<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
     xmlns:atom="http://www.w3.org/2005/Atom"
     xmlns:content="http://purl.org/rss/1.0/modules/content/"
     xmlns:media="http://search.yahoo.com/mrss/">
  <channel>
    <title>Joonyoung Park</title>
    <description>Welcome to Joon&apos;s blog where I share my thoughts on technology, programming, and life.</description>
    <link>https://joon.upthere.ai</link>
    <language>en-us</language>
    <atom:link href="https://joon.upthere.ai/rss.xml" rel="self" type="application/rss+xml" />
    <lastBuildDate>Fri, 13 Nov 2026 04:09:00 GMT</lastBuildDate>
    
    <item>
      <title>ATSC 3.0 + 5G Broadcast on the Same RF: How the Knobs Actually Work</title>
      <link>https://joon.upthere.ai/2026/05/11/atsc3-5gb-tdm-configuration/</link>
      <description>A practical, field-tested guide to time-multiplexing ATSC 3.0 with 5G Broadcast on a single RF channel — three knobs, two boxes, and what makes them agree.</description>
      <content:encoded><![CDATA[<blockquote>
<p>A practical, field-tested guide to time-multiplexing ATSC 3.0 with 5G Broadcast on a single RF channel. Written for broadcast engineers, system integrators and 3GPP-side colleagues who want a quick mental model before diving into A/321, A/322, or 3GPP TS 38.331. DigiCaster Scheduler (DCS) is used as the reference implementation throughout.</p>
</blockquote>
<h2 id="why-this-is-suddenly-a-question" tabindex="-1">Why this is suddenly a question</h2>
<p>Two things changed in 2024–2025. <strong>5G Broadcast Rel-19</strong> finalised the <em>CAS-muting</em> feature: the 5GB cell can be told to stay quiet on its <strong>Cell Acquisition Subframes</strong> (PSS/SSS/PBCH) for a programmed pattern of subframes per cycle, freeing the RF for someone else to transmit during those gaps. And <strong>ATSC 3.0</strong> has always supported coexistence with non-3.0 signals on the same channel, by design — A/321 §6 lets a bootstrap announce, in a single 5-bit field, “the next ATSC frame won’t arrive sooner than X milliseconds.”</p>
<p>Put those two together and you have a credible path to a single 6 MHz (or 5 MHz, or 7/8 MHz) RF channel that carries both ATSC 3.0 and 5G Broadcast, time-sliced, with both receiver populations happy. Field trials are running now; the Las Vegas station is one of them.</p>
<p>The catch: getting the timing right requires three independent knobs to agree, and two of them live in <em>different boxes operated by different teams</em>. This brief lays out the knobs, what their valid values are, and how to set them in a real product.</p>
<h2 id="the-picture" tabindex="-1">The picture</h2>
<p>One TDM cycle, looking at the RF over time:</p>
<p><img src="/images/atsc3-tdm-cycle.svg" alt="One TDM cycle: ATSC 80 ms frames bracketing a 1200 ms 5G Broadcast CAS-active window"></p>
<p>Three knobs, three owners:</p>
<table>
<thead>
<tr>
<th>Knob</th>
<th>Where it lives</th>
<th>Owner</th>
<th>Quantum</th>
</tr>
</thead>
<tbody>
<tr>
<td>CAS-muting cycle (active + inactive duration)</td>
<td>5G Broadcast scheduler / 3GPP NR-MBS layer</td>
<td>5GB operator</td>
<td><strong>1 ms</strong> (NR subframe)</td>
</tr>
<tr>
<td>ATSC 3.0 frame duration</td>
<td>ATSC scheduler (e.g. DCS) → carried in L1-Basic preamble</td>
<td>ATSC operator</td>
<td><strong>5 ms</strong> (when locked to time)</td>
</tr>
<tr>
<td><code>min_time_to_next</code> (next-frame promise)</td>
<td>ATSC bootstrap (A/321)</td>
<td>ATSC operator</td>
<td><strong>50 / 100 / 200 / 500 / 1000 ms</strong> sliding</td>
</tr>
</tbody>
</table>
<p>The pattern is stable when <strong>CAS cycle ≈ ATSC frame interval</strong>, and <code>min_time_to_next</code> is the <strong>smallest legal value ≥ CAS cycle</strong>.</p>
<h2 id="what-the-spec-actually-constrains" tabindex="-1">What the spec actually constrains</h2>
<h3 id="knob-2-%E2%80%94-atsc-frame-duration" tabindex="-1">Knob 2 — ATSC frame duration</h3>
<p>A/322 §9.2 puts the L1-Basic preamble in one of two modes via a single bit, <code>L1B_frame_length_mode</code>:</p>
<ul>
<li><strong>Symbol-aligned</strong> (mode = 1, the common DCS factory default). Frame ends on whatever OFDM symbol boundary the FFT/GI/symbol-count math produces. You don’t pick the duration; you read it. A typical 6 MHz / 8K FFT / GI 1024 / 54-symbol subframe configuration produces a frame about <strong>76.67 ms</strong> long. The spec carries <code>L1B_time_offset</code> (16 bits) so the receiver can recover where the nearest millisecond boundary fell.</li>
<li><strong>Time-aligned</strong> (mode = 0). Frame duration is <em>chosen</em> by the operator and signalled in <code>L1B_frame_length</code> — a <strong>10-bit</strong> field in <strong>5 ms units</strong>, with legal range 10–1000 (so 50 ms to 5000 ms in 5 ms steps). The natural OFDM symbol math almost never lands on a 5-ms grid, so the leftover slack is redistributed across every non-preamble symbol’s guard interval via <code>L1B_excess_samples_per_symbol</code> (13 bits). Receivers stay aligned because every symbol grew by the same number of samples.</li>
</ul>
<p>For TDM with 5GB, you almost always want <strong>time-aligned mode</strong>. The 5 ms quantum is the smallest control you have; commit to it.</p>
<h3 id="knob-3-%E2%80%94-min_time_to_next" tabindex="-1">Knob 3 — <code>min_time_to_next</code></h3>
<p>A/321 §6, Table 6.3. Five bits, sliding-scale resolution, signalled in <em>Bootstrap Symbol 1</em> of every ATSC frame. The semantics are:</p>
<blockquote>
<p>bootstrap A promises that the next bootstrap (B) will arrive <strong>no earlier than T(X)</strong> ms and <strong>no later than T(X+1)</strong> ms, where X is the 5-bit value and T is the lookup table.</p>
</blockquote>
<p>The table is dense at the low end (50 ms steps from index 0 to 6) and gets sparser as values grow. Around the values most relevant to 5GB CAS cycles:</p>
<pre><code>  index 14  →  1100 ms   (window upper edge 1200)
  index 15  →  1200 ms   (window upper edge 1300)
  index 16  →  1300 ms   (window upper edge 1500)   ← granularity jumps to 200
  index 17  →  1500 ms
  ...
