Quote:
|
Originally Posted by Mush Martin
That was a large part of the OIC experience they had to convince the
admiralty that not only could they use it to review historical situations
for insight as the germans predicted they would be able to do. they
also had to convince them that the submarine tracking room could
forecast future movements of individual uboats and raiders based
on established patterns of normal behaviour and precedented patterns
of deviant behaviour. they were lucky to have Roger winn as there
First attempt analyst an intelligent and insightful man he was indeed
able to well predict a better than average percentage of movements
and patterns and showed an uncanny insight for being able to
pull together unrelated threads into a firm prediction of future
action, more importantly with the precedent and patterns established
So also were his succesors able to also.
the OIC predicted the channel dash two weeks or so prior to it
happening based the decrypts and traffic patterns and movements
of Kriegsmarine units such as destroyers and minesweepers
(each of which used totally different cyphers)
as well as Photo Reconassaince and traffic from OKM.
Sadly "A series of unfortunate events" including two seperate
photo reconnasaince flights in different areas that missed seing
the sortie due to mechanical failure, allowed the german units
to gain an early leadoff and they managed to maintain the
initiative through the voyage where inertia ruled the day
for the coastal and air forces.
M
|
One of the major problems with using non-cryptanalytical signals intelligence is convincing people that it means something.
The sinking of HMS Glorious is a perfect example: Harry Hinsley at BP was able to deduce through traffic analysis and HF/DF that major elements of the Kriegsmarine were going to sortie into the North Sea. He passed the warning on to OIC, who promptly ignored it. Not long after, the carrier HMS Glorious (who's captain, had he survived, should have been courtmartialed for not posting lookouts or have a CAP up in the air) was attacked and sunk by the Scharnhorst and Gneisenau. The OIC realized that had it listened to Hinsley and broadcast a warning, the Glorious might not have had her guard down, and might not have been surprised. It was an expensive lesson.
Still, the intelligence provided by Hinsley wasn't specific. It merely said that major units of the Kriegsmarine could be expected to sortie from the Skagerrak. That kind of intelligence, while useful in a defensive sort of way, has little offensive value: It doesn't tell you the where and when necessary to "get thar firstest with the mostest"
Another major problem is that TA and HF/DF are more susceptable to being 'spoofed'. The Japanese did this with the Kido Butai that attacked Pearl Harbor: They left the radio operators from their ships on shore in Japan to simulate normal traffic while the actual task force maintained radio silence. You can also do what the Allies did before D-Day: set up a communications structure that simulates an entire Army group, when in reality it is just a relative handful of operators. It's hard to do this, though, because the fake radio traffic has to be cohesive and consistent with what one would expect of an actual unit. This was simplified for the Allies, because the idea was that this was an Army group training, so pro-forma practice traffic solely to exercise the operators would be what you expect to see.
It is harder to do that kind of thing on an operational basis (ie., the putative units are supposed to be front line units), as pretty quickly the analysts will discover that what is being said in the traffic is not being done in real life. It's a fool me once, won't get fooled again kind of thing. You can get away with it on an occasional basis, as the US showed before the Battle of Midway: They knew, through decryption of the JN-25 code, that the Japanese were going to attack a target they designated as 'AF'. The analysts at station HYPO in Hawaii thought 'AF' might be Midway (based on the fact that in previous decrypts the 'A' series stood for Hawaiian locations), but Washington disagreed. To resolve the issue, they cabled Midway to send via radio a message that their desalination plant had broken down. A couple of days after this was done, we intercepted a Japanese message stating that 'AF is short of water'. This was a one-shot deal, however. Keeping up a deception that is internally and externally consistent would be extremely difficult.
However, the inferential nature of Traffic Analysis, outside of cryptanalysis of the actual message contents, means that the spoofing side can keep the charade going for longer. If successful to any degree, it can also make the other side doubt the intelligence derived from traffic analysis/DF, which is a significant benefit.