Wednesday, June 29, 2005

CANUSA anniversary?

As part of its participation in the UKUSA community, Canada is reported to have bilateral SIGINT agreements with the U.K. and the U.S. The latter is apparently called the CANUSA Agreement (although some sources say CANUS Agreement), and today may be its 56th anniversary.

CANUSA signVery little is publicly known about the CANUSA agreement. One thing we do know is that Canada-U.S. negotiations for the agreement were well underway by mid-1948. According to a U.S. Air Force memorandum dated 7 June 1948 that described a draft version of the agreement, it is modelled at least in part on the U.K.-U.S. BRUSA agreement (presumably the 1946 version) and governs Canada-U.S. co-operation on Communications Intelligence, which, for the agreement's purposes, is "understood to comprise all processes involved in the collection, production and dissemination of information derived from the communications of countries other than the U.S.A., the British Empire, and the British Commonwealth of Nations."

According to the memorandum, the agreement provides for the exchange of COMINT information "on the request of each authority to meet the requirements of the COMINT centers for assistance in the efficient discharge of their mutually agreed-upon COMINT activities and undertakings" and "on a 'need to know' basis as determined by the originating authority." It also provides for the exchange of COMINT liaison officers between Canada and the