  index 31  →  RESERVED, MUST NOT be signalled
</code></pre>
<p>Two practical consequences:</p>
<ol>
<li>The spec lets you write <strong>any</strong> value, but the <em>receiver-perceived window</em> widens as you climb the scale. Past index 16 every step opens a 200 ms slack on the late side; past ~index 22 it’s 500 ms; etc. That slack is what absorbs cycle drift between the 1-ms-quantised 3GPP grid and the ATSC-side cadence.</li>
<li>The scale has <strong>gaps</strong> in the range that matters most for 5GB Phase-2 patterns (1200 ↔ 1300 ↔ 1500). If your CAS cycle lands in a gap (e.g. 1280 ms) you cannot signal it exactly; you pick the nearest legal value <em>above</em> the cycle and accept that some emission windows will have no ATSC frame ready. Your transmitter fills those with plain OFDM to keep the AGC happy.</li>
</ol>
<h2 id="a-four-step-recipe" tabindex="-1">A four-step recipe</h2>
<ol>
<li><strong>Find the 5GB inactive window</strong> in milliseconds. From the 3GPP parametrisation (K_CAS, N_CAS, etc.), compute the two numbers: cycle period and inactive duration. Both are integer milliseconds.</li>
<li><strong>Set the ATSC frame to fit the inactive window.</strong> Round the inactive duration up to the nearest 5 ms. That value goes in <code>L1B_frame_length</code> (in 5-ms units).</li>
<li><strong>Pick <code>min_time_to_next</code> from A/321 Table 6.3</strong> — the smallest table value that is <strong>strictly greater than</strong> the CAS cycle. If the cycle is 1280 ms, that’s index 16 (1300 ms).</li>
<li><strong>Sanity-check the inequality.</strong> <code>min_time_to_next</code> − <code>L1B_frame_length</code> must be ≥ inactive window. (If it isn’t, the bootstrap promise overlaps your own current frame, which is meaningless.)</li>
</ol>
<p>When <code>min_time_to_next</code> &gt; CAS cycle, the emission rate exceeds the production rate and the gateway buffer drains; the system is stable and the “too few ATSC frames” condition shows up as occasional OFDM-filler windows. When <code>min_time_to_next</code> &lt; CAS cycle, frames pile up and the transmitter restarts on overflow — typical failure mode is a periodic restart, on the order of tens of minutes, set by the buffer depth.</p>
<h2 id="configuring-digicaster" tabindex="-1">Configuring DigiCaster</h2>
<p>DCS exposes the two ATSC-side knobs in two places under <strong>PHYSICAL SETTING → Preamble → L1 Basic Parameters</strong>:</p>
<p><img src="/images/atsc3-dcs-ui-panel.svg" alt="DigiCaster L1 Basic Parameters panel: Frame Length mode, Frame Length, Custom Frame Time, min_time_to_next"></p>
<p>The mapping back to the spec is one-for-one:</p>
<ul>
<li><strong>Frame Length mode</strong> is <code>L1B_frame_length_mode</code> (1 = Symbol, 0 = Time).</li>
<li><strong>Frame Length [ms]</strong> is <code>L1B_frame_length × 5 ms</code>. DCS rejects anything that isn’t a multiple of 5 — that’s the spec’s 10-bit-in-5-ms-units encoding showing through, not a DCS limitation.</li>
<li><strong>Custom Frame Time = yes</strong> lets the bootstrap override the natural ATSC frame cadence with a chosen <code>min_time_to_next</code> value.</li>
<li><strong>min_time_to_next</strong> is exactly the A/321 5-bit field, expressed in milliseconds; pick from Table 6.3.</li>
</ul>
<p>The recipe in DCS terms: switch <em>Frame Length mode</em> to Time, type the inactive-window value (rounded up to a multiple of 5), set <em>Custom Frame Time</em> to yes, type the chosen <code>min_time_to_next</code>, click Apply.</p>
<h2 id="a-field-example%3A-the-las-vegas-trial" tabindex="-1">A field example: the Las Vegas trial</h2>
<ul>
<li><strong>Channel:</strong> 6 MHz, single RF.</li>
<li><strong>Payloads:</strong> ATSC 3.0 ≈ 537 kbps; 5G Broadcast ≈ 4.45 Mbps on a 5 MHz inner bandwidth.</li>
<li><strong>3GPP CAS-muting:</strong> K_CAS = 30, N_CAS = 8 → 1200 ms CAS-active + 80 ms inactive = <strong>1280 ms cycle</strong>.</li>
<li><strong>ATSC frame length:</strong> symbol-aligned natural 76.67 ms; time-aligned recommended <strong>80 ms</strong> (= 16 × 5 ms, smallest 5-ms multiple ≥ 76.67).</li>
<li><strong><code>min_time_to_next</code>:</strong> <strong>1300 ms</strong> (A/321 index 16). Was 1200 ms initially, which produced 80 ms-per-cycle of buffer growth and a ~17-minute periodic transmitter restart; the bump to 1300 reversed the inequality and the system has been stable since.</li>
<li><strong>Steady-state cost:</strong> the 20 ms / 1280 ms ≈ 1.5 % of cycles where no ATSC frame is queued get filled with plain OFDM by the exciter to maintain AGC. Sony chipset and EiTV ATSC 3.0 Analyzer Box receive cleanly throughout.</li>
</ul>
<p>If Castanet 5G can shift K_CAS to make the CAS cycle land at exactly 1300 ms (e.g. 1220 ms active + 80 ms inactive), the OFDM-filler events disappear and the system runs in perfect lock-step. That is the long-term target.</p>
<h2 id="where-the-standards-process-landed" tabindex="-1">Where the standards process landed</h2>
<p>As of <strong>A/327:2026-04</strong> (published 14 April 2026), this is no longer pending work. The amendment that started life as the S32 working draft <em>A/327 Amendment — TDM of ATSC with Other Broadcasting Systems</em> (2024-03-26) is now folded into the revised Recommended Practice itself. A/327 — <em>Guidelines for the Physical Layer Protocol</em> — is the authoritative reading for TDM coexistence; treat this brief as a practitioner-level walkthrough that points back to it, not as a substitute. Anything in here that disagrees with A/327:2026-04 should be assumed wrong, not the other way around.</p>
<h2 id="further-reading" tabindex="-1">Further reading</h2>
<ul>
<li><strong>ATSC A/321:2025-07</strong> <em>System Discovery and Signaling</em>, §6, Table 6.3 — <code>min_time_to_next</code> 5-bit field and its sliding scale.</li>
<li><strong>ATSC A/322:2025-07</strong> <em>Physical Layer Protocol</em>, §9.2 — L1-Basic preamble syntax, <code>L1B_frame_length_mode</code>, and the time-aligned vs. symbol-aligned branches.</li>
<li><strong>ATSC A/327:2026-04</strong> <em>Guidelines for the Physical Layer Protocol</em> (Recommended Practice, published 14 April 2026) — the authoritative reference for TDM of ATSC 3.0 with other broadcasting systems.</li>
<li><strong>3GPP TS 38.331 (Rel-19)</strong> — CAS-muting parametrisation on the 5GB side.</li>
</ul>
]]></content:encoded>
      <pubDate>Mon, 11 May 2026 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2026/05/11/atsc3-5gb-tdm-configuration/</guid>
      <category>broadcast</category>
      <category>ATSC-3.0</category>
      <category>5G-Broadcast</category>
      <category>DigiCaster</category>
      <category>TDM</category>
      <media:content url="https://joon.upthere.ai/images/atsc3-5gb-tdm-cover.jpg" medium="image" />
    </item>
    <item>
      <title>Broadcast Once, Deliver Everywhere: SES as an Operational Bet</title>
      <link>https://joon.upthere.ai/2026/04/27/ses-broadcast-once-deliver-everywhere/</link>
      <description>How DigiCAP&apos;s Signaling Export Service turns ATSC 3.0 broadcast into operational infrastructure — one signaling source, two delivery paths, no drift.</description>
      <content:encoded><![CDATA[<p>ONE Media’s vision is bold: a single ATSC 3.0 broadcast that reaches every screen — TVs, tablets, phones — without a separate streaming workflow. DigiCAP’s Signaling Export Service (SES) is what makes that possible, and we’ve demonstrated it at NAB 2026 with Sinclair.</p>
<p>Before I get to what SES does, look at this chart.</p>
<p><img src="/images/strassmann_management_vs_operational_chart.png" alt="Where the IT budget goes, by company success tier"></p>
<p>It’s from Paul Strassmann’s <a href="https://archive.org/details/businessvalueofc0000stra"><em>The Business Value of Computers</em> (1990)</a> — the book Steve Jobs said had stung him personally. Strassmann surveyed hundreds of companies and found that <em>total</em> IT spending didn’t correlate with profitability at all. But the <strong>split</strong> did. The least successful companies put 70% of their IT budget into <strong>management productivity</strong> — PCs, spreadsheets, email, the tools that help white-collar staff move paper around faster. The most successful ones inverted it: 70% on <strong>operational productivity</strong>, the software that runs the actual core of the business.</p>
