiding.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> DFS channels give 5 GHz most of its room, but an access point must give them up to radar.</p>

  </div>
</section>


<section class="lab-stage lab-stage-show" id="show">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">2</span>
    <span class="lab-stage-kind mono">show</span><h2>The channel strip</h2>
  </header>
  <div class="lab-stage-body">
    <p>The tool below draws a band as a strip of frequencies, from low on the left to high on the right, with tick
marks for each channel number. Each access point gets a lane with a colored block showing exactly which
frequencies its channel covers.</p>
<ul>
<li>A block with a <strong>red outline</strong> partly overlaps another access point&rsquo;s channel.</li>
<li>A block with a <strong>dashed amber outline</strong> shares a channel with another access point.</li>
<li>On 5 GHz, the shaded parts are the four sub-bands, and the amber ones need DFS.</li>
</ul>
<p>Under the strip, the tool counts how many channels of the chosen width fit without overlapping, and lists every
pair of access points that share or overlap.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> One lane per access point; red is overlap, amber is sharing.</p>

  </div>
</section>


<section class="lab-stage lab-stage-follow" id="follow">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">3</span>
    <span class="lab-stage-kind mono">follow along</span><h2>Fix six access points</h2>
  </header>
  <div class="lab-stage-body">
    <link rel="stylesheet" href="/css/rflab.min.6470f6e7f26d78ca57cbe2d4ea3c5e64bc6a1cf89d42def3f2f99e04198472e4.css" integrity="sha256-ZHD25/JteMpXy&#43;LU6jxeZLxqHPidQt7z8vmeBBmEcuQ=">
<script src="/js/rfwidgets.77d3a23a25f98a9ab3916711c8500f082c2d4eb4f4ff9254870cef0a3461f09e.js" integrity="sha256-d9OiOiX5ipqzkWcRyFAPCCwtTrT0/5JUhwzvCjRh8J4=" defer></script>
<figure class="rflab rfw" data-rfwidget="channels" aria-label="The channel strip: six access points placed on the 2.4, 5 or 6 GHz channels, with shared and overlapping channels marked">
  <div class="rfw-app" data-rfw-app><p class="rfw-muted">This tool needs JavaScript. The text below covers the same ideas.</p></div><figcaption><p>Try these, in order:</p>
<ol>
<li>The six access points start on 2.4 GHz on channels <strong>1, 3, 6, 6, 9 and 11</strong>. Six pairs partly overlap. Look at
AP 2 on channel 3: its block covers half of channel 1&rsquo;s and half of channel 6&rsquo;s.</li>
<li>Move AP 2 to <strong>11</strong> and AP 5 to <strong>1</strong>. Now every access point is on 1, 6 or 11. Nothing overlaps, but three
pairs share a channel: with six access points and three clear channels, some have to share. In a real
building you&rsquo;d make sure the ones sharing are far apart.</li>
<li>Switch to <strong>5 GHz</strong> at 20 MHz: 25 clear channels, plenty for six access points.</li>
<li>Change the width to <strong>80 MHz</strong>: 6 clear channels. Then <strong>160 MHz</strong>: only 2, and both include DFS channels.</li>
<li>Switch to <strong>6 GHz</strong>: 7 clear channels even at 160 MHz.</li>
</ol>
</figcaption>
</figure>
<p class="lab-recap"><span class="mono">in one sentence</span> On 2.4 GHz you run out of clear channels fast. On 5 and 6 GHz the width you pick decides how many you get.</p>

