Talk:Signals intelligence/Archive 1

Latest comment: 14 years ago by 24.218.32.186 in topic Development thoughts
Archive 1

Untitled

Does anyone know how to add : http://en.wikipedia.org/wiki/Norwegian_Intelligence_Service Next to FAS with the Norwegian flagg, at the end of this page. ?

External links: I have restored two links to commercial sites. I have no financial interest in either. They are of interest because (a) Narus is one of the suppliers of hardware involved in the current AT&T / NSA kerfuffle, and (b) Applied Signal Technology is cited in Duncan Campbell's report to the European Parliament, "Interception Capabilities 2000". People interested in SIGINT may well find the spooky stuff sold by these companies interesting. Harumphy 23:11, 23 May 2006 (UTC)

Still not compliant with WP:EL, if you think there is something noteworthy about them then bring it into the article in text and use the link to expand on the topic. Without some form of anchor in the article they're nothing more than linkspam.ALR 07:19, 24 May 2006 (UTC)
WP:EL does indeed say that external sites "that primarily exist to sell products or services" should not normally be linked to. It does not rule out such links in every case. Given the massive secrecy and dearth of primary (or credible secondary) source info about SIGINT, it's necessary to grab what little there is and make the most of it. IMHO that justifies an exception in this case. This does not make it linkspam, unless you are using a very different definition of the term. More pragmatically, if you are interested in SIGINT and want to learn more, the content of these sites is very interesting, as Duncan Campbell's work has already shown. Harumphy 18:03, 24 May 2006 (UTC)
Perhaps it'd be better to say something, then, like "investigative journalist Campbell has shown/argued that...", rather than add links to websites, which would mean little to most people without any context. — Matt Crypto 18:19, 24 May 2006 (UTC)
Whilst I see your point, without context the links are meaningless to the casual reader. Grabbing what you can and making the most of it risks creating the proverbial 'castle in the air'. Using your argument it would be legitimate to add links to Microsoft, various flavours of Linux, Sun Microsystems etc just because community use those OSs.ALR 06:56, 25 May 2006 (UTC)
I see there is now a page about Narus anyway, which solves that problem. I might create a page about App. Sig. Tech. later. Harumphy 10:06, 25 May 2006 (UTC)

WWI Traffic Analysis Incident

I referred to an error caused by an admiral, Jellicoe I believe, going to the traffic analysts and asking the position of a station ID. That was the ID of the German fleet headquarters, and he was told it was on shore. Since he demanded only that specific information, no one told him that when the High Seas Fleet sortied, the commander used a different ID. This led to the British not being prepared to meet the fleet, since the admiral did not understand it had sortied.

What I don't remember is whether the resulting battle was Dogger Bank, or just a late reaction to Jutland. The conversation, IIRC, is in Kahn's The Codebreakers, which I don't have physically at hand. Can anyone document this reference? Howard C. Berkowitz 14:03, 1 October 2007 (UTC)

It's on pp. 272-273 of Kahn, 1967 ed. The admiral was Jellicoe, the call sign "DK", the response to the query about where DK was located was "in the Jade River." The ensuing battle was Jutland. The result was that Jellicoe later ignored a correct decrypted report on steering orders given to the German fleet and thus missed his opportunity for a decisive battle. Jellicoe's faith in cryptographic intelligence was also shaken by a decrypted report that placed the German cruiser Regensburg near him. It turns out the navigator on the Ravensburg was off by 10 miles in his position calculation.--agr 14:19, 1 October 2007 (UTC)

Direction of the article?

I've added a good deal in an article that should be "National Means of Technical Verification", but I misnamed it "National Technical Means of Verification" and am waiting to fix it.

SIGINT isn't in enormous detail; it's mostly MASINT and some IMINT. I'm not sure I'm brave enough to go and redo MASINT, as it's such a catch-all although there are some subtle unifying aspects. The hardest thing to follow is that MASINT might use a SIGINT sensor to capture signals, but then things branch: there could be ELINT or COMINT analysis on the deliberate signals captured by a SIGINT platform, but MASINT goes into the inadvertent side characteristics of the signal. The way these intertwine, and, for that matter, the bureaucratic management of them in the US, are not, to put it mildly, simple.

