<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: NT Maritime</title>
    <description>The latest articles on DEV Community by NT Maritime (@ntmaritime).</description>
    <link>https://dev.to/ntmaritime</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4206678%2F1980c741-7fcd-41ab-9f15-5633384e262b.jpg</url>
      <title>DEV Community: NT Maritime</title>
      <link>https://dev.to/ntmaritime</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/ntmaritime"/>
    <language>en</language>
    <item>
      <title>When the Satellite Drops: Designing Voice and Messaging for Intermittent Backhaul at Sea</title>
      <dc:creator>NT Maritime</dc:creator>
      <pubDate>Wed, 28 Oct 2026 10:11:26 +0000</pubDate>
      <link>https://dev.to/ntmaritime/when-the-satellite-drops-designing-voice-and-messaging-for-intermittent-backhaul-at-sea-4cdh</link>
      <guid>https://dev.to/ntmaritime/when-the-satellite-drops-designing-voice-and-messaging-for-intermittent-backhaul-at-sea-4cdh</guid>
      <description>&lt;p&gt;In our &lt;a href="https://dev.to/ntmaritime/voice-and-messaging-at-sea-engineering-crew-communications-over-ship-wifi-1d40"&gt;first post&lt;/a&gt; we argued that a ship should be treated as an edge site: run communication services on local servers on board, and keep onboard traffic onboard. This post picks up the part we only touched on, which is what happens to the traffic that &lt;em&gt;does&lt;/em&gt; need to leave the ship when the satellite link is slow, lossy, or simply gone.&lt;/p&gt;

&lt;p&gt;Every vessel's uplink is shared, asymmetric, and subject to conditions nobody on board controls. Even with modern LEO service, there will be moments when the link degrades or drops. A communications stack that only works when the link is healthy will, from the user's point of view, simply be broken some of the time. So the question is not "how do we avoid outages" but "how do we behave well during them."&lt;/p&gt;

&lt;h2&gt;
  
  
  Separate the two failure domains
&lt;/h2&gt;

&lt;p&gt;The most important design decision is to make sure a backhaul outage is &lt;em&gt;only&lt;/em&gt; a backhaul outage. If the onboard servers handle signaling, media relay, message storage, and voicemail locally, then a crew member calling another crew member on a different deck never notices the satellite went down. Only off-ship features are affected.&lt;/p&gt;

&lt;p&gt;In practice that means:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Local identity and auth.&lt;/strong&gt; Sessions should validate against onboard services. If every login or token refresh needs a round trip to a shore data center, an outage logs everyone out.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Local directory and presence.&lt;/strong&gt; Who is reachable, and on which device, should be answerable from the ship.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Local push path where possible.&lt;/strong&gt; Mobile OSes rely on Apple and Google push services, which sit on the far side of the satellite. Plan explicitly for how a sleeping phone gets woken up for an onboard call when that path is unavailable, and test it with the uplink unplugged.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;If you can pull the uplink cable in a test environment and onboard calling, messaging, and voicemail all keep working, you have the foundation right.&lt;/p&gt;

&lt;h2&gt;
  
  
  Store-and-forward for everything that can wait
&lt;/h2&gt;

&lt;p&gt;Real-time voice and video cannot be queued, but almost everything else can. Messages, voicemail, delivery receipts, call logs, and sync data all fit a store-and-forward model:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Accept the item locally and acknowledge it to the sender right away, so the client is not left spinning.&lt;/li&gt;
&lt;li&gt;Persist it durably on the onboard server.&lt;/li&gt;
&lt;li&gt;Forward it when the link allows, with retries and backoff.&lt;/li&gt;
&lt;li&gt;Report the final state (delivered, failed) back to the sender.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;This is the same idea email has used for decades, and it suits ships well. The key engineering details are:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Idempotency.&lt;/strong&gt; After a flaky link, you will retransmit things the other side already received. Give every message a stable ID so duplicates are dropped rather than shown twice.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Ordering.&lt;/strong&gt; Per-conversation sequence numbers or timestamps from a consistent source keep threads readable when items arrive in bursts.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Durability.&lt;/strong&gt; A queue held only in memory turns a server restart during an outage into data loss.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Avoid the reconnection stampede
&lt;/h2&gt;

&lt;p&gt;When the link comes back, every queued message, every client reconnect, every sync job, and every software update check wants bandwidth at the same moment. On a shared, asymmetric uplink this can saturate the link right when people are trying to make their first off-ship call in an hour.&lt;/p&gt;

