Re: SETI Encounter 2001/Dutil Message


Yvan Dutil (Yvan.Dutil@drev.dnd.ca)
Wed, 08 Sep 1999 08:36:59 -0400


At 12:16 AM 9/8/99 +0100, you wrote:
>Having just caught up with the Encounter 2001 message (there's
>certainly less technical information on their site than
>>SETI@Home's), the following thoughts and questions struck me
>after a quick look at the maths section and the overall structure.

By teh way, where did you find your informations?

>Why is the Mersenne prime introduced before power-of and subtract
>operators are defined (although they might be used to define those
>operators, the next question may be relevant)?

I know this a problem. This is cause by the necessity of using as
efficiently as possible the space available. This is the problem cause
by the close format format approche we used. In the next edition, we
will try a new scheme of coding called "open format" which will allow
us to change the length of each page. To do this we will use a ternary
modualtion.

>Why is power-of not represented by an operator character - the notation
>seems to be based on human typographic conventions, not on any fundamental
>principle (the only advantage is that the operator precedence is rather
>clearer in the Mersenne prime)?

Many peoples had this comment. We did this because it save on caracter. Since
we cant change the length of a line this as been prove useful in some case.

>Why is the mathematical message at the start of the transmission (my guess
>is commercial reasons, although the fast code for the greetings might also
>make the receiver think it not worth decoding)? I would want at least
>one copy at the end to allow time to bring the most sensitive receivers to
>bear and to optimise them for the signal.

I thougth about this scheme, but I had not much control over this part
since the
russian was responsible for the transmission

the scheme follow as been this one:

Start of Session
 Dutil Message
 Braastad Message
 Arecibo Message
 Encounter 2001 Staff Message.

             Part II
End of Session

The Part I was transmitted three times at a rate 100 bits per second, and
the Part II - only once at a rate 2000 bits per second.

>I've established that there are three signalling states, 0, 1 and pause,
>but how are these used to represent clock and data information, e.g. is
>1 0 1 sent as P1P0P1P... with each of equal length? Where are runs of
>Ps used?

We define time using, the central frequency, and the length of one pixel.

>What pulse shape is used?

Frequency modulation, reactangular shaped.

>What is the frequency stability?

one part in 10^14 in think.

>Are the borders shown in the GIFs actually transmitted (I would hope so,
>as they provide synchronisation patterns)?

Yes, the farme is one of trick, which really help the decoding.

>Is there any clock run-in and training sequence? I guess that, if you
>have the whole sampled signal to post-process, you don't really need a
>clock run-in, but training is less clear; depending on the nature of
>the propagations distortions and how they vary with time, it might pay
>off to include training sequences at beginning and end (you can play the
>data backwards, or interpolate) and ideally at intermediate points).
>
>(Training sequences are predictable patterns which start simple and
>then build up in complexity, so that you can calibrate the channel with
>more and more accuracy.)

No, we dont have these. We were too limited in time. However, the message
shoudl not suffer to much from distortion induce by the interstellar medium.
It should not be a problem for a few thousard ligth years.

Yvan Dutil



This archive was generated by hypermail 2.0b3 on Sun Oct 10 1999 - 15:46:34 PDT