<p>Jobs’ line on this, from his <a href="https://youtu.be/Gk-9Fd2mEnI?si=6UYS6rBeCyLGfVUM">1992 MIT Sloan talk</a>:</p>
<blockquote>
<p>“You can’t go down to your local computer store and buy an app that will help you do stock trading, or will help you run a hospital, or will help you in whatever operational part of your business you want to automate.”</p>
</blockquote>
<p>Operational software has to be custom-built for the shape of the business. That’s why most companies default to management productivity — it’s purchasable. Operational is harder, slower, and where the winners spend.</p>
<h2 id="what-this-looks-like-for-a-broadcaster" tabindex="-1">What this looks like for a broadcaster</h2>
<p>For a station group like Sinclair, the operational core is delivery — getting programs from the studio onto a viewer’s screen, reliably, everywhere. Management productivity for a broadcaster is the suite of tools that runs the office. Useful, but not what differentiates the business.</p>
<p>The traditional way to extend broadcast reach to non-tuner devices has been to bolt on a streaming pipeline: a parallel encoder, a separate packager, a second CDN, and two sets of service metadata that quietly drift apart. That’s overhead, not operational transformation.</p>
<p>SES is the operational alternative. It takes the signaling DigiCaster already produces — the same channels, schedules, and service descriptions broadcast over the air — and publishes it to the cloud, continuously, on the same standard (ATSC A/331) any compliant receiver already understands. One signaling source. One service catalog. Two delivery paths. No drift.</p>
<h2 id="why-it-counts-as-operational" tabindex="-1">Why it counts as operational</h2>
<p><a href="https://www.linkedin.com/posts/joonyoung-park-6478a24_one-media-technologiess-vision-is-bold-activity-7451973414912925696-9wb6?utm_source=share&amp;utm_medium=member_desktop&amp;rcm=ACoAAADX2xcB4aPiZfn2KrIJ0q0tLP4irna3ygM">SES</a> (Signaling Export Service) isn’t an app you can order off a shelf. It’s custom-built for how a NextGen TV plant actually operates — DigiCaster’s data model, ATSC 3.0’s signaling envelopes, real-world failover between RF and IP. That’s the multi-year, executive-patience work Strassmann was pointing at.</p>
<p>Sinclair’s bet on ATSC 3.0 has always been a bet on broadcasting becoming national IP infrastructure rather than local RF. SES is one of the operational tools that makes that bet payable.</p>
]]></content:encoded>
      <pubDate>Mon, 27 Apr 2026 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2026/04/27/ses-broadcast-once-deliver-everywhere/</guid>
      <category>broadcast</category>
      <category>ATSC-3.0</category>
      <category>DigiCAP</category>
      <category>SES</category>
      <category>NAB</category>
      <media:content url="https://joon.upthere.ai/images/strassmann_management_vs_operational_chart.png" medium="image" />
    </item>
    <item>
      <title>Television&apos;s Centennial Secret: The Infrastructure That Refuses to Die</title>
      <link>https://joon.upthere.ai/2026/01/27/television-centennial/</link>
      <description>100 years of broadcast television—from Baird&apos;s first demo to ATSC 3.0&apos;s dual-use future</description>
      <content:encoded><![CDATA[<p>On January 26, 1926, John Logie Baird demonstrated television to a room of skeptical journalists in London’s Soho. The Times buried the story on page 9.</p>
<p>One hundred years later, broadcast television infrastructure remains—outlasting radio’s dominance, surviving cable, weathering the internet, and now coexisting with streaming.</p>
<p>The secret? Broadcast never stopped evolving. From Baird’s 30-line mechanical scanner to the BBC’s 405-line service in 1936. From black-and-white to color. From analog to digital. Each generation predicted broadcast’s death. Each time, the infrastructure adapted.</p>
<p>ATSC 3.0 represents the centennial evolution—and perhaps the most significant pivot yet. This is not just better television. It is infrastructure that <em>senses</em>.</p>
<p>The same transmitters delivering 4K entertainment now illuminate the sky for passive radar systems. Research shows ATSC 3.0’s signal structure enables detection ranges eight times greater than previous standards. Baird wanted us to receive images. A century later, those signals bounce off aircraft and return information.</p>
<p>Broadcast’s survival secret: remaining useful in ways its inventors never imagined. The signal keeps giving.</p>
<hr>
<p><em>Sources: <a href="https://www.bbc.co.uk/historyofthebbc/tv-100/">BBC History of Television</a>, <a href="https://www.strath.ac.uk/whystrathclyde/news/2026/fromstrathclydetotheworld100yearssincetelevisionsfirstdemonstration/">University of Strathclyde</a></em></p>
]]></content:encoded>
      <pubDate>Tue, 27 Jan 2026 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2026/01/27/television-centennial/</guid>
      <category>broadcast</category>
      <category>ATSC-3.0</category>
      <category>television</category>
      <category>passive-radar</category>
      <media:content url="https://joon.upthere.ai/images/John_Logie_Baird_1600x600.gif" medium="image" />
    </item>
    <item>
      <title>Paper Accepted: ATSC 3.0 Passive Radar at SPIE Defense + Security 2026</title>
      <link>https://joon.upthere.ai/2026/01/08/spie-passive-radar-accepted/</link>
      <description>Paper on ATSC 3.0 passive radar accepted at SPIE Defense + Security 2026</description>
      <content:encoded><![CDATA[<p>I am pleased to share that my paper, <em>“Understanding the signal structure in ATSC 3.0 signals for passive radar detection,”</em> has been accepted for presentation at SPIE Defense + Security 2026.</p>
<p>The paper will be presented at the Signal Processing, Sensor/Information Fusion, and Target Recognition XXXV conference in National Harbor this April.</p>
<p>This work builds on the passive radar concepts I have been writing about in this <a href="/2025/11/24/passive-radar-introduction/">blog series</a>. It provides a comprehensive analysis of how ATSC 3.0’s signal structure—the same broadcasts delivering next-generation television—can be optimized for passive radar applications.</p>
<p>The research examines ATSC 3.0’s ambiguity function characteristics and introduces mismatched filter techniques that significantly reduce range ambiguities compared to existing DVB-T approaches.</p>
<p>For those following the passive radar series here, this paper represents the technical foundation behind what I have been describing. The conference presentation will be an opportunity to discuss these findings with the defense and signal processing community.</p>
<p>More details to come as the conference approaches.</p>
]]></content:encoded>
      <pubDate>Thu, 08 Jan 2026 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2026/01/08/spie-passive-radar-accepted/</guid>
      <category>passive-radar</category>
      <category>ATSC-3.0</category>
      <category>research</category>
      <category>SPIE</category>
      <media:content url="https://joon.upthere.ai/images/SPIE.png" medium="image" />
    </item>
    <item>
      <title>Signals of Opportunity: The Illuminators Powering Passive Radar (Part 2 of 4)</title>
      <link>https://joon.upthere.ai/2025/12/15/signals-of-opportunity/</link>
      <description>Why broadcast TV signals emerged as the gold standard for passive radar illumination</description>
      <content:encoded><![CDATA[<p><em>This is Part 2 of a 4-part series on passive radar. See <a href="/2025/11/24/passive-radar-introduction/">Part 1: Introduction to Passive Radar</a> for the foundation.</em></p>
<hr>
<h2 id="what-can-we-borrow%3F" tabindex="-1">What Can We Borrow?</h2>