Comments are welcome about whether things like BEADWINDOW are too specialized for SIGINT, as they are more communications security than communications intelligence -- but if you don't do them, the other guy will be able to collect COMINT on you. Incidentally, I don't yet see a reason for separating ELINT and COMINT into separate articles. BEADWINDOW and related operational COMSEC practices don't fit nicely elsewhere, as most of the COMSEC articles are about cryptography. Howard C. Berkowitz 11:40, 3 October 2007 (UTC)

Howard
I think a lot of what you're adding is useful, particularly around the relationships between the various disciplines. I think that the style needs working on, it's very conversational at the moment, but that's something which can be resolved once all the content is in place. However I do think that quite a lot is verging on, if not striding into, the realms of original research. Some of that may be a result of the writing style, but quite a lot is vulnerable to challenge if you don't substantiate it with sources.
I would add a health warning on sourcing, the WP approach to reliability is pretty flawed, there appears to be no willingness to consider sources on their merits for reliability, consistency etc as an OSINT or all-source analyst would.
ALR 17:14, 11 October 2007 (UTC)

Moving SIGINT discussion from HCBerkowitz's talk page

This article is growing and needs to be split in a coherent way, and it interacts, at the very least, with the MASINT material I'm writing, as well as existing, if stubby, articles on ELINT and COMINT. Organizational ideas gratefully accepted! Howard C. Berkowitz 14:09, 16 October 2007 (UTC)

Nice work expanding the article! Big improvement. Binksternet 02:20, 1 October 2007 (UTC)

How about SIGINT = Current & a separate History of Sigint Perhaps Sigint in WWI & Sigint in WWII out separately, as they can then link sideways into other WWI & WWII articles. But that still leaves any Sigint before WWI (?) & Sigint Interwar & Sigint After WWII to put somewhere, as well as breaking up the development thread! So History of Sigint?? (I think the comments belong here on the Talk part!) Hugo999 13:55, 16 October 2007 (UTC)

That is a good idea. Perhaps I should move all this to the SIGINT talk page.
I wonder if there might be a main article on the principles and definitions, and then split out some of the history. You are quite correct that could break up the flow, and there is a question about the scope of historical articles. Arbitrarily, I started discussing things as of World War I, but there certainly were things in the 19th century directed at telegraphs, and encryption & steganography seemingly as old as writing. Howard C. Berkowitz 14:05, 16 October 2007 (UTC)

First spinoff

If for no other reason that editing was getting too slow, I created SIGINT by Alliances, Nations and Industries. It's not the greatest title, but it works, and lets me reduce the size of the main SIGINT article. Even this spinoff is not a small article. Contributions to the introduction of this article would be especially welcome. —Preceding unsigned comment added by Hcberkowitz (talkcontribs) 22:13, 16 October 2007 (UTC)

Second Spinoff

I'm creating SIGINT in Modern History as a second spinoff, again for pure size reasons. The spinoff is almost 100K, and I still have more material for it. Is it feasible to try to split out further work, and articles, by period? That will work best before the Cold War, as after that point, there were projects spanning decades, far worse than the interwar UK and US efforts in the twenties and thirties. Howard C. Berkowitz 22:37, 19 October 2007 (UTC)

Another option

Back to the basice quetion, if ELINT and SIGINT should be merged - well, as to the definition - both elint and comint are sub-desciplines of SIGINT. so if comint is merged with sigint, so should elint. BUT, and its a big "but", since sigint is such a wide topic, and both elint and comint are huge topics, with, ofcourse, some alikenesses, though even more - differences, i asume that perhaps it will be best to have 3 different articles - ELINT, COMINT, and short general information on the concepts of both, under SIGINT. Regards! Comint 12:21, 25 October 2007 (UTC)

COMINT isn't now an article, but simply points back to SIGINT. ELINT is less detailed than other topics in the existing SIGINT. SIGINT does need to contain the common techniques. As long as it does, and there is simultaneous buildout of both COMINT and ELINT, then breaking them out may be appropriate. I've already split three sub-articles out of SIGINT. Did you want to take on the separate articles?
Again, things do need to be sourced and can't be an advertisement. EOB is a valid thing to have in the article, but the techniques to produce it are not unique to Genesis. Howard C. Berkowitz 15:48, 25 October 2007 (UTC)

EOB additions

I'm not just going to revert the additions related to Electronic Order of Battle right now, but I will note that the article that contained the material before was deleted several times under the Speedy delete process and discussed on the Milhist talk page with a deletion review which endorsed the decision.

