[Spooks] E10 Happenings

Swartz, Robert K [email protected]
Tue, 06 May 2003 11:51:27 -0600


Hi,


Thank you so much for your effort using your data base to extract potential
repeats

for E10.  I have been looking at your results and have pulled out the
records from the spy numbers 

data base for further study.  So far I have only done the ones for ART.


1. I feel quite confident of the ones for 0100z/0200z and 2200z, not only
because I 

did them myself, but there is some good evidence for it, especially between
0100z and 0200z

with message 95/LGHJU.  The three are strongly tied together by message
93/XTFFU seen at

0100z on 14, 17 April, 0200 on 14, 17 April and 2200z on 13 April.  Other
repeated 

messages include 93/FAVWZ at 0200z between 18 and 30 April and at 0100z
between 18 April and 

01 May.  Additionally, there is the repeat of 74/XIRGQ at 0200z on 21 and 24
March and at 2200z

on 30 March.  (interestingly, the next sked at 0200z on 31 March was ART2).

Since that time frame, heavy interference and my personal schedule have
prevented 

further in depth collection.  Looking back over history for these skeds, I
cannot correlate any 

other repeats involving any of them prior to 2 March 2003 when 95/LGHJU was
reported at 2200z as 

a repeat of the 0200z sked.  More significantly than finding no matches
prior to this time is the 

fact that there were no conflicts (different messages passed on the
different skeds on the same day).

2.  I pulled all the records that applied to the other potential ART
repeats.

a. 2230z/2330z involving 33/HRQZD.  

Between  25 and 28 April 2000, 33/HRQZD was reported active three times on
3415k and 5435k at 2330z.

On 3 May 2000, 33/HRQZD was reported active at 2230z on both 3415k and
5435k.  (Record ids 23613 and

23614).  Also on 3 May 2000 at the same time, on the same two rfs,  30/HLMBJ
was reported. (Record ids

23593 and 23583).  This conflict is probably a result of the mislogging of
the sked time on 3 May for

33/HRQZD. Note that 33/HRQZD was not reported active at 2330z on 3 May.  The
reporter listed for this 

activity is Gallus Galus.  A further search of the 2230z and 2330z sked
activity revealed no other 

potential matches.

b.  2030z/2130z involving 90/KXQRR

Four records are involved

46443  2003 03 14 2030z 5435k		ART 90 KXQRR
46444  2003 03 14	2030z 5435k		ART2 

46475  2003 03 15 2130z 5435k		ART 90 KXQRR
46475	 2003 03 16	2130z	5435k		ART 90 KXQRR


This is also a case of duplicate reporting in the same time slot and RF.
The lack of reporting

on the 14th for the 2130z time slot reinforces the conclusion that this is
an error in record
 
date vice a repeated message.  The reporter listed for this activity is also
Gallus Galus.


	
3.  I  will continue to research the potential repeats that have been
provided as time permits 

and will report them to the group.  Any comments, suggestions, etc are
greatly appreciated.


Bob Swartz+




-----Original Message-----
From: jmm [mailto:[email protected]]
Sent: Friday, May 02, 2003 7:55 PM
To: [email protected]
Cc: [email protected]
Subject: Re: [Spooks] E10 Happenings


Hi,

I just took this opporunity to sync my DB and draw some conclusions. IDs 
referred to hereunder are the serial record number in Smolinski's DB.

Swartz, Robert K wrote:
>
> 
> 
> To all:
> 
> I have been doing some monitoring of the E10 family and would like to
share
> some stuff with you all.
> 
> 
> 1.  There appears to be some daily repeats of messages.
> 
> 	ART repeats the same message at 0100z, 0200z, and 2200z.
> 
> 	EZI repeats at 2000z and 2200z and at 0130z and 2130z.
> 

Now, time for some SQL wizardry with 'GROUP BY' tricks :

My DB has some 47722 logs, 16703 of them are E10 likes. It is mainly 
drawn from Smolinski's DB but with some extra correction (logs expressed 
in MHz removed when not able to correct them, typos corrected when 
noticed on time and asked to original submitters, logs in the future of 
their date of submitting due to confusion in date formats, etc...)

If I take apart logs the comments of which contains the followong 
strings 'CIO2', 'MIW2', 'KPA2', 'SYN2', 'VLB2', there are 10838 records 
remaining mainly composed of regular H+00 and H+30 transmissions plus 
the E10As.

After having put all logs starting between H+01 and H+29 to H+00 and 
those starting between H+31 and H+19 to H+30, I grouped the comments and
the resulting starting times

This gives 5203 distinct couples. I then do a selection retrieving 
record with identical comments and different starting times, and 
manually browse the results.

Bingo !