  </div>
</section>


<section class="lab-stage lab-stage-challenge" id="challenge">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">4</span>
    <span class="lab-stage-kind mono">challenge</span><h2>A row of six classrooms</h2>
  </header>
  <div class="lab-stage-body">
    <p>A school corridor has six classrooms in a row, one access point in each. Each access point can hear its
<strong>neighbors</strong> on either side, but not the ones two rooms away.</p>
<ol>
<li>Pick 2.4 GHz channels so that no two neighbors share or overlap.</li>
<li>Pick 5 GHz channels at <strong>40 MHz</strong> for the same six rooms. The school is next to a regional airport, so it
wants <strong>no DFS channels</strong>.</li>
</ol>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>On 2.4 GHz you only have 1, 6 and 11. The rule is only about <em>neighbors</em>: rooms two apart can reuse a channel.</p>
</div>
</details>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>Without DFS, 5 GHz keeps UNII-1 (36 to 48) and UNII-3 (149 to 165). At 40 MHz that&rsquo;s 36-40, 44-48, 149-153 and
157-161.</p>
</div>
</details>
<details class="lab-solution">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg>
 <span>show the solution</span></summary>
  <div class="lab-solution-body">
    <ol>
<li><strong>1, 6, 11, 1, 6, 11.</strong> Every neighbor is on a different clear channel. Rooms 1 and 4 share channel 1, but
they&rsquo;re three rooms apart and can&rsquo;t hear each other.</li>
<li><strong>36-40, 44-48, 149-153, 157-161, 36-40, 44-48.</strong> Four non-DFS 40 MHz channels for six rooms, so two get
reused, again only by rooms far apart.</li>
</ol>
<p>The pattern, reuse a channel only where the access points can&rsquo;t hear each other, is the heart of channel
planning. Lesson 5 shows how far an access point&rsquo;s signal actually reaches, which tells you how far apart
&ldquo;can&rsquo;t hear each other&rdquo; needs to be.</p>
  </div>
</details>