I stand by my position at the time, the concept is certainly of value, but advertising for a specific product set is not.

ALR 13:32, 25 October 2007 (UTC)

The material reads like an ad in places and is not intelligible in others. Links were dead, pointed to paid pages, or had little to do with what was being sourced. Howard C. Berkowitz 07:34, 26 October 2007 (UTC)

Suggestions

HCB...some suggestions just from what I've seen by a cusory glance over:

1. The lead needs reworked, I believe. It's a bit too in-depth for the opening lead...perhaps simply mentioning it as part of the intelligence cycle dealing with Signals Intelligence...then have a first section explaining the technical definitions, etc.
2. It needs a lot more citations for an article this length.
3. The length itself is a problem. I'm not versed enough to tell you where to split it, or if it should simply be parsed down instead.
4. There is one graphic in the whole article...it could be improved with some more diagrams or pictures or the like.

That's what I've seen for now. If you'd like, when I have the time I can do a full read through of the article. Just let me know.Cromdog (talk) 18:51, 4 February 2008 (UTC)

I'm open to suggestions on the lead. Remember that this article has several spinoffs. You will also find many more references in the other articles. You may see references there that should be replicated here. It's an enormous subject, and I would truly appreciate constructive criticism.
There should be 3 graphics:
  1. Wullenweber array
This one now appears, didn't last time.
  1. Spectrum analyzer, which seems broken One appears to be broken, and I'll probably try to draw a better picture of SIGINT spectrum analysis, maybe using a waterfall display (not that there are many sources on them).
This one is definetly broken...
  1. Electronic order of battle flow diagram.
Checks out ok.Cromdog (talk) 20:23, 4 February 2008 (UTC)

Are you not seeing at least one other, since I'm finding the spectrum analyzer broken? I'm sketching realistic SIGINT spectrum analysis displays, and wonder if they will need their own article to explain! :=(> (I have a beard) Howard C. Berkowitz (talk) 19:06, 4 February 2008 (UTC)

I have class shortly, but this evening I'll give this one a thorough run thru and see what I can come up with. This one would probably be B-class if it had more inline cites and a little parsing down, but I'll give more details later. Congratulations on the fine work on these intelligence articles, I've been running across most of them as I work through the MILHIST Unassessed articles.Cromdog (talk) —Preceding comment was added at 20:28, 4 February 2008 (UTC)

Boer Radio Interception

The section on Boer radio interception ("The Boers had captured some British radios, and, since the British were the only people transmitting at the time, had signals rather obvious to intercept.") is incorrect, and seems to be based on misunderstanding of the cited source. Page 4 of cited text:

As early as 1900 in the Boer War, the Royal Navy in South Africa appears

to have used wireless sets inherited from the Royal Engineers to signal from the neutral port of Lourenço Marques “information relative to the enemy” albeit in violation of international law. [71] This first use of radio for intelligence purposes depended, of course, on the inability of others to intercept the signals, but in 1900,

only the British in that part of the world had any wireless capability.

It could also be supplemented by discussion of Philip Pienaar, "a Boer telegraphist, (who) once got "the entire plan of campaign for the next four weeks" by tapping General Hamilton's telegraph line."

Zipzipzip (talk) 19:20, 12 April 2008 (UTC)

"Defensive SIGINT" Section Intro Paragraph

This section's opening paragraph is crazy and needs to be rewritten entirely. I tried fixing a sentence, but then realized the entire thing was silly. Please help. 65.191.115.91 (talk) 02:59, 14 April 2008 (UTC)

Mostly fixed now, thanks to user 'Hcberkowitz'. Needs a few more eyes to go over it. 65.191.115.91 (talk) 01:05, 16 April 2008 (UTC)

Defensive SIGINT as howto?

Absolutely disagree; a basic library of actual documents on doing defensive SIGINT fills a large bookcase, or, more likely, a large safe.

The defensive issues here all have been involved in one or more international incidents, such as the Soviet bugging of the U.S. embassy in Moscow, or the revelations about British RAFTER intercepts of foreign diplomatic communications. In other words, everything here have policy implications, with the description being the minimum needed to describe the nature of the threat and countermeasure.

