Hi Markus
The communal sever is running the
decision making part of the dynamic system ,
along with distributing the call signs for
the list , the call signs are also
vetted before distribution to prevent
'random' events
The web linking transports from the Rx's call
/ s/n / % fade / time or call / s/n /
time [ no % fade for dynamic ] , there is
no DSP shared and as such cannot
influence the decision of each Rx station
The satellite Rx's provide the
time stamp , and as can be observed , [
if the windows software is used .. ] the
spots from the dynamic detectors are
transferred to the server , then after a
finite delay . are distributed to the Opera users
. with the ''??'' stamp 'single' or 'false'
or no '??' then validated .
The Opera system functions , with or
without web linking . in a stand alone ,
dynamic detection , the s/n provides the
confidence level , other than that , If the TX
is web-linked , the psk-map data base holds
the TX beacon .. without this , then its difficult
to evaluate , unless there exists a string
of decodes / detections , from which a profile may
be established or other Rx stations have
decoded/detected the station
There being no master clock , the
system runs as simultaneous unique individual
decodes/detection's , the only not allowed condition ,
would be a random detection , occurring
within the time validation window of a genuine Tx
..But that event would need to share
the call sign with a actual event to
be of importance
In this , three Dynamic Rx stations
share the same time stamp , and are validated [
green] two would of also validated each
other.
18:38 477 VK3ELV de VK5FQ Op8 Deep Search 437 mi -36 dB
in
18:38 477 VK3ELV de VK2XGJ Op8 Deep Search 290 mi -36 dB
18:38 477 VK3ELV de VK5CV Op8 Deep Search 431 mi -36 dB
A local off air decode ,
simply provides a live input to the server ,
which in turn evaluates the situation and is
suggested for remote or QRP TX sites.
Dynamic is only provided
for 477K and 136 K , all other bands
are single mode
73-Graham
G0NBD
Sent: Friday, May 22, 2015 6:52 PM
Subject: Re: LF: 477 VK5CV de VK2XGJ Op8 690 mi -17 dB F:8% in
Dapto, Austealia with 25w
Hi Graham,
> ... note Phil is also running RX and may
provide shortly a secondary , off-air decode
to provide a time stamp.
maybe this is not such a good
idea. When the timestamp from a local decode is propagated over the Opera web
link, all other active decoders immediately get this info. A sceptic (are
we?) could suspect that this might be used by the detectors to
correlate their last-received noise against the known message and
timeframe, using an arbitrarily low detection threshold - a la "You've
heard it too, haven't you? You surely have...". This could make false positives
very likely, but no one can tell them from a real one. The Opera
program then has four minutes (one on MF) to output such a
hit as a delayed dynamic deep search decode. afterwards, the matching
timestamps would be used to falsely "verify" a false detection,
which had really been provoked by exactly
that timestamp.
In my opinion, it would enhance
credibility to not propagate local decodes, but let the transmitter
only keep a "secret" log of their emissions. This could then be used in an
a posteriori comparison against correlation results, which
have been created independently without a priori knowledge of transmit
times.
All the best,
Markus
Sent: Friday, May 22, 2015 3:19 PM
Subject: LF: 477 VK5CV de VK2XGJ Op8 690 mi -17 dB F:8% in Dapto,
Austealia with 25w
|