<p>Passive radar systems need existing transmitters to illuminate targets—what engineers call “emitters of opportunity.” The options are surprisingly diverse: FM radio stations, digital television broadcasts (DVB-T), cellular base stations, GPS satellites, and even low-earth orbit constellations like Starlink.</p>
<p>Each signal type offers different characteristics. The question becomes: which signals work best for detecting aircraft or drone?</p>
<h2 id="the-ideal-illuminator" tabindex="-1">The Ideal Illuminator</h2>
<p>Four factors determine how well a signal performs for passive radar:</p>
<p><strong>High power</strong> means longer detection range. FM stations can transmit at 250 kW, while cellular towers operate around 100 W—a 2,500× difference that translates directly into coverage.</p>
<p><strong>Wide bandwidth</strong> enables better range resolution. DVB-T’s 6-8 MHz bandwidth can distinguish targets separated by just 25 meters. FM’s narrow bandwidth cannot.</p>
<p><strong>Continuous transmission</strong> ensures 24/7 surveillance capability. Broadcast signals never stop. Cellular traffic fluctuates with demand.</p>
<p><strong>Favorable signal structure</strong> matters for extracting weak echoes. OFDM signals—used in digital TV—have mathematical properties that make target detection more reliable than analog transmissions.</p>
<h2 id="why-broadcast-tv-dominates" tabindex="-1">Why Broadcast TV Dominates</h2>
<p>Digital television emerged as the gold standard for passive radar. DVB-T combines all four requirements: wide bandwidth for resolution, OFDM structure for reliable detection, high transmit power for range, and continuous operation for persistent coverage.</p>
<p>This is not theoretical. Operational systems already exploit broadcast TV: Hensoldt’s TwinVis in Germany, ERA’s Vera-NG in the Czech Republic, and Silentium Defence’s Maverick in Australia—all using DVB-T. The United States and South Korea, now transitioning to ATSC 3.0, present unexplored territory.</p>
<p>But DVB-T has a limitation. Its pilot structure—the reference symbols broadcast engineers use for channel estimation—creates a side effect: false targets appearing every 12.8 km in range. This ambiguity constrains how far a single-transmitter system can reliably operate.</p>
<p><em>Next in this series: Why do these ghost targets appear, and how do engineers suppress them?</em></p>
]]></content:encoded>
      <pubDate>Mon, 15 Dec 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/12/15/signals-of-opportunity/</guid>
      <category>passive-radar</category>
      <category>OFDM</category>
      <category>broadcast-signals</category>
      <category>signal-processing</category>
      <media:content url="https://joon.upthere.ai/images/silentium-defence-maverick-m8-passive-radar63.jpg" medium="image" />
    </item>
    <item>
      <title>The Cost of Forgetting Humans</title>
      <link>https://joon.upthere.ai/2025/12/13/cost-forgetting-humans/</link>
      <description>On the tension between corporate efficiency and human dignity.</description>
      <content:encoded><![CDATA[<p>A company exists to generate value for its shareholders. This is not controversial—it is the legal and economic foundation of corporate structure. Executives have fiduciary duties. Boards answer to investors. Profit is not a dirty word.</p>
<p>But here is what gets forgotten: shareholders do not write code, serve customers, or solve problems. People do.</p>
<p>The modern corporation has developed an impressive vocabulary for depersonalizing its workforce. Employees become “resources” to be allocated, “headcount” to be optimized, “FTEs” on a spreadsheet. This language is efficient. It is also revealing.</p>
<p>When people become abstractions, it becomes easier to treat them as expendable. Decisions that affect livelihoods get reduced to arithmetic. The human element—dignity, contribution, loyalty—disappears from the calculus.</p>
<p>This is not just an ethical concern. It is a strategic blind spot.</p>
<p>Companies that forget the humans behind the headcount eventually discover an uncomfortable truth: institutional knowledge walks out the door. Discretionary effort evaporates. The best people leave first, because they can. What remains is a hollowed-out organization wondering why the numbers no longer work.</p>
<p>Appreciation is not about perks or platitudes. It is about recognizing that without the people who show up and do the work, there is no company at all. Just a legal entity with no one left to make it real.</p>
<p>Shareholders may own the company. But people <em>are</em> the company.</p>
]]></content:encoded>
      <pubDate>Sat, 13 Dec 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/12/13/cost-forgetting-humans/</guid>
      <category>work</category>
      <category>culture</category>
      <media:content url="https://joon.upthere.ai/images/forgetting-humans.jpeg" medium="image" />
    </item>
    <item>
      <title>Introducing ATSC-RAG: An AI-Powered Knowledge Assistant for ATSC 3.0 Specifications</title>
      <link>https://joon.upthere.ai/2025/12/13/atsc-rag/</link>
      <description>A Retrieval-Augmented Generation system that makes ATSC 3.0 and B2X specifications searchable and queryable with AI, providing authoritative answers with source citations.</description>
      <content:encoded><![CDATA[<h1 id="introducing-atsc-rag" tabindex="-1">Introducing ATSC-RAG</h1>
<p>I’ve been working on a project that I’m excited to share with the ATSC community: <strong>ATSC-RAG</strong>, a Retrieval-Augmented Generation (RAG) system designed specifically for querying ATSC 3.0 and B2X specifications.</p>
<p>The system enables intelligent querying of broadcast standards documentation using modern AI techniques. Instead of manually searching through hundreds of pages of specifications, you can simply ask questions in natural language and get accurate, cited answers.</p>
<h1 id="how-it-works" tabindex="-1">How It Works</h1>
<p>ATSC-RAG combines several technologies to deliver authoritative answers:</p>
<ul>
<li><strong>Large Language Model</strong>: Qwen3-14B with 32K context window for understanding and generating responses</li>
<li><strong>Vision Language Model</strong>: Qwen3-VL-30B for analyzing diagrams and visual content from specifications</li>
<li><strong>Vector Database</strong>: Milvus storing approximately 50,000 document chunks for similarity search</li>
<li><strong>MCP Server</strong>: Integration with Claude Code and other AI assistants via Model Context Protocol</li>
</ul>
<p>When you ask a question like “What is a PLP in ATSC 3.0?”, the system retrieves relevant sections from the indexed specifications, then generates a comprehensive answer with source citations pointing to specific documents and page numbers.</p>
<p>The system runs entirely in one single Docker container, making deployment straightforward for anyone who wants to set it up locally.</p>
<h1 id="value-to-the-atsc-community" tabindex="-1">Value to the ATSC Community</h1>
<p>Working with broadcast standards can be challenging. The ATSC 3.0 suite includes dozens of specifications covering physical layer, link layer, signaling, video, audio, security, and more. Finding the right information often means searching across multiple documents.</p>
<p>ATSC-RAG addresses this by:</p>
<ul>
<li><strong>Making specifications searchable</strong>: Ask questions in plain English and get relevant answers</li>
<li><strong>Providing source citations</strong>: Every answer includes references to specific documents and pages</li>
<li><strong>Supporting visual content</strong>: Diagrams and figures are analyzed and indexed alongside text</li>
<li><strong>Integrating with AI tools</strong>: The MCP server allows Claude Code and other AI assistants to query specifications directly</li>
</ul>
<p>For engineers, researchers, and anyone working with ATSC 3.0, this can significantly reduce the time spent hunting for information.</p>
<h1 id="looking-for-contributors" tabindex="-1">Looking for Contributors</h1>
