Precise timestamping from the packet forwarder

Home › Forums › Conduit: AEP Model › Precise timestamping from the packet forwarder

Tagged: 

Viewing 12 posts - 1 through 12 (of 12 total)
  • Author
    Posts
  • February 15, 2019 at 4:08 pm #27256
    Christopher Hunt
    Participant

    Hi there,

    We have a new IP67 Conduit with AEP installed. I was wondering if there’s supposed to be support for the precise timestamping of packets received by the packet forwarder. Normally, these packets are forwarded on using the time field in the JSON rxpk object. I’ve begun tracing packets to see, but I suspect that the time field isn’t being provided, or if it is, then it isn’t precise.

    Thanks for any clarification.

    Cheers,
    -C

    February 15, 2019 at 4:47 pm #27257
    Steve Kovarik
    Moderator

    Hello -C

    IP67 Conduits with AEP do not support precise time stamping.
    Currently only mLinux models MTCDTIP-xxx-275L-xxx support precise
    time stamping for geolocation using TDoA.

    -Best Regards

    February 15, 2019 at 5:02 pm #27258
    Christopher Hunt
    Participant

    Hi Steve,

    Thanks so much for the fast reply. Is the lack of precise time stamping a hardware limitation or software?

    Kind regards,
    Christopher

    February 15, 2019 at 5:45 pm #27259
    Steve Kovarik
    Moderator

    Hi Christopher

    Precise time stamp models are different hardware.
    Why/what do you need precise time stamp for?

    -Best Regards

    February 15, 2019 at 6:25 pm #27260
    Christopher Hunt
    Participant

    Hi Steve,

    Without precise timestamps, the Multitech and other networking hardware that it is connected to is, of course, subject to clock drift. Unless NTP is available then the observation times will be incorrect.

    NTP and the internet, in general, cannot be assumed to be available for us. One of our projects in flight targets some sub-antarctic islands off New Zealand. There is no internet other than the occasional, assumed-unreliable satellite connectivity. Plenty of time for clock drift.

    HTH.

    Cheers,
    -C

    February 15, 2019 at 10:53 pm #27261
    Christopher Hunt
    Participant

    Hang on, shouldn’t I be able to use the GPS time (tmms field)? While not quite precise, it is certainly reasonable.

    `
    tmms | number | GPS time of pkt RX, number of milliseconds since 06.Jan.1980
    `

    March 22, 2019 at 4:03 pm #27511
    Christopher Hunt
    Participant

    I’ve deployed our most recent Network Server that now takes advantage of tmms. However, I’m not seeing that field come in either. Can you please advise on how I may receive the GPS time with a packet forwarder receive packet?

    March 22, 2019 at 6:25 pm #27512
    Jason Reiss
    Keymaster

    See AEP 1.7.2 or mLinux 4.1.6 on downloads page to use gpsd for time sync and packet forwarder timestamps.

    March 22, 2019 at 9:16 pm #27513
    Christopher Hunt
    Participant

    Thanks Jason. I’ll look at upgrading the firmware. So, just to confirm, I should expect to see the tmms field populated by the packet forward along with its receive packets…?

    March 22, 2019 at 11:27 pm #27514
    Christopher Hunt
    Participant

    I just tried upgrading to 1.7.2 from 1.6.4 via a reliable LAN connection. No cigar. I then thought, perhaps try 1.7.0 first. No cigar. Each time the UI reported that there was some error half way through uploading the firmware.

    I then had to power-cycle the Multitech Conduit to recover. This is using AEP. Any further advice? Thanks.

    March 25, 2019 at 8:06 am #27517
    Jason Reiss
    Keymaster

    Did you try the upgrade after a reboot?

    April 3, 2019 at 1:37 am #27592
    Christopher Hunt
    Participant

    Yes, I did try the upgrade after a reboot.

  • Author
    Posts
Viewing 12 posts - 1 through 12 (of 12 total)
  • You must be logged in to reply to this topic.