Howard C. Berkowitz (talk) 15:43, 20 May 2008 (UTC)

The problem is the language, not the content. Some excerpts:
  • One must begin by defining the threat
  • When using radio transmitters, use directional antennas
  • If possible, avoid transmitting when
  • avoid using regular commercial power to transmit
  • Use highly variable transmission schedules
These feel as coming straight from a manual. Hint: The Imperative mood shouldn't be used in a wikipedia article. Morana (talk) 20:34, 20 May 2008 (UTC)
Fair enough. Looking at my personal time, I've come to the position that I am simply not going to worry about the MOS. As long as the information is present, I am totally unaffected if all, or none, of my articles ever advance beyond Start class over stylistic issues. It would be too dangerous to my professional writing to get into the habit of following "encyclopedic" style.
It would be totally irrelevant to me if all my articles suddenly were made Featured Article. The idea of "encyclopedic" style is, more and more, making me look for alternate venues.
Feel free to edit, but I'm not going to take the time for non-content related issues, and "mood" is, to me, unrelated to content. Howard C. Berkowitz (talk) 21:10, 20 May 2008 (UTC)

Scope and term definition of SIGNINT

From the article it is quite obvious that Signals Intelligence is defined as interception of radio waves. However, interception in cable is a significant part of intelligence today. Does that fall under another definition? MaxPont (talk) 08:35, 10 July 2008 (UTC)

SIGINT is the process of producing intelligence from intercepted signals, regardless of the medium through which those signals are conveyed. Signal Intercept is the process of collecting the material which is then processed, in the context of an intelligence requirement, to produce the intelligence. All that includes the intercept of traffic in glass...
ALR (talk) 12:11, 10 July 2008 (UTC)
OK, intercepting a copper/fiber sea cable with a submarine could be considered Sigint. But how about intercepting traffic on a higher level (such as IP packets)? And how about if IP traffic passing a certain router is monitored and filtered as a part of a government enacted (secret) program? IMO that should be defined as surveillance (or wiretapping)? My point is that there is a need to define Sigint in a way that does not include everything (e.g. overhearing people talking to each other in a bar). How to draw the line? MaxPont (talk) 13:17, 10 July 2008 (UTC)
That example would be considered collection not SIGINT per se. Intercepting packets, similarly would be a collection activity. Collected material only becomes intelligence when it's been processed in the context of an Intelligence Requirement.
From an analysis perspective any intelligence product would be informed by collection through a range of mechanisms whether that be signal, imagery, open, human or other sources. Analysis of a single mode of collection would not be considered as all source and would have inherent vulnerabilities. Overhearing someone in a bar would be a HUMINT, or potentially ACINT (although that would be an ultra-purist position), collection, making use of at collection might involve tasking of other collection modes to corroborate the data.
ALR (talk) 13:35, 10 July 2008 (UTC)

OK, but that does not clarify the scope for the term SIGINT in the article. The distintion between Collection and Intelligence is not really made in the article and I believe it is only of interest for the experts. I don't think that the article should be split in two - Collection and Intelligence. IMO we should define the articles in the way the terms are used in normal language. Are we to expand the article with all possible ways of intercepting signals - I am sceptical? MaxPont (talk) 13:56, 10 July 2008 (UTC)

In general I would agree, although I would suggest you take the issues up with Howard directly since he wrote the majority of this content.
Intelligence is a collection of disciplines and the relationships between them are as important, if not more important, than the elements of each individually. Intelligence analysis is a holistic discipline. Frankly individual methods of SIGINT collection are of limited value, so I wouldn't advocate going into that much detail, although verifiable sources owuld be difficult for the more esoteric aspects anyway.
I keep thinking about cleaning up this range of articles, but it's a pretty formidable task, for which I don't think I have enough time.
ALR (talk) 14:06, 10 July 2008 (UTC)


The "see also" "ELINT" link is an internal link, not an external link to another page. That's weird and non-standard afaik. Lastdingo (talk) 00:01, 30 August 2008 (UTC)

Development thoughts

I'm going to bring together my thoughts based on a range of discussions going on at the moment. And this in itself might be a tad bitty. I'd say that someone having gripped this article is extremely useful, but what falls out of that is a need to think strategically about the whole Int domain. Unfortunately there is quite a lot of garbage in Wikipedia on the subject, and we've probably got an opportunity to address that.