  </div>
</section>

<p>Back to <a href="/lessons/wireless-101/task-1-1/">lesson 1</a> for dB and dBm, or on to
<a href="/lessons/wireless-101/task-3-1/">lesson 3</a>, where the antenna decides which way the signal goes. The
listen-before-talk idea is close to the hub from <a href="/lessons/networking-101/task-2-1/">Networking 101&rsquo;s lesson 2.1</a>:
one shared medium, one talker at a time.</p>
]]></content:encoded></item><item><title>Radio basics: waves, dB and dBm</title><link>https://networkdoodles.com/lessons/wireless-101/task-1-1/</link><pubDate>Mon, 16 Nov 2026 09:00:00 -0500</pubDate><dc:creator>David Henderson</dc:creator><guid>https://networkdoodles.com/lessons/wireless-101/task-1-1/</guid><description>What a Wi-Fi signal is, why engineers count it in dB and dBm, and how to read −67 dBm and 25 dB SNR, with a link budget you can change. Beginner lesson.</description><content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[
<section class="lab-stage lab-stage-explain" id="explain">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">1</span>
    <span class="lab-stage-kind mono">explain</span><h2>What is a Wi-Fi signal?</h2>
  </header>
  <div class="lab-stage-body">
    <h3 id="ripples-in-a-pond">Ripples in a pond</h3>
<p>Drop a stone in a pond and ripples spread out in rings. They get weaker as they spread, because the same
energy is shared out over a bigger and bigger ring. They bounce off the bank, and they pass around a reed
but not through a rock.</p>
<p>A Wi-Fi signal is a ripple too, but in the electric and magnetic field around an antenna instead of in water.
It&rsquo;s a <strong>radio wave</strong>, the same kind of thing as FM radio, phone signal and the light from a lamp, just at a
different speed of wiggle. Everything this course covers is the pond: how the ripples spread, fade, bounce and
get blocked.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> A Wi-Fi signal is a radio wave spreading out from an antenna, like ripples in a pond.</p>
<h3 id="frequency-and-wavelength">Frequency and wavelength</h3>
<p>Two numbers describe a wave:</p>
<ul>
<li><strong>Frequency</strong>: how many times it wiggles each second. One wiggle a second is one <strong>hertz</strong> (Hz). Wi-Fi wiggles
billions of times a second: <strong>gigahertz</strong> (GHz). Wi-Fi uses three bands of frequencies, around <strong>2.4 GHz</strong>,
<strong>5 GHz</strong> and <strong>6 GHz</strong> (lesson 2 is all about them).</li>
<li><strong>Wavelength</strong>: how far apart the ripples are. Radio waves all travel at the speed of light, so the faster the
wiggle, the shorter the wave.</li>
</ul>
<table>
  <thead>
      <tr>
          <th>Band</th>
          <th>Frequency</th>
          <th>Wavelength</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td>2.4 GHz</td>
          <td>2,437 MHz (channel 6)</td>
          <td>12.3 cm, about 5 inches</td>
      </tr>
      <tr>
          <td>5 GHz</td>
          <td>5,180 MHz (channel 36)</td>
          <td>5.8 cm, about 2¼ inches</td>
      </tr>
      <tr>
          <td>6 GHz</td>
          <td>6,135 MHz (channel 37)</td>
          <td>4.9 cm, about 2 inches</td>
      </tr>
  </tbody>
</table>
<p>The wavelength matters more than you&rsquo;d think. An antenna is usually about half a wavelength long, which is why
5 GHz antennas are shorter. And a shorter wave is easier to stop with a wall, which you&rsquo;ll see in lesson 4.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Frequency is wiggles per second; wavelength is the distance between ripples. Higher frequency, shorter wave.</p>
<h3 id="why-decibels">Why decibels?</h3>
<p>An access point sends about a tenth of a watt, less than a night light. By the time it reaches a laptop across
the office, there&rsquo;s about <strong>0.00000001 milliwatts</strong> left: a hundred-millionth of a milliwatt. Numbers like that
are miserable to work with. So engineers use <strong>decibels</strong> (dB), which count <em>how many times bigger or smaller</em>:</p>
<ul>
<li><strong>+3 dB</strong> is <strong>twice</strong> the power; −3 dB is half.</li>
<li><strong>+10 dB</strong> is <strong>ten times</strong>; −10 dB is a tenth.</li>
<li>+20 dB is a hundred times, +30 dB a thousand.</li>
</ul>
<p>The magic is that multiplying becomes adding. An antenna that focuses the signal 2.3 times, and a wall that lets
through a third of it, become <em>+3.7 dB</em> and <em>−4.8 dB</em>. Add them up and you have the answer.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> dB counts &ldquo;how many times&rdquo;: +3 dB doubles, +10 dB is ten times, and you add instead of multiply.</p>
<h3 id="dbm-decibels-against-a-milliwatt">dBm: decibels against a milliwatt</h3>
<p>A plain dB is a <em>change</em>. To say how strong a signal <em>is</em>, you need a starting point. <strong>dBm</strong> means <em>decibels
compared with one milliwatt</em>:</p>
<table>
  <thead>
      <tr>
          <th>dBm</th>
          <th>Power</th>
          <th>Where you&rsquo;d see it</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td>30 dBm</td>
          <td>1 W</td>
          <td>the most the US allows at 2.4 GHz, before the antenna</td>
      </tr>
      <tr>
          <td>20 dBm</td>
          <td>100 mW</td>
          <td>a home router turned all the way up</td>
      </tr>
      <tr>
          <td>0 dBm</td>
          <td>1 mW</td>
          <td>the reference point</td>
      </tr>
      <tr>
          <td>−30 dBm</td>
          <td>1 µW (a millionth of a watt)</td>
          <td>a laptop right next to the access point</td>
      </tr>
      <tr>
          <td>−67 dBm</td>
          <td>about 0.2 nW</td>
          <td>the usual target for voice and video calls</td>
      </tr>
      <tr>
          <td>−95 dBm</td>
          <td>about 0.3 pW</td>
          <td>the <strong>noise floor</strong>: the radio hiss everywhere</td>
      </tr>
  </tbody>
</table>
<p>Signals at a laptop are almost always negative. <strong>Closer to zero is stronger</strong>: −50 dBm is a great signal,
−80 dBm is a poor one.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> dBm is power compared with 1 mW. At the laptop it&rsquo;s negative, and closer to zero is better.</p>
<h3 id="rssi-noise-and-snr">RSSI, noise and SNR</h3>
<p>Your laptop reports the strength it hears as <strong>RSSI</strong> (<em>received signal strength indicator</em>). On most laptops and
phones you can see it in dBm (lesson 4 shows how), though some show a percentage or bars instead, and every
maker scales those differently. dBm is the number engineers trust.</p>
<p>A strong signal isn&rsquo;t enough on its own; it has to stand out from the <strong>noise</strong>, the background hiss from other
electronics and from physics itself. The gap between them is the <strong>signal-to-noise ratio</strong> (<strong>SNR</strong>), and in dB
it&rsquo;s just a subtraction:</p>
<div class="highlight"><pre tabindex="0" class="chroma"><code class="language-text" data-lang="text"><span class="line"><span class="cl">SNR = signal − noise = −67 dBm − (−95 dBm) = 28 dB
</span></span></code></pre></div><p>The more SNR, the faster the connection can go, because the radio can pack more bits into each wiggle without
mistaking noise for data. Around <strong>25 dB</strong> or more is comfortable for voice and video calls.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> RSSI is how loud the access point is; SNR is how much louder than the noise. Aim for −67 dBm and 25 dB.</p>