<p>I’ve developed this system to help with my own work in the ATSC community, but I believe it could benefit many others. The project is currently hosted at <a href="https://github.com/jp127266/atsc-rag">github.com/jp127266/atsc-rag</a>, a personal repo.</p>
<p>I’m open to contributors who are interested in:</p>
<ul>
<li>Improving the RAG pipeline and answer quality</li>
<li>Adding support for additional document formats</li>
<li>Implementing Phase 2 (GraphRAG for relationship queries)</li>
<li>Implementing Phase 3 (Ontology)</li>
<li>Testing and providing feedback</li>
<li>Documentation and examples</li>
</ul>
<p>With enough community interest and contributions, I hope to fully open-source this project for the benefit of the entire ATSC community. If you’re working with ATSC 3.0 specifications and interested in AI-powered tools, I’d love to hear from you.</p>
]]></content:encoded>
      <pubDate>Sat, 13 Dec 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/12/13/atsc-rag/</guid>
      <category>ATSC 3.0</category>
      <category>AI</category>
      <category>RAG</category>
      <category>MCP</category>
      <category>ontology</category>
      <category>open source</category>
      <media:content url="https://joon.upthere.ai/images/atsc-rag.png" medium="image" />
    </item>
    <item>
      <title>Why I&apos;m Not Worried About the AI Bubble</title>
      <link>https://joon.upthere.ai/2025/11/28/ai-bubble/</link>
      <description>Boom-bust cycles make transformative technology cheaper</description>
      <content:encoded><![CDATA[<p>A colleague recently raised concerns about the AI bubble. With massive investments pouring into anything AI-related and uncertain returns, his worry felt reasonable. I didn’t have a good counter-argument at the time.</p>
<p>Then I heard IBM CEO Arvind Krishna frame it differently. He pointed to the dot-com era: yes, 8 out of 10 heavily-invested companies went bankrupt. But would anyone say the Internet failed to deliver on its promise? Amazon and Google alone have probably paid for all the capital invested during that period. And the failed companies? Their assets didn’t vanish—they were acquired at ten cents on the dollar by others who built successful businesses on that foundation.</p>
<p>Krishna sees this as a strength of the US capital system. Other countries desperately try to keep all companies alive, diluting resources across zombies. The US lets failures fail. The system works.</p>
<p>This pattern isn’t new. The Railway Mania of the 1840s saw wild speculation and mass bankruptcies, yet it built 6,220 miles of British rail infrastructure. The 1920s electrification boom crashed spectacularly—the Dow Utilities index fell from 144 to 17—but by 1930, nearly 90% of American urban homes had electricity. Each bubble built lasting infrastructure that outlived the financial carnage.</p>
<p>The mechanism is consistent: mania attracts capital, capital builds infrastructure, bust makes that infrastructure cheap. The railways still run. The power grid still hums. The Internet still connects us.</p>
<p>AI will follow the same path. There will be tears. Many companies will fail. But the infrastructure—the data centers, the trained models, the tooling—will remain. And it will be cheaper for everyone who comes after.</p>
<p>Let the system work.</p>
<hr>
<p><em>Source: Arvind Krishna’s interview on <a href="https://open.spotify.com/episode/3mczFPNvXQnZgrCq0mIEke">Spotify</a></em></p>
]]></content:encoded>
      <pubDate>Fri, 28 Nov 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/11/28/ai-bubble/</guid>
      <category>technology</category>
      <category>AI</category>
      <category>investment</category>
      <category>innovation</category>
      <media:content url="https://joon.upthere.ai/images/dot-com-bubble.webp" medium="image" />
    </item>
    <item>
      <title>The ATSC 3.0 Transition: Why Removing the Deadline Opens the Door</title>
      <link>https://joon.upthere.ai/2025/11/26/atsc-3-transition/</link>
      <description>The FCC just removed the 2027 ATSC 3.0 transition deadline. Industry groups wanted mandates. But the voluntary approach opens bigger opportunities for broadcasters willing to embrace IP infrastructure.</description>
      <content:encoded><![CDATA[<p>The FCC just eliminated the 2027 deadline for the ATSC 3.0 transition. Broadcast industry groups pushed for mandates—forced migration, required tuners in TVs, hard cutoff dates. They didn’t get them. The transition remains voluntary, and that’s exactly what it should be.</p>
<p>ATSC 3.0 isn’t just “better broadcast TV.” It’s an infrastructure innovation. The real breakthrough is IP as the transport protocol. And if you look at the history of the telecom industry, you’ll see what that means.</p>
<h2 id="the-ip-transport-innovation" tabindex="-1">The IP Transport Innovation</h2>
<p>IP didn’t just make telecom networks more efficient. It transformed the entire industry structure. Before IP, phone companies were local or regional players, constrained by circuit-switched infrastructure. IP eliminated those boundaries. Local carriers became national players. Markets expanded. Revenues grew. Consolidation followed—mergers and acquisitions reshaped the competitive landscape. Yes, jobs were lost (remember all those switchboard operators), but the industry fundamentally expanded.</p>
<p>The ATSC 3.0 transition follows the same pattern. IP as the transport protocol means broadcasters aren’t just upgrading their transmission technology. They’re gaining the infrastructure to become national players. Broadcast-as-a-Service. Cloud-native operations. The ability to compete beyond geographic boundaries. The market opportunity isn’t in delivering 4K video to local viewers—it’s in operating as infrastructure providers at national scale.</p>
<h2 id="what-digicap-is-building" tabindex="-1">What DigiCAP Is Building</h2>
<p>At DigiCAP, we’re focused on making this infrastructure transition operationally feasible. Our flagship product, DigiCaster, is now fully containerized. That means broadcasters can deploy Broadcast-as-a-Service architectures—scalable, cloud-native, infrastructure-as-code. No more rack-mounted hardware locked into single facilities.</p>
<p>We’re also automating the broadcast operations with CCMS (Centralized Certificate Lifecycle Management System) and CDMS (Centralzied DigiCaster Management System). As broadcast systems get more complex, traditional GUIs become unsustainable. The cost of training operators on increasingly complicated interfaces doesn’t scale. So we’re experimenting with what I call the “Super-UI”—using generative AI with LLM integration and Model Context Protocol (MCP) to create operator interfaces that are both more capable and more intuitive. Complex backend, simple frontend. That’s the future of broadcast operations.</p>
<p>The next step? Proactive problem-solving. Broadcast backends are mission-critical—downtime isn’t an option. We’re evolving the Super-UI to diagnose issues, suggest fixes, and guide operators through resolution. Not just a simpler interface, but a smarter one.</p>
<h2 id="why-the-voluntary-timeline-matters" tabindex="-1">Why the Voluntary Timeline Matters</h2>
<p>IP-based broadcast infrastructure requires different operational expertise than traditional RF engineering. Containerization. Automation. Cloud-native architectures. API-driven workflows. These aren’t skills you develop overnight, and they’re not skills you can mandate through regulation.</p>
<p>The voluntary transition period gives broadcasters time to build this operational capacity. Infrastructure modernization isn’t about flipping a switch—it’s about developing organizational capabilities. The broadcasters who use this time to invest in infrastructure readiness will be positioned to compete as national players.</p>
<p><em>The ATSC 3.0 transition is not only about better broadcast TV, but also about becoming infrastructure providers in an IP-based world. The deadline doesn’t matter. Operational readiness does.</em></p>
<hr>
<p><strong>Sources:</strong></p>
<ul>