Good thoughts, though. I'm relatively new with Wikipedia, but the discussions here about looking at the overall structure of a topic. When I first started dealing with some of the computer networking issues, there were complaints that I had to retain things already there, or keep a structure that reflected some vendors' marketing approaches, or some simplifications for teaching & certifications. It seemed a bit odd to be told I had to keep an old secondary or tertiary definition for something I had had a hand in building. Howard C. Berkowitz 19:26, 16 October 2007 (UTC)
It's patchy, I think a lot of people don't really understand what the technology allows and the administranium is poorly managed. What WP needs is some leadership, but that's another debate. It's largely up to those who are active in an area to decide how to structure it, but MilHist has some pretty good guidance on a common style and most are pretty much up to the mark in terms of coming up with credible content.ALR 13:42, 19 October 2007 (UTC)
It may be, with respect to SIGINT capabilities, that the one-eyed man is the king in the Kingdom of the Blind. When thinking, in the same way, about MASINT, it is the Kingdom of the Public Stereotype of the Monty Python Village Idiot, blindfolded, ears plugged, and in a plaster body cast. It was, with some difficulty, that I avoided referencing most of the Village Idiot skit, and there were definitely occasions where wink wink nudge nudge and Blue Parrot fit, to say nothing of the sheep in the wall. Thinking of some conversations with my former retired 3-star boss, and some of his tales from when he was an active recon pilot, Bravely Ran Sir Robin was the recon motto. In the MASINT articles, there is a reasonable amount of sourced, as well as editorial, comment about how one subdiscipline complements another.
Sorry, I didn't sleep well, but it's not irrelevant to try to remember the name of the UK officer, between the wars, brought to trial over violating the Official Secrets Act, one of the charges being disclosing that the head of SIS was called "C". When the judge accused him of having said that British Intelligence was like a Marx Brothers movie, the defendant replied "No, M'lud. I said that compared to the Secret Intelligence Service, a Marx Brothers movie was a crystal-clear portrait of reality." I partially revised the US security classification article last night, and the Marx Brothers came to mind in terms of both the reality of the system and the Wikiperception of the system.
Seriously, yes. I would like to see a consensus outline of what might well be an Intelligence and Security and [foo] project. Security comes in here as first the high-level balance between dissemination and protection (thinking of Wohlstetter and other serious analyses of Pearl Harbor). [foo] is a miscellany including good things such as National means of technical verification and bad things such as political misuse of the intelligence community. I don't know what it's called in the UK, but part of [foo] is what the Soviets called Correlation of Forces and the US tends to call Net Assessment -- when the senior strategic planners of force development sit down and balance it against own-force capability, and perhaps decides on changes in force structure. Howard C. Berkowitz 16:39, 19 October 2007 (UTC)

I think there would be value in an overarching Intelligence article, drawing together the various threads; HUMINT, SIGINT, MASINT, IMINT, OSINT etc and putting that all in the context of Intelligence Support to Operations or Intelligence Support to Government. Neither title is ideal as we need to cover the whole spectrum from Grand Strategic right down to Combat Int.

Especially at the higher organizational levels, the terminology and relationships get a bit sticky. There are reasonable points at which someone balances their own capabilities against their opponents, although, in my experience, that is a process that needs to be kept away from most of the analysts, except possibly (and I'm interested if there's a difference between US and UK terminology here) at the level of the very senior people who coordinate the formal estimates. Beyond that, however, there has been a Pentagon group that then compared the intelligence with our own capabilities. The US term for that, when it wasn't being politicized, was "net assessment". I much prefer the Soviet term of art, "correlation of forces".
Continuing at the high-level/relationship to top government, you probably know that there was something of a US scandal, at least to those with some intelligence exposure, where the "Office of Special Plans" would pick the most ideologically desirable material and send that to the White House. The media called this "stovepiping", as distinct from the more classic definition in the intelligence world, which I added to the Wiki article: IMINT not talking to ELINT not talking to HUMINT. Howard C. Berkowitz 19:26, 16 October 2007 (UTC)
I'm aware of the issue you refer to, and to the risk in the community. It's certainly something that can be talked about in the abstract, and I think Keegan has written on the issue in the military context. there is also the issue of political interference in the process, although finding much that is both reliable and in the public domain about that could be a challenge.
The starting point, on which I wrote a bit, might be going back to the issue of stovepiping in terms of not using different intelligence disciplines to cross-check before it ever gets to politicians. One primary source, starting about 2000-2001, is a series of white papers and Congressional testimony from NSA Directors about rethinking roles and missions for the 21st Century. There are also some papers at the CIA website, including excerpts from the journal, Studies in Intelligence and the Sherman Kent "occasional papers" that talk about the need to adapt to new world issues, such as transnational terrorism as opposed to the threat of Soviet major attack.Howard C. Berkowitz 16:39, 19 October 2007 (UTC)
I just found http://www.fas.org/blog/secrecy/docs/seminar.pdf, which was an NSA briefing to press members on balancing accuracy and detail. Skimpy and overly redacted, it's still interesting. Howard C. Berkowitz 16:46, 19 October 2007 (UTC)

