Loading...
 
Janus general discussion forum

Janus general discussion forum

Forum List Topic List

Forums » Janus general discussion forum » Revision of JANUS STANAG next year in March
 
Thread actions
Print this page
  • Print all pages

    • Revision of JANUS STANAG next year in March

    Posted by ivor.nissen on 2018-08-11 06:42 posts: 17
    Dear JANUS friends,

    in the ANEP you can find our concerns about the "Values provided in Table in Annex B are likely not to be implementable in the hardware to be used in the given accuracy." page V

    The same for ANNEX C. Here the case is more difficult! The values of
    the table C-1 are not fitting to the formula, in my view. If you are
    using 50 digits, the correct basis is not

    1.176769793407883 it is
    1.176769793407882894...

    to generate the values of the table.

    If you have the accuracy (only MAPLE, MATHEMATICA, ... - not on a
    simple computer) you can calculate the index correctly

    Index = ln(time in s)/ln(1.176769793407882894...) + 21

    We have to change it to fix "human readable" numbers:

    Proposal: frac(1000*FunctionValue(Index))/1000

    BR Ivor
    Reads: 585328
    Link
    Posted by ivor.nissen on 2020-10-30 18:51 posts: 17
    Dear JANUS friends,

    in the NATO standard ANEP-87 - „Digital underwater signalling standard for network node
    discovery and interoperability“, appendix B and C two time scaling formulas are being used.

    The problem:

    1) The values provided in Table in ANNEX B are likely not to be implementable in the hardware to be used in the given accuracy. page V

    2) The same for ANNEX C. Here the case is more difficult! The values of the table C-1 are not
    fitting to the formula. If you are using 50 digits, the correct basis is not

    1.176769793407883 it is
    1.176769793407882894...

    to generate the values of the table.

    If you have the accuracy (only MAPLE, MATHEMATICA, ... - not on a
    simple computer) you can calculate the index correctly, otherwise NOT!

    Different computers, different time slots!

    3) For some/many applications we need not the time, we need the number of bits
    in form of a length field.

    In Janus we have the situation 2 chips = 1 bit (with the coding). In 2.1.11 2. the Chip
    duration is Cd = 0.00625 s. With this we have an interaction between bit and time!

    Each index in the table in ANNEX B and C can interpret as time AND length in bit!
    like (index N, time = intN * Cd, length_in_bits = intN)

    With this concept we have a length field without changing the allocation map, can
    transmit JANUS cargo also in case of repetitions

    Best regards

    Ivor
    Link