&lt;p&gt;Some useful habits:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Prioritize.&lt;/strong&gt; Real-time voice signaling and media first, then user messages, then receipts and sync, with bulk transfers last.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Jitter retries.&lt;/strong&gt; Randomized backoff spreads reconnects out instead of synchronizing them.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Rate-limit the drain.&lt;/strong&gt; Flush queues at a controlled rate rather than all at once.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Mark traffic.&lt;/strong&gt; QoS markings on the onboard network and at the edge router help the scheduler do the right thing.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Degrade gracefully for real-time media
&lt;/h2&gt;

&lt;p&gt;For off-ship calls, the link may be up but poor. Adaptive codecs and bitrates, jitter buffers sized for the link's actual behavior, and clear fallbacks help: drop video to audio, and when audio is not viable, steer the user toward a message or voicemail rather than leaving them on a call that keeps breaking up. A call that is clearly downgraded is a better experience than one that silently fails.&lt;/p&gt;

&lt;h2&gt;
  
  
  Be honest in the UI
&lt;/h2&gt;

&lt;p&gt;Much of the user experience during an outage is about expectations. A few small touches go a long way:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Show a clear connectivity state that distinguishes "onboard only" from "fully connected," so users know onboard calls still work.&lt;/li&gt;
&lt;li&gt;Mark messages as &lt;em&gt;queued&lt;/em&gt; versus &lt;em&gt;delivered&lt;/em&gt;, rather than showing an error or a fake success.&lt;/li&gt;
&lt;li&gt;Tell users before they place an off-ship call when quality is likely to be poor.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Hiding the limitation does not make it go away. Being upfront that service quality depends on the ship's connection builds more trust than pretending the network is always fine.&lt;/p&gt;

&lt;h2&gt;
  
  
  Test the bad days, not the good ones
&lt;/h2&gt;

&lt;p&gt;Lab networks are usually too good. Use network emulation (for example Linux &lt;code&gt;tc netem&lt;/code&gt;) to add latency, jitter, packet loss, and full outages, and run scripted scenarios: an outage mid-call, a server restart while messages are queued, a hundred clients reconnecting at once. The goal is to see the failure modes in a test rig before the crew sees them in the middle of an ocean.&lt;/p&gt;

&lt;h2&gt;
  
  
  Takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Keep onboard services fully independent of the backhaul, including auth, directory, and push planning.&lt;/li&gt;
&lt;li&gt;Use durable, idempotent store-and-forward for anything that is not real-time.&lt;/li&gt;
&lt;li&gt;Prioritize and pace traffic when the link returns.&lt;/li&gt;
&lt;li&gt;Degrade media gracefully and tell users what is happening.&lt;/li&gt;
&lt;li&gt;Test with the uplink impaired or unplugged.&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;At NT Maritime we run local servers on board and build crew and passenger calling, video, messaging, and voicemail over ship WiFi. More at &lt;a href="https://ntmaritime.com" rel="noopener noreferrer"&gt;ntmaritime.com&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>voip</category>
      <category>networking</category>
      <category>distributedsystems</category>
      <category>maritime</category>
    </item>
    <item>
      <title>Voice and Messaging at Sea: Engineering Crew Communications Over Ship WiFi</title>
      <dc:creator>NT Maritime</dc:creator>
      <pubDate>Sun, 25 Oct 2026 13:43:09 +0000</pubDate>
      <link>https://dev.to/ntmaritime/voice-and-messaging-at-sea-engineering-crew-communications-over-ship-wifi-1d40</link>
      <guid>https://dev.to/ntmaritime/voice-and-messaging-at-sea-engineering-crew-communications-over-ship-wifi-1d40</guid>
      <description>&lt;p&gt;Most VoIP and messaging stacks are designed with an unspoken assumption: there is a reasonably fast, reasonably stable internet connection between the client and the server. On a ship, that assumption breaks. A vessel is effectively a small, floating campus network with a single, shared, variable uplink to the rest of the world. If you build communications for that environment, the architecture looks different.&lt;/p&gt;

&lt;p&gt;This post walks through the main engineering constraints of running calling, video, and messaging over a ship's WiFi/IP network, and the design patterns that follow from them.&lt;/p&gt;

&lt;h2&gt;
  
  
  The network you actually have
&lt;/h2&gt;

&lt;p&gt;A typical ship network has two very different halves:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;The onboard LAN/WLAN.&lt;/strong&gt; Access points throughout the vessel, switching, and local servers. This part is under the operator's control and can be engineered like any campus network.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The backhaul.&lt;/strong&gt; Everything leaving the ship goes over a satellite link. Newer LEO constellations have changed the picture a lot. For example, Starlink is commonly cited at up to 220 Mbps down, 40 Mbps up, with latency under 99 ms. But the link is still shared by everyone onboard, it is asymmetric, and its quality depends on conditions you do not control.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;The practical consequence: &lt;strong&gt;service quality depends on the ship's WiFi connection.&lt;/strong&gt; Any honest design has to start from that limitation rather than hide it.&lt;/p&gt;