I've seen the suggestion that it could be Military Intelligence but would disagree. Whilst many of the strategic intelligence organisations have military antecedents the development of int activity are as much diplomatic and political as they are military. UK examples are the Security Service and the Secret Intelligence Service both of which are derived from military organisations prior to WWI although under the auspices of the Home office and Foreign Office now.

Sounds like you're a newcomer to UK intelligence. What you just said would have shocked Sir Francis Walsingham, whom I don't think was especially military. :=)
Not a newcommer, by a long shot :)
What I was getting at was the industrialised intelligence machinery which we have now is predominantly derived from military origins. Those military origins drew on the methods and skills from before, but developed a much more secure funding base through the state. Wasn't it Walsingham who bankrupted himself to keep his network going?
ALR 13:42, 19 October 2007 (UTC)
I think it was. On a different note, when Patton was commanding his first armored division, he bought most of the mechanics' tools out of personal funds and from commercial sources. My armored friends tell me that one commercial wrench is still really necessary, and the gunners personally buy, to keep the 25mm gun on the Bradley working happily. Howard C. Berkowitz 16:39, 19 October 2007 (UTC)

A capping piece could reasonably cover the Collection Co-ordination and Intelligence Requirements Management (CCIRM) process, AKA the Intelligence Cycle, and put the analysis activities in context. That should allow an opportunity to discuss cross disipline issues in the round, rather than just the ELINT/ COMINT/ MASINT relationship that's being considered here. It might make it easier to deal with, as it is the relationships can be quite esoteric just within the SIGINT domain. I do see a challenge in sourcing that discussion, anything here in the UK that deals with all source is classified, although the NATO OSINT publications are UNCLAS which might help.

There's a fair bit of US material, both in the press and in discussions of "all source analyst workstations", that can give some perspective. Unfortunately, I know a fair number of unclassified things that aren't sourced anywhere, ranging from my intelligence seminar in graduate school to working as a technical advisor for a retired agency head, in his first civilian job.
The personal knowledge issue is something that I've found difficult, but now it's more a question of using that to nuance the writing, and to inform my questioning of sources.ALR 13:42, 19 October 2007 (UTC)

In terms of the structure of this article, I think there would be value in something along the lines of; overview, SIGINT principles, subordinate disciplines, collection methods and a brief overview of Sovereign collection, bi-lateral and multi-lateral collection. Detail of the various national histories and capabilities can be spun out to daughter articles. My instinct is to avoid much detail about individual national collection platforms, by their very nature the public domain source material about platforms isn't all that reliable. I'm seeing that in the Echelon article at the moment, almost everything is based on Duncan Campbell and his writing is confusing.

That makes sense. In my user page sandbox, I have a draft into which I put multinational, national, and industrial components. You may want to look at that (I assume it's accessible). There will be the mechanical problem of going back and forth to be sure all the citations in the original document show up in the daughters.
Your points are well taken about Campbell and some of the other journalists. In some cases, I don't know how "quality control" fits with the Wiki tradition of "no original research". Is it OR, for example, to point out that while Danish journalists said an agreement had been made between their service and OSS, that OSS had been out of existence for about two years before? Oh, there were various intemediate groups, like CIG, OPC, and...(the other operations unit that wasn't OPC, and whose name I don't remember). In passing, I saw some references to technical sensors, which variously had impossible geometry or, unfortunately, I knew from my personal experience didn't work that way, but I doubt there's one classified or unclassified document that had the information.
I'd like to apply some cross-checking to Campbell. In some cases, following his footnotes, he quotes Jeffrey Richelson, whom I've usually found to be very accurate. I give less weight to something from a counterculture journalist !!!
I've had a look at your draft, it goes a little further than I've been intending. Part of the problem is that Campbells' conclusions just aren't tenable for anyone with even a cursory knowledge, although he does base his assumptions on some credible material. time pressures are an issue for me at the moment, so it's taking longer than I'd anticipated.ALR 13:42, 19 October 2007 (UTC)