  </div>
</section>


<section class="lab-stage lab-stage-show" id="show">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">2</span>
    <span class="lab-stage-kind mono">show</span><h2>From the access point to the laptop</h2>
  </header>
  <div class="lab-stage-body">
    <p>Every Wi-Fi link is a chain of gains and losses. Engineers call the sum a <strong>link budget</strong>:</p>
<table>
  <thead>
      <tr>
          <th>Step</th>
          <th>What happens</th>
          <th>Typical</th>
      </tr>
  </thead>
  <tbody>
      <tr>
          <td><strong>Transmit power</strong></td>
          <td>what the access point&rsquo;s radio puts out</td>
          <td>14 to 20 dBm</td>
      </tr>
      <tr>
          <td><strong>Cable</strong></td>
          <td>a little lost between the radio and the antenna</td>
          <td>−1 dB</td>
      </tr>
      <tr>
          <td><strong>Antenna gain</strong></td>
          <td>the antenna focuses the signal toward you (lesson 3)</td>
          <td>+2 to +10 dBi</td>
      </tr>
      <tr>
          <td><strong>The air</strong></td>
          <td>the signal spreads out over distance (lesson 4)</td>
          <td>−60 to −80 dB</td>
      </tr>
      <tr>
          <td><strong>Walls</strong></td>
          <td>each one takes a share (lesson 4)</td>
          <td>−3 dB to −30 dB each</td>
      </tr>
      <tr>
          <td><strong>The laptop&rsquo;s antenna</strong></td>
          <td>small, and not pointed anywhere in particular</td>
          <td>about 0 dBi</td>
      </tr>
  </tbody>
</table>
<p>Two words worth knowing from that table:</p>
<ul>
<li><strong>dBi</strong>: an antenna&rsquo;s gain, compared with an imaginary antenna that sends the same amount in every direction.
Lesson 3 explains it.</li>
<li><strong>EIRP</strong> (<em>effective isotropic radiated power</em>): transmit power, minus the cable, plus the antenna gain. It&rsquo;s
what actually leaves the antenna in its best direction, and it&rsquo;s what regulators limit.</li>
</ul>
<p>Start at the transmit power, add the gains, subtract the losses, and what&rsquo;s left is the RSSI at the laptop.</p>
<p class="lab-recap"><span class="mono">in one sentence</span> Signal at the laptop = transmit power − cable + antenna gain − air − walls. That&rsquo;s a link budget.</p>