<li><a href="https://current.org/2025/11/fccs-atsc-3-0-rules-would-slow-transition-to-next-gen-tv-broadcasting/">FCC’s ATSC 3.0 rules would slow transition to next-gen TV broadcasting</a> - <a href="http://Current.org">Current.org</a></li>
<li>One US government report estimated the number of <a href="https://en.wikipedia.org/wiki/Switchboard_operator">switchboard operators</a> working for AT&amp;T at 161,669, when empoyment peaked in 1929.</li>
<li><a href="https://modelcontextprotocol.io/docs/getting-started/intro">What is the MCP?</a></li>
</ul>
<p><strong>Related:</strong></p>
<ul>
<li><a href="https://joon.upthere.ai/2025/11/20/leo-satellites-disrupting-telecom/">LEO Satellites and the Next Telecom Revolution</a></li>
</ul>
]]></content:encoded>
      <pubDate>Wed, 26 Nov 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/11/26/atsc-3-transition/</guid>
      <category>ATSC 3.0</category>
      <category>broadcasting</category>
      <category>infrastructure</category>
      <category>FCC</category>
      <media:content url="https://joon.upthere.ai/images/atsc-3-transition.jpg" medium="image" />
    </item>
    <item>
      <title>The Silent Watchers: An Introduction to Passive Radar (Part 1 of 4)</title>
      <link>https://joon.upthere.ai/2025/11/24/passive-radar-introduction/</link>
      <description>How passive radar systems detect aircraft using borrowed radio signals—from WWII secrets to modern anti-stealth technology</description>
      <content:encoded><![CDATA[<p>Imagine detecting aircraft without transmitting a single signal. Passive radar systems accomplish this remarkable feat by “borrowing” existing broadcast transmissions—your FM radio, digital TV, or cellular signals—and using them to track targets. Unlike traditional radar that broadcasts its own signal and listens for echoes, passive radar operates silently, using signals already in the air.</p>
<h2 id="a-brief-history" tabindex="-1">A Brief History</h2>
<p>The concept isn’t new. In 1935, Robert Watson-Watt demonstrated the first passive radar by detecting aircraft using BBC radio transmissions at Daventry. During World War II, Germany’s Klein Heidelberg receivers exploited British radar signals to detect Allied bombers. After decades of dormancy, NATO revived passive radar in the 1990s, recognizing its potential against stealth aircraft—which are designed to deflect active radar but can’t hide from signals bouncing off them at different angles.</p>
<h2 id="why-it-matters-today" tabindex="-1">Why It Matters Today</h2>
<p>Passive radar offers compelling advantages that have made it commercially viable:</p>
<p><strong>Cost-effective operation:</strong> Systems cost 10-50× less than traditional radar. No high-power transmitter means lower capital costs and minimal power consumption.</p>
<p><strong>Covert surveillance:</strong> Without emissions, passive radar is virtually undetectable. Adversaries can’t locate or jam what they can’t see transmitting.</p>
<p><strong>Anti-stealth capability:</strong> The bistatic geometry (separate transmitter and receiver locations) defeats stealth design, which relies on reflecting signals away from their source.</p>
<p><strong>Operational today:</strong> Commercial systems are deployed worldwide—Hensoldt’s TwinVis in Germany, ERA’s Vera-NG in the Czech Republic, and Silentium Defence’s Maverick in Australia.</p>
<p><strong>Enhanced with AI:</strong> Modern deployments fuse data from multiple passive radar sensors with AI pattern-matching algorithms, further improving detection accuracy and reducing false alarms.</p>
<h2 id="the-evolution-continues" tabindex="-1">The Evolution Continues</h2>
<p>Passive radar systems today primarily use DVB-T broadcast signals. But broadcast technology is evolving. Next-generation TV standards like ATSC 3.0—now operational in South Korea and expanding across the United States—offer more flexible signal structures. These advanced systems may enable significant improvements: larger coverage areas without false targets, configurable patterns for different missions, and better urban performance.</p>
<p>As this series will explore, the convergence of modern broadcast technology and passive sensing creates new commercial opportunities. Companies developing surveillance and sensing applications can leverage existing broadcast infrastructure—no need to deploy expensive transmitters. This opens doors for startups, defense contractors, and technology firms to build cost-effective radar systems for civilian markets: airport security, border monitoring, critical infrastructure protection, and beyond.</p>
<p><em>Next in this series: What signals can we actually use? We’ll explore the various “emitters of opportunity” that make passive radar possible, from FM radio to satellites, and why broadcast TV has emerged as the gold standard.</em></p>
<p><strong>P.S.</strong> For deeper technical background, see Rohde &amp; Schwarz’s <a href="https://www.rohde-schwarz.com/us/solutions/aerospace-and-defense/land/emso-testing/whitepaper-an-introduction-to-passive-radar-systems_256615.html?change_c=true">Introduction to Passive Radar Systems</a>.</p>
]]></content:encoded>
      <pubDate>Mon, 24 Nov 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/11/24/passive-radar-introduction/</guid>
      <category>radar</category>
      <category>passive-radar</category>
      <category>defense-technology</category>
      <category>signal-processing</category>
      <media:content url="https://joon.upthere.ai/images/passive-radar-concept.png" medium="image" />
    </item>
    <item>
      <title>Harmonizing ATSC 3.0 and 5G Broadcast: A New Business Opportunity</title>
      <link>https://joon.upthere.ai/2025/11/24/atsc-5g-broadcast-integration/</link>
      <description>DigiCAP and Castanet 5G demonstrate seamless integration of 5G Broadcast frames into ATSC 3.0 infrastructure at NAB NY Show</description>
      <content:encoded><![CDATA[<p>At the NAB NY Show in early October 2025, DigiCAP and Castanet 5G (C5G) showcased something the broadcast industry has been theorizing about for years: <strong>seamless integration of 5G Broadcast into ATSC 3.0 infrastructure</strong>.</p>
<p>In a private demonstration, we at DigiCAP helped C5G insert 5G Broadcast frames directly into the gaps generated by an ATSC 3.0 modulator. This wasn’t a hack or workaround—we used the very mechanisms ATSC 3.0 was designed to support: the bootstrap signal and the <code>min_time_to_next</code> property.</p>
<h2 id="why-this-matters" tabindex="-1">Why This Matters</h2>
<p>ATSC 3.0 was architected with forward compatibility in mind. The standard explicitly accommodates new protocols through its timing properties. What we demonstrated was that this design principle actually works in practice.</p>
<p>By leveraging the natural gaps in ATSC 3.0 frame timing, we multiplexed 5G Broadcast frames without disrupting the primary ATSC 3.0 service. The yellow frames (ATSC 3.0) and green frames (5G Broadcast) coexist on the same infrastructure, sharing spectrum efficiently.</p>
<h2 id="the-technical-approach" tabindex="-1">The Technical Approach</h2>
<p>The key was respecting ATSC 3.0’s bootstrap structure. The <code>min_time_to_next</code> property tells receivers when to expect the next frame. We used those intervals to inject 5G Broadcast data, maintaining perfect synchronization.</p>
<p>This isn’t theoretical anymore. The demo proved that broadcasters can deploy 5G Broadcast capabilities without replacing their ATSC 3.0 equipment—just smart integration.</p>
<h2 id="what%E2%80%99s-next%3F" tabindex="-1">What’s Next?</h2>
<p>This demonstration opens doors for broadcasters to leverage both standards simultaneously. Imagine delivering traditional broadcast content via ATSC 3.0 while providing targeted, mobile-optimized streams through 5G Broadcast—all from the same transmitter.</p>
<p>The infrastructure is ready. The technology works. Now it’s about deployment strategy.</p>
<hr>
<p><strong>P.S.</strong> For more details about Castanet’s broader 5G Broadcast strategy and market positioning, see Jeff Baumgartner’s coverage in <a href="https://www.lightreading.com/5g/castanet-takes-aim-at-5g-broadcast">Light Reading: “Castanet Takes Aim at 5G Broadcast”</a>.</p>