Some daughter articles already exist, but do need some work themselves; UKUSA is inadequate, HF/DF could be subsumed into a broader article about DFing etc. Dropping the rather extensive USian history into a daughter article would restore some proportionality. I'll concede that US OPSEC is pretty questionable, leading to a high volume of available sourcing on it, but it's not the only nation in the world engaged in the art of intelligence.

I'm not sure existing ELINT has much in it besides cut and paste. Traffic analysis has some content, and there's a direction finding page distinct from HF/DF. There's also some question about how much belongs in an intelligence article (e.g., DF) versus goniometry. Cryptanalysis vs. COMINT is yet another area. I consider DF to be applicable both to COMINT and ELINT, and I'll make an argument that traffic analysis does as well. Is cryptanalysis a common technique when dealing with frequency agile devices or radar pulses that make use of encryption for authentication/ECCM?
HF/DF, as a concept in the Battle of the Atlantic, is historically significant, but the oscilloscope techniques they used are several generations obsolete. Even so, there's a question of what makes a generation -- at a minimum, a Wullenweber type as a next generation, then the variants that consider polarization, but I'd consider the current state of the art to be time of signal arrival at precisely time-synchronized receivers. Howard C. Berkowitz 19:26, 16 October 2007 (UTC)

I've noted some discussion of BDA, I'd consider this to be an element of targetting, rather than a discipline in it's own right.

There used to be an USAF manual of targeting online, but I think it's been pulled. Especially considering extending BDA to the obscenity of "body count", Sam Adams War of Mumbers and other Vietnam-era sources are useful. The US is doing quite a bit of technical development on integrating BDA with operations, my favorite, which I am told actually works, is a variant on Tactical Tomahawk, where, slightly before impact, the missile reels out a video camera and transmitter on a fiber cable perhaps a quarter-mile long, which sends back imagery of final target selection and the explosion, before it joins its parent in death. Howard C. Berkowitz 16:39, 19 October 2007 (UTC)
Honestly, I'm not sure where it fits. Clearly, some BDA is necessary to decide if you need to hit a target again. Thinking at an aggregate level, there were widely differing estimates of the damage done to Iraqi forces in Desert Storm, with the operators on the optimistic side. This was also a problem in Vietnam, as bad as body count may be -- see Sam Adams' War of Numbers.
I'm using my own knowledge of how things are organised in the UK for this. For us the BDA is carried out by the same people, since it's essential to keeping the target lists current and meaningful for the weapon carriers. At the tactical level then it's just part of the estimate process. I think that the issue is complicated when HUMINT from the tactical level informs target list updates at the operational and strategic levels.ALR 13:42, 19 October 2007 (UTC)
For the US, the weapon carriers frequently overestimate damage, perhaps due to pressure from politically-oriented senior commanders. The more conservative CIA estimate often turn out to be more accurate. Somewhere, I saw a piece or two that discussed the motivation and symmetry of the US TENCAP (Tactical Exploitation of National Capabilities) and TIARA (Tactical Intelligence and Related Activity). Especially in Congressional research, there's some thoughtful open source, if I can just find it. There is some discussion of how SCI target photography was downgraded to a level where the attack pilots could look at pictures of what they were to hit.
Perhaps another datum here appears in the repeated citing, of Wiki and elsewhere, of a raw intelligence report about Saddam's use of CW in Iraq, which rather casually used the term "chemical warfare". That report, which was full of caveats, has been repeatedly used to justify a US consensus that white phosphorus is a chemical weapon. Howard C. Berkowitz 16:39, 19 October 2007 (UTC)

I'd consider Information Operations as an Int discipline, part of the preparation of the battlespace, encompassing soft communications as well as offensive electronic attack. Use of ECM is an Information Ops activity at the very basic, admittedly tactical, level.