  ART 33 HRQZD                               | 2230
  ART 33 HRQZD                               | 2330
  ART 74 XIRGQ                               | 0200
  ART 74 XIRGQ                               | 2200
  ART 90 KXQRR                               | 2030
  ART 90 KXQRR                               | 2130
  ART 93 FAVWZ                               | 0100
  ART 93 FAVWZ                               | 0200
  ART 95 LGHJU                               | 0200
  ART 95 LGHJU                               | 2200

  EZI 149 NDKZK                              | 0130
  EZI 149 NDKZK                              | 2130
  EZI 29 JKZLO                               | 1730
  EZI 29 JKZLO                               | 1800
  EZI 69 LXNIY                               | 2000
  EZI 69 LXNIY                               | 2200
  EZI 74 OBFBC/6840                          | 0230
  EZI 74 OBFBC/6840                          | 0330
  EZI 74 OBFBC/9130                          | 0230
  EZI 74 OBFBC/9130                          | 0330
  EZI 84 JECAT                               | 2000
  EZI 84 JECAT                               | 2200
  EZI 89 JCOVQ                               | 1400
  EZI 89 JCOVQ                               | 1530


  JSR 64 CLXXC                               | 2100
  JSR 64 CLXXC                               | 2130

  ULX 64 ITDBP                               | 0100
  ULX 64 ITDBP                               | 0130
  ULX 74 NUVGG                               | 0230
  ULX 74 NUVGG                               | 1800
  ULX 91 LWWVU                               | 0000
  ULX 91 LWWVU                               | 0230
  ULX 91 YTSWF                               | 0230
  ULX 91 YTSWF                               | 1800

  YHF (36)SGDNM                              | 0200
  YHF (36)SGDNM                              | 0230
  YHF (45) KJKJM                             | 0200
  YHF (45) KJKJM                             | 0230
  YHF (55) RQBRG                             | 0200
  YHF (55) RQBRG                             | 0230
  YHF (76)FZZMY                              | 0200
  YHF (76)FZZMY                              | 0230
  YHF 50 NLMTI                               | 2100
  YHF 50 NLMTI                               | 2300
  YHF 58 XWWRP                               | 2100
  YHF 58 XWWRP                               | 2300
  YHF 72 VUYDS                               | 2100
  YHF 72 VUYDS                               | 2300
  YHF 73 SHTSB                               | 2100
  YHF 73 SHTSB                               | 2300
  YHF 89 YSCGB                               | 2130
  YHF 89 YSCGB                               | 2200
  YHF 9 BYVRC                                | 0130
  YHF 9 BYVRC                                | 0230

So, as we can see, there are many apparent repeats. I say 'appearent' 
because this approach suffers some flaws.

  - It relies on a consistent type of commenting a log
    For instance :
    1 log by monitor A at 1230z 'EZI 62 ZAGXS' and the same
    tx at 1800 but with 'EZI 62 ZAGXS weak' will not be detected

- It is not exempt from input flaws; although it could be easy to
   determine with the number of occurences from the original table.

Ok, so far so good. The existence of repeats has been verified. An we 
already have some leads to check.

For instance :

"Are these repeats consistents over a long time, or do they appear for 
some time and then vanish ? "

"In the later case, are repeats more frequents when there are string 
msgs ? "

A more thorough analysis is needed, but as it is nearly 0200B, I will 
stop here (Pillow is calling and it has a loud voice). Maybe will I 
publish a fully detailed study on that topic if interest is shown.



> 
> 	It is really difficult for me to get on my receiver between 2000z
> and 2200z.
> 
> 	Is anyone interested in covering those skeds?
> 
> 

---8<----

 From the top of my head, the longest ever running msg I can remember 
was 68 FSTYL on YHF.

First log by John Maky
2000-01-13 0230 7918 (id=18721)
Last  log by Gallus Galus
2000-03-07 0230 7918 (id=21292)

So it ran for at least 55 days.

The previous log for YHF at this time slot with a different msg is from 
  1999-12-30 (id=18283), the following one dates from 2000-03-21.

So, at the most 68 FSTYL ran from 1999-12-31 to 2000-03-20, i.e. 
possibly for 80 days.

> 
> 3. We do get some long running messages on E10 on occasion.  That is, the
> same message
> 	is repeated at the same time every day (or at least it is there
> every time we look at it).
> 	Currrently the longest one that I have observed is from EZI on the
> 2330z sked.  Message
> 	79/LRYYX was first heard 1 April 2003,  observed last night 1 May.
> Not the longest 
> 	runner, but the longest we have as of right now.
> 
> 
> Would appreciate any responses, additional information on E10.  Don't
forget
> to send your logs in to
> the data base.
> 
> 
> 
> Bob Swartz
> Lovettsville, VA
>