Mode S preamble detection

Hello!

This is my first post on the site.

I would like to ask a question if I may. I am trying to decode Mode S
reply signals for my diploma thesis. The first thing I must get up and
going is preamble detection. My teacher suggested that the analog
processing before sampling would remove any DC component of the signal. He
also suggested using a 16 MHz sampling frequency.

Seeing as how the coding is Manchester, I was thinking of using a matched
filter (time reversed and delayed version of the "1" bit symbol as an
impulse response), checking the output values at 16 sample intervals and
using an FSM to recognize the correct sequence: a "1" symbol would output
16, a "0" symbol would output -16 and a null symbol (meaning the spaces
between pulses in the 2-3; 5-6; 6-7 and 7-8 us timeslots), which would mean
checking for the following sequence of filter outputs: 16 16 0 -16 -16 0 0
0.

Does anybody have a better idea on how to do this?

I was thinking of using a correlator, but the problem is all those long
sequences of "0" bits which ruin correlation (as anything multiplied by 0
would also yield 0).

Thank you,
     Tudor.

On Nov 19, 6:26&#4294967295;am, "tudorelu87" <tudorel...@yahoo.com> wrote:
> Hello!
>
> This is my first post on the site.
>
> I would like to ask a question if I may. I am trying to decode Mode S
> reply signals for my diploma thesis. The first thing I must get up and
> going is preamble detection. My teacher suggested that the analog
> processing before sampling would remove any DC component of the signal. He
> also suggested using a 16 MHz sampling frequency.
>
> Seeing as how the coding is Manchester, I was thinking of using a matched
> filter (time reversed and delayed version of the "1" bit symbol as an
> impulse response), checking the output values at 16 sample intervals and
> using an FSM to recognize the correct sequence: a "1" symbol would output
> 16, a "0" symbol would output -16 and a null symbol (meaning the spaces
> between pulses in the 2-3; 5-6; 6-7 and 7-8 us timeslots), which would mean
> checking for the following sequence of filter outputs: 16 16 0 -16 -16 0 0
> 0.
>
> Does anybody have a better idea on how to do this?
>
> I was thinking of using a correlator, but the problem is all those long
> sequences of "0" bits which ruin correlation (as anything multiplied by 0
> would also yield 0).
>
> Thank you,
> &#4294967295; &#4294967295; &#4294967295;Tudor.

I doubt most people here know what "Mode S" means....
are you talking about aircraft ATC transponders?

Mark

>On Nov 19, 6:26=A0am, "tudorelu87" <tudorel...@yahoo.com> wrote:
>> Hello!
>>
>> This is my first post on the site.
>>
>> I would like to ask a question if I may. I am trying to decode Mode S
>> reply signals for my diploma thesis. The first thing I must get up and
>> going is preamble detection. My teacher suggested that the analog
>> processing before sampling would remove any DC component of the signal.
H=
>e
>> also suggested using a 16 MHz sampling frequency.
>>
>> Seeing as how the coding is Manchester, I was thinking of using a
matched
>> filter (time reversed and delayed version of the "1" bit symbol as an
>> impulse response), checking the output values at 16 sample intervals
and
>> using an FSM to recognize the correct sequence: a "1" symbol would
output
>> 16, a "0" symbol would output -16 and a null symbol (meaning the
spaces
>> between pulses in the 2-3; 5-6; 6-7 and 7-8 us timeslots), which would
me=
>an
>> checking for the following sequence of filter outputs: 16 16 0 -16 -16
0 =
>0
>> 0.
>>
>> Does anybody have a better idea on how to do this?
>>
>> I was thinking of using a correlator, but the problem is all those
long
>> sequences of "0" bits which ruin correlation (as anything multiplied by
0
>> would also yield 0).
>>
>> Thank you,
>> =A0 =A0 =A0Tudor.
>
>I doubt most people here know what "Mode S" means....
>are you talking about aircraft ATC transponders?
>
>Mark
>
>
>Yes, exactly.

On Nov 19, 6:26=A0am, "tudorelu87" <tudorel...@yahoo.com> wrote:
> Hello!
>
> This is my first post on the site.
>
> I would like to ask a question if I may. I am trying to decode Mode S
> reply signals for my diploma thesis. The first thing I must get up and
> going is preamble detection. My teacher suggested that the analog
> processing before sampling would remove any DC component of the signal. H=
e
> also suggested using a 16 MHz sampling frequency.
>
> Seeing as how the coding is Manchester, I was thinking of using a matched
> filter (time reversed and delayed version of the "1" bit symbol as an
> impulse response), checking the output values at 16 sample intervals and
> using an FSM to recognize the correct sequence: a "1" symbol would output
> 16, a "0" symbol would output -16 and a null symbol (meaning the spaces
> between pulses in the 2-3; 5-6; 6-7 and 7-8 us timeslots), which would me=
an
> checking for the following sequence of filter outputs: 16 16 0 -16 -16 0 =
0
> 0.
>
> Does anybody have a better idea on how to do this?
>
> I was thinking of using a correlator, but the problem is all those long
> sequences of "0" bits which ruin correlation (as anything multiplied by 0
> would also yield 0).
>
> Thank you,
> =A0 =A0 =A0Tudor.

Have you seen this?

http://adsb.tc.faa.gov/WG3_Meetings/Meeting12/1090-WP-12-15.pdf