]]></content:encoded>
      <pubDate>Mon, 24 Nov 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/11/24/atsc-5g-broadcast-integration/</guid>
      <category>broadcasting</category>
      <category>5g</category>
      <category>technology</category>
      <category>ATSC 3.0</category>
      <media:content url="https://joon.upthere.ai/images/atsc-5g-multiplexing.webp" medium="image" />
    </item>
    <item>
      <title>The Interactive Television Paradox: How YouTube Delivered What Broadcasters Couldn&apos;t</title>
      <link>https://joon.upthere.ai/2025/11/22/interactive-tv-paradox/</link>
      <description>Broadcasters spent decades pursuing interactive television through apps and middleware. YouTube made it real with comments and chat.</description>
      <content:encoded><![CDATA[<p>For decades, broadcasters chased the dream of interactive television. They tried everything: apps embedded in MPEG transport streams, middleware platforms, companion screens, red-button services. Countless standards emerged—MHP, HbbTV, ATSC 3.0 applications—each promising to transform passive viewing into active engagement.</p>
<p>None of them succeeded at scale.</p>
<p>The irony? Interactive television finally became reality, but not where broadcasters expected. It happened on YouTube.</p>
<h2 id="the-dream-vs.-the-reality" tabindex="-1">The Dream vs. The Reality</h2>
<p>Watch any live YouTube stream. The creator reads comments in real-time, answers questions, changes direction based on viewer feedback. Someone suggests a topic—the streamer pivots. Viewers vote on decisions. The chat influences what happens next, live, as it unfolds.</p>
<p>This is the interactive television broadcasters imagined. Not apps running alongside content. Not voting on pre-determined outcomes. Real-time, bidirectional, responsive content creation shaped by audience participation.</p>
<p>Broadcasters built elaborate technical infrastructure trying to achieve this. YouTube just needed a chat box and creators willing to listen.</p>
<h2 id="why-broadcasters-failed" tabindex="-1">Why Broadcasters Failed</h2>
<p>The problem wasn’t technical capability. Broadcasters had the technology to enable interactivity. What they lacked was the right model.</p>
<p>Traditional broadcasting is fundamentally one-to-many. A network creates content, transmits it, and millions receive the same thing. Interactivity was bolted on as a feature—an afterthought to the core broadcast model.</p>
<p>YouTube inverted this. The platform made interactivity the foundation, not the feature. Content isn’t broadcast to passive recipients; it’s created in conversation with an active community.</p>
<h2 id="the-medium-vs.-the-service" tabindex="-1">The Medium vs. The Service</h2>
<p>This reveals something crucial: <strong>broadcasting as a medium remains relevant, but broadcasting as a television service model is fading.</strong></p>
<p>The medium—one transmitter reaching many receivers efficiently—still has value for live events, emergencies, and bandwidth-constrained scenarios. But the service model—networks controlling what, when, and how content is consumed—lost to internet streaming’s flexibility and interactivity.</p>
<p>Broadcasters spent resources trying to make their service model interactive when the internet had already solved it differently. They asked “How do we add interactivity to broadcasting?” when they should have asked “What role does broadcast infrastructure play in an interactive media landscape?”</p>
<h2 id="the-path-forward" tabindex="-1">The Path Forward</h2>
<p>YouTube, Netflix, and streaming services didn’t make broadcasting obsolete—they made the traditional television service model obsolete. The distinction matters.</p>
<p>Smart broadcasters will stop trying to recreate YouTube on ATSC 3.0 or HbbTV. Instead, they’ll recognize what broadcasting does uniquely well: efficient, reliable, one-to-many distribution. That’s infrastructure, not a service.</p>
<p>The interactive television paradox is complete: the industry that invented television couldn’t deliver interactive TV because they were too focused on preserving television as they knew it. YouTube succeeded because it never tried to be television at all.</p>
<p>Sometimes the future of your medium means accepting it’s no longer your service.</p>
]]></content:encoded>
      <pubDate>Sat, 22 Nov 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/11/22/interactive-tv-paradox/</guid>
      <category>broadcasting</category>
      <category>technology</category>
      <category>streaming</category>
      <category>innovation</category>
      <media:content url="https://joon.upthere.ai/images/interactive-tv.jpg" medium="image" />
    </item>
    <item>
      <title>Walled Gardens vs Open Systems: Why Broadcasters Need to Embrace the Pipe</title>
      <link>https://joon.upthere.ai/2025/11/20/walled-gardens-broadcasters/</link>
      <description>The telecommunications industry learned to abandon walled gardens and focus on infrastructure. It&apos;s time for broadcasters to follow suit.</description>
      <content:encoded><![CDATA[<p>The debate between walled gardens and open systems is nothing new. It emerged at the dawn of the telecommunications age, when carriers attempted to control both the infrastructure and the services running on top of it. Eventually, they learned a crucial lesson: innovation thrives in openness.</p>
<p>Telecommunications companies abandoned their walled gardens and pivoted to what they do best—providing better pipes. They became infrastructure providers, and the result was transformative. All the meaningful innovation happened in the open Internet ecosystem, built on top of these neutral pipes. Applications, services, and platforms flourished because developers could build freely without asking permission from gatekeepers.</p>
<p>Yet broadcasters haven’t learned this lesson. They still cling to their walled gardens, maintaining tight control over content distribution and access. This might explain why they’ve consistently lagged behind Internet innovations. While the digital world moved to on-demand, interactive, and personalized experiences, traditional broadcasting remained largely one-way and inflexible.</p>
<p>The introduction of ATSC 3.0 and next-generation broadcasting technologies presents a pivotal opportunity. Instead of using these advances to build more sophisticated walled gardens, broadcasters should ask themselves: How can we become better pipes? How can we bring Internet-style innovation to our infrastructure?</p>
<p>The answer isn’t to compete with Internet services by replicating them within a closed system. It’s to provide robust, efficient broadcast infrastructure that Internet innovators can build upon. Embrace the role of the pipe, and let the ecosystem flourish on top.</p>
<p>The telecommunications industry already proved this model works. It’s time for broadcasters to follow.</p>
]]></content:encoded>
      <pubDate>Thu, 20 Nov 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/11/20/walled-gardens-broadcasters/</guid>
      <category>broadcasting</category>
      <category>ATSC 3.0</category>
      <category>open systems</category>
      <category>infrastructure</category>
      <media:content url="https://joon.upthere.ai/images/walled-garden.jpg" medium="image" />
    </item>
    <item>
      <title>LEO Satellites: The Next Telecommunications Revolution</title>
      <link>https://joon.upthere.ai/2025/11/20/leo-satellites-disrupting-telecom/</link>
      <description>How Low Earth Orbit satellites are set to disrupt global telecommunications, making roaming fees and international plans obsolete.</description>
      <content:encoded><![CDATA[<p>The telecommunications industry has experienced seismic shifts before. Mobile phones made long-distance calling obsolete, transforming what was once an expensive, premium service into something we take for granted. But mobile networks had a fundamental limitation: national borders. LEO (Low Earth Orbit) satellites are about to shatter that constraint entirely.</p>