You've just described some of the most difficult areas. It depends, I suppose, on how soft you make the disciplines. Subtle software attacks can go along with heavy-duty jamming, but some US sources are putting psychological operations into information operations. To me, the disciplines are too different to treat as a unit, although they both are information in the broad sense. Howard C. Berkowitz 19:26, 16 October 2007 (UTC)
Again I'm basing that on my experience of the UK. IO is predominantly about influence, rather than electronic attack. That said, I'm not sure how much public domain material is about on how we do things.ALR 13:42, 19 October 2007 (UTC)

I think there needs to be some serious consideration of source validity and reliability. Much of what's in there at the moment is unsupported, and whilst there is some I agree with there is some that I'd argue was at best very out of date, or lacking in context to the extent that what can be inferred is incorrect. Given the amount of garbage around on Int activities I think we need to be pretty rigorous about finding decent source material.

While the gods of procedure revise it every so often, there is value -- I think it's NATO as well as US -- of the marking of reports as A-1 or F-6, to report separately source reliability and the plausibility/confirmation of an actual report. Again, this is something for a general article.
I don't remember if it was you or someone else, but I've been thinking of a diagram that ties the Boyd OODA loop to the classic intelligence cycle, which I'm going to try to draw. Whether that is OR or editorial license is TBD. Howard C. Berkowitz 16:39, 19 October 2007 (UTC)
I wonder if the Military History or another Project can define something that wouldn't be considered OR, but would present contradictions among sources as a comment about source quality.Howard C. Berkowitz 19:26, 16 October 2007 (UTC)

I'm wary of rehashing a lot of Campbell in the suite of articles, I don't see him as particularly reliable and much of his evidence to the European Parliament was pretty speculative. I also think that we need to be reasonably proportionate, he provides some interesting insight into intercept issues, but that's the main value.

My impression is that whenever the EP assumed a station intercepted something specific, they had considered the geometry of the antenna, orbit, footprint, and perhaps the receiver sensitivity. In some cases, Campbell seems to infer a capability to a site not based on physical parameters such as those I mentioned, but only because a detachment from some unit elsewhere associated with ECHELON was assigned there. The latter isn't enough.

In terms of the Defence against SIGINT section I'd look at a daughter article dealing with Mitigating the SIGINT threat, since most of the defences are really mitigations, difficult to avoid your signals being intercepted, so you make using that intercept as difficult as possible.

Ever since I got stuck in a poorly ventilated SCIF with a cigar smoker, I sometimes wonder if the mitigations are worth it. :-) Howard C. Berkowitz 19:26, 16 October 2007 (UTC)
You have my sympathies :) Fortunately mine was wholly smoke free, although after a 12 hour watch it didn't need that to become unpleasant...
I was thinking more in the round, we protect our traffic with crypto and keymat, yet the crypto and keymat also need protected, through physical security, personnel security, audit, process etc. The physical and electronic protection of the equipment being related to that. There are open source books which bring all of that together and talk about cracking crypto through various routes and could be used as sources for the article.
ALR 13:42, 19 October 2007 (UTC)
Keymat is one of the least understood aspects, especially with civilian crypto. The best discussion I've seen is in Internet Cryptography by Smith. Howard C. Berkowitz 16:39, 19 October 2007 (UTC)

Nonetheless the foundations are there. thats clearly related to something about more generally mitigating the intelligence threat.

ALR 17:41, 16 October 2007 (UTC)

Geez - this whole article is disjointed, and contains either dated, incomplete, innacurate, or even unlrelated info. One example: why are 'code talkers' discussed in the section on COMINT _Voice Interception_ ...and little else (besides some apparently partial cut/paste segments from some manual or training aid? Seems to me code talker stuff more correctly would belong to _COMSEC_ (if anywhere), and in the context it IS used it should be marked as more clearly for illustrative purposs as an issue COMINT collectors would need to be able to recognize, detect, and deal with if encountered. The analysis stuff...well...you tell me?? I'd contribute, but I don't have time for it, I've been out of the field a while, and don't really want to be part of it -- especially in the state is is now. But as a former professional in the feild, I suggest complete overhaul should done with a planned, offline, restructuring and re-writing...or something. As it is now, despite some sections being 'okay', overall it's an embarassment to both Wikipedia and the SIGINT/MilInt professions IMHO. 24.218.32.186 (talk) 01:26, 6 April 2010 (UTC)