欢迎!

分享和讨论最佳内容和新营销理念,打造专业形象,共同成为更好的营销人员。

注册

You need to be registered to interact with the community.
This question has been flagged
6 Views

Do not synchronize through the acoustic link if you need tight timing. Acoustic travel time is large and variable, and it changes with sound speed, geometry and platform motion, so a timestamp carried over the modem inherits all of that uncertainty.

Better approaches, from tightest to loosest:

  • Local hardware trigger. A wire or a shared pulse from the same clock starts both logger and modem. This is the only route to sub-millisecond alignment.
  • Holdover oscillators. Give each node a temperature-compensated oscillator, synchronize before deployment, and accept drift over the mission; log the drift estimate.
  • Acoustic two-way ranging. Exchange messages both ways and average the travel time to cancel the unknown offset. Typical accuracy is milliseconds, not microseconds, and depends on a stable sound-speed estimate.
  • Post-hoc correction. If none of the above is possible, record local timestamps and correct them afterwards using modelled travel time and measured sound-speed profiles.

Whatever the method, log the arrival time of every frame: the timing metadata is usually worth more than the payload when you are reconstructing an event.

Avatar
Discard