<h2 id="the-mobile-revolution%E2%80%99s-unfinished-promise" tabindex="-1">The Mobile Revolution’s Unfinished Promise</h2>
<p>Remember when making a long-distance call meant planning ahead and watching the clock? Mobile telecommunications changed everything. What was once a premium service with complex pricing tiers became simple: one plan, unlimited calls across the country. Distance became irrelevant within national borders.</p>
<p>But the revolution stopped at the edge of each nation. Travel abroad, and suddenly you’re back in the world of premium pricing, roaming fees, and complex international plans. The technology that made domestic distance meaningless couldn’t cross political boundaries.</p>
<h2 id="enter-leo-satellites" tabindex="-1">Enter LEO Satellites</h2>
<p>Low Earth Orbit satellites are positioned to finish what mobile networks started. These satellites orbit at altitudes between 500 and 2,000 kilometers—much closer than traditional geostationary satellites—creating a mesh network that blankets the entire planet.</p>
<p>The implications are profound: <strong>telecommunications is about to become truly global</strong>.</p>
<h2 id="one-plan%2C-one-planet" tabindex="-1">One Plan, One Planet</h2>
<p>Imagine this: You’re on a business trip to Tokyo, vacation in Paris, or visiting family in São Paulo. You pull out your phone, and it just works. No roaming charges. No international plan add-ons. No SIM card swaps. The same plan you use at home works everywhere on Earth.</p>
<p>This isn’t science fiction. Companies like SpaceX’s Starlink, OneWeb, and Amazon’s Project Kuiper are already deploying thousands of LEO satellites. While initially focused on internet connectivity, the technology fundamentally enables global voice and data services without traditional cellular infrastructure.</p>
<h2 id="the-disruption-ahead" tabindex="-1">The Disruption Ahead</h2>
<p>The telecom industry’s current business model relies heavily on geographical monopolies and partnerships. Carriers charge premium rates for international roaming because customers have no choice—they need connectivity abroad, and the carrier knows it.</p>
<p>LEO satellite networks eliminate this leverage entirely. Why pay roaming fees when your device can connect directly to a global satellite network? Why negotiate complex international partnerships when one satellite constellation serves the entire planet?</p>
<h2 id="what-makes-this-different" tabindex="-1">What Makes This Different</h2>
<p>Previous satellite phone systems existed but remained niche due to high costs and bulky equipment. LEO satellites change the equation:</p>
<ul>
<li><strong>Lower latency</strong>: Closer orbits mean faster communication</li>
<li><strong>Higher bandwidth</strong>: Modern satellites can handle data-intensive applications</li>
<li><strong>Smaller devices</strong>: Direct-to-smartphone connectivity is becoming reality</li>
<li><strong>Competitive pricing</strong>: Mass deployment drives costs down</li>
</ul>
<p>The technology that once required specialized equipment is being integrated directly into standard smartphones. Apple’s iPhone 14 already includes satellite emergency messaging. Full satellite connectivity for regular calls and data is the next logical step.</p>
<h2 id="the-ripple-effects" tabindex="-1">The Ripple Effects</h2>
<p>When telecommunications goes global, the implications extend beyond cheaper roaming:</p>
<ul>
<li><strong>Remote work</strong>: Truly work from anywhere on the planet</li>
<li><strong>Emergency services</strong>: Global coverage for disaster response</li>
<li><strong>Developing nations</strong>: Connectivity without expensive infrastructure</li>
<li><strong>Maritime and aviation</strong>: Seamless coverage over oceans and remote areas</li>
<li><strong>IoT devices</strong>: Global tracking and monitoring without network gaps</li>
</ul>
<h2 id="the-incumbent-response" tabindex="-1">The Incumbent Response</h2>
<p>Traditional telecom carriers aren’t sitting idle. Some are partnering with satellite providers, integrating satellite connectivity into their networks. Others are investing in their own satellite capabilities. The smart ones recognize that the choice isn’t whether to adopt this technology, but how quickly.</p>
<p>Those who resist or move too slowly risk becoming the long-distance carriers of the mobile era—companies that failed to adapt to a paradigm shift and faded into irrelevance.</p>
<h2 id="challenges-ahead" tabindex="-1">Challenges Ahead</h2>
<p>The transition won’t happen overnight. Regulatory frameworks need updating. Spectrum allocation must be negotiated. Ground infrastructure requires upgrading. And countries that derive revenue from telecom monopolies may resist.</p>
<p>But the technological trajectory is clear. Just as mobile networks made long-distance calls obsolete within countries, LEO satellites will make international roaming obsolete globally.</p>
<h2 id="the-future-is-global" tabindex="-1">The Future Is Global</h2>
<p>We’re witnessing the final chapter of telecommunications fragmentation. The era of thinking about phone service in terms of regions, countries, or continents is ending. The next generation won’t understand why we once paid extra to make calls across borders.</p>
<p>LEO satellites aren’t just improving telecommunications—they’re completing the revolution that mobile networks began. They’re making the entire planet feel as connected as a single city once did.</p>
<p>The question isn’t if this will happen, but how quickly incumbent carriers will adapt. Because whether they’re ready or not, the future of telecommunications is already orbiting overhead.</p>
<hr>
<p><em>The telecommunications revolution is entering its next phase. And this time, the whole world is the network.</em></p>
]]></content:encoded>
      <pubDate>Thu, 20 Nov 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/11/20/leo-satellites-disrupting-telecom/</guid>
      <category>technology</category>
      <category>telecommunications</category>
      <category>satellites</category>
      <category>innovation</category>
      <media:content url="https://joon.upthere.ai/images/leo-satellites.jpg" medium="image" />
    </item>
    <item>
      <title>Welcome to My Blog</title>
      <link>https://joon.upthere.ai/2025/11/19/welcome/</link>
      <description>An introduction to my personal blog and what I plan to write about.</description>
      <content:encoded><![CDATA[<p>Welcome to my personal blog. This is a space where I’ll be sharing my thoughts, experiences, and insights on various topics.</p>
<h2 id="what-to-expect" tabindex="-1">What to Expect</h2>
<p>I plan to write about:</p>
<ul>
<li><strong>Technology and Design</strong>: Thoughts on software development, design principles, and emerging technologies</li>
<li><strong>Personal Growth</strong>: Lessons learned and reflections on career and life</li>
<li><strong>Projects</strong>: Updates on side projects and experiments</li>
<li><strong>Book Notes</strong>: Summaries and insights from books I’m reading</li>
</ul>
<h2 id="why-i%E2%80%99m-starting-this" tabindex="-1">Why I’m Starting This</h2>
<p>Writing helps me clarify my thinking and serves as a record of my journey. If others find value in these posts, that’s a wonderful bonus.</p>
<h2 id="get-in-touch" tabindex="-1">Get in Touch</h2>
<p>Feel free to reach out if you’d like to discuss any of the topics I write about.</p>
<hr>
<p>Thanks for visiting, and I hope you find something interesting here.</p>
]]></content:encoded>
      <pubDate>Wed, 19 Nov 2025 00:00:00 GMT</pubDate>
      <author>Joonyoung Park</author>
      <guid>https://joon.upthere.ai/2025/11/19/welcome/</guid>
      <category>introduction</category>
      <category>technology</category>
      <category>design</category>
      <media:content url="https://joon.upthere.ai/images/welcome.jpg" medium="image" />
    </item>
  </channel>
</rss>