&lt;h2&gt;
  
  
  Design principle 1: keep onboard traffic onboard
&lt;/h2&gt;

&lt;p&gt;A large share of communication on a ship is between people on the same ship: crew coordinating across departments, or passengers in the same travel group. Sending a call from deck 4 to deck 9 out to a cloud server and back over satellite is wasteful and adds latency and jitter for no reason.&lt;/p&gt;

&lt;p&gt;The better pattern is to run signaling and media services locally on the vessel, so that:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Ship-to-ship-user calls, messages, and voicemail never touch the backhaul.&lt;/li&gt;
&lt;li&gt;Onboard communication keeps working when the satellite link degrades or drops.&lt;/li&gt;
&lt;li&gt;The uplink is reserved for traffic that genuinely has to leave the ship (off-ship calls, sync, updates).&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;This is essentially an edge deployment: treat the vessel as an isolated site that occasionally syncs with the shore, not as a thin client of a central cloud. Teams that have built for air-gapped environments will recognize the mindset.&lt;/p&gt;

&lt;h2&gt;
  
  
  Design principle 2: assume NAT, roaming, and sleeping phones
&lt;/h2&gt;

&lt;p&gt;On the client side, the endpoints are usually smartphones on WiFi, not desk phones on a wired VLAN. That brings a familiar list of VoIP headaches, amplified:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;NAT and firewalls.&lt;/strong&gt; SIP and RTP have well-known NAT traversal issues. WebRTC (with ICE/STUN/TURN) or carefully configured SIP keep-alives and media relays are typical answers. Relays should live onboard so media does not hairpin through the satellite.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;AP roaming.&lt;/strong&gt; People walk through steel bulkheads between access points mid-call. The WLAN design (coverage, roaming behavior, QoS marking for voice) matters as much as the app.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Background delivery.&lt;/strong&gt; Mobile OSes aggressively suspend apps. Incoming calls and messages generally rely on Apple Push Notifications and Google Firebase Cloud Messaging to wake the device, which means you need a plan for how push reaches a phone on an isolated network where the push services sit across the satellite link.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Design principle 3: pick protocols for the link, not the whiteboard
&lt;/h2&gt;

&lt;p&gt;Messaging protocols vary a lot in how they behave on constrained, intermittent links. XMPP is a common choice for chat and presence. MQTT and AMQP are good fits for lightweight pub/sub and device telemetry. SIP and WebRTC handle real-time voice and video. Using each where it fits, rather than forcing everything through one stack, makes it easier to cope with packet loss and reconnection storms after an outage.&lt;/p&gt;

&lt;p&gt;For voice and video specifically, think about adaptive codecs and bitrate, jitter buffers sized for real link behavior, and graceful downgrade (video to audio, audio to message) when off-ship quality drops.&lt;/p&gt;

&lt;h2&gt;
  
  
  Design principle 4: integrate with the ship's existing systems
&lt;/h2&gt;

&lt;p&gt;Ships already run PBXs, sometimes legacy TDM, plus property and passenger management systems. Ripping all of that out is rarely realistic. A common modernization path is TDM-to-IP and POTS-to-SIP migration that keeps existing equipment in service while modernizing the backend, plus APIs and middleware to tie in identity (who is this cabin/crew member?) and provisioning.&lt;/p&gt;

&lt;h2&gt;
  
  
  Replacing DECT with apps
&lt;/h2&gt;

&lt;p&gt;A concrete example of all of the above is crew handsets. Many vessels rely on DECT phones for internal crew communication. Replacing them with an app on a phone the crew already carries means calling, video calling, messaging, and voicemail all run over the ship's WiFi, but only if the onboard-first architecture, WLAN roaming, and push delivery are done well. Otherwise you have traded a dedicated radio system for a flaky app.&lt;/p&gt;

&lt;h2&gt;
  
  
  Takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Treat the ship as an edge site; keep local traffic local.&lt;/li&gt;
&lt;li&gt;Engineer the WLAN for voice, not just browsing.&lt;/li&gt;
&lt;li&gt;Plan for NAT, roaming, and mobile push on an isolated network.&lt;/li&gt;
&lt;li&gt;Choose protocols per workload and design for degradation.&lt;/li&gt;
&lt;li&gt;Modernize around existing PBX and passenger systems rather than replacing everything.&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;At NT Maritime we work on this problem, running local servers on board and building crew and passenger communications (calling, video, messaging, voicemail) over ship WiFi, including crew DECT phone replacement. More at &lt;a href="https://ntmaritime.com" rel="noopener noreferrer"&gt;ntmaritime.com&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>voip</category>
      <category>networking</category>
      <category>telecom</category>
      <category>maritime</category>
    </item>
  </channel>
</rss>