  </div>
</section>


<section class="lab-stage lab-stage-follow" id="follow">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">3</span>
    <span class="lab-stage-kind mono">follow along</span><h2>Climb down the dB ladder</h2>
  </header>
  <div class="lab-stage-body">
    <p>The top half converts between mW and dBm. The bottom half is a link budget, drawn as a staircase from the
access point down to the laptop.</p>
<link rel="stylesheet" href="/css/rflab.min.6470f6e7f26d78ca57cbe2d4ea3c5e64bc6a1cf89d42def3f2f99e04198472e4.css" integrity="sha256-ZHD25/JteMpXy&#43;LU6jxeZLxqHPidQt7z8vmeBBmEcuQ=">
<script src="/js/rfwidgets.77d3a23a25f98a9ab3916711c8500f082c2d4eb4f4ff9254870cef0a3461f09e.js" integrity="sha256-d9OiOiX5ipqzkWcRyFAPCCwtTrT0/5JUhwzvCjRh8J4=" defer></script>
<figure class="rflab rfw" data-rfwidget="ladder" aria-label="The dB ladder: a power converter between mW and dBm, and a link budget from an access point to a laptop drawn as a staircase of gains and losses">
  <div class="rfw-app" data-rfw-app><p class="rfw-muted">This tool needs JavaScript. The text below covers the same ideas.</p></div><figcaption><p>Try these, in order:</p>
<ol>
<li>Set the power to <strong>20 dBm</strong> (100 mW). Press <strong>+3 dB</strong>: 200 mW. Press it again: 400 mW. Now <strong>+10 dB</strong> from
20 dBm: 1,000 mW, a whole watt.</li>
<li>In the link budget, the default is a 17 dBm access point with 1 dB of cable and a ceiling antenna (+3.7 dBi),
40 ft away on 5 GHz. The laptop gets about <strong>−49 dBm</strong>: an excellent signal.</li>
<li>Drag the distance to <strong>80 ft</strong>. The signal drops by about <strong>6 dB</strong>: twice the distance costs 6 dB in open air.</li>
<li>Add a <strong>drywall</strong> wall: −4 dB. Add <strong>poured concrete</strong>: about −26 dB. Watch the staircase fall below the green
−67 dBm line.</li>
<li>Switch the band to <strong>2.4 GHz</strong>. The signal climbs a long way: the air costs about 6.5 dB less, and the
concrete about 11 dB less. Lesson 4 explains why.</li>
</ol>
</figcaption>
</figure>
<p class="lab-recap"><span class="mono">in one sentence</span> Each step on the ladder is a dB you can add or subtract. The staircase ends at the RSSI.</p>

  </div>
</section>


<section class="lab-stage lab-stage-challenge" id="challenge">
  <header class="lab-stage-head">
    <span class="lab-stage-n mono">4</span>
    <span class="lab-stage-kind mono">challenge</span><h2>Is it good enough for a video call?</h2>
  </header>
  <div class="lab-stage-body">
    <p>A café&rsquo;s access point sends <strong>20 dBm</strong> through <strong>2 dB</strong> of cable into a <strong>6 dBi</strong> omni antenna. A customer sits
where the air costs <strong>75 dB</strong> and one <strong>brick</strong> wall (6 dB on 2.4 GHz) is in the way. Their laptop&rsquo;s antenna counts
as 0 dBi, and the noise floor is −95 dBm.</p>
<ol>
<li>What&rsquo;s the access point&rsquo;s EIRP?</li>
<li>What&rsquo;s the RSSI at the laptop?</li>
<li>What&rsquo;s the SNR, and is it good enough for a video call?</li>
<li>How many dB would the café need to find to hit −67 dBm?</li>
</ol>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>EIRP is transmit power, minus the cable, plus the antenna gain.</p>
</div>
</details>
<details class="lab-hint">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m20 20-4-4"/></svg>
 <span>hint</span></summary>
  <div class="lab-hint-body"><p>RSSI is the EIRP, minus the air, minus the walls, plus the laptop&rsquo;s antenna gain.</p>
</div>
</details>
<details class="lab-solution">
  <summary class="mono"><svg class="ico" viewBox="0 0 24 24" width="18" height="18" fill="none" stroke="currentColor" stroke-width="1.7" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg>
 <span>show the solution</span></summary>
  <div class="lab-solution-body">
    <ol>
<li>EIRP = 20 − 2 + 6 = <strong>24 dBm</strong> (about 250 mW).</li>
<li>RSSI = 24 − 75 − 6 + 0 = <strong>−57 dBm</strong>.</li>
<li>SNR = −57 − (−95) = <strong>38 dB</strong>. Comfortably good enough for a video call.</li>
<li>None: −57 dBm is already <strong>10 dB better</strong> than the −67 dBm target. They have headroom for a second wall, or a
few more tables farther away.</li>
</ol>
<p>Turning the power up isn&rsquo;t always the answer anyway. The laptop has to answer back with its own, weaker radio,
and lesson 5 explains why a loud access point and a quiet laptop make a lopsided conversation.</p>
  </div>
</details>

  </div>
</section>

<p>Next: <a href="/lessons/wireless-101/task-2-1/">lesson 2</a> cuts the bands into channels. If any of the network words here
are new, <a href="/lessons/networking-101/task-1-1/">Networking 101&rsquo;s lesson on hosts</a> is the place to start.</p>
]]></content:encoded></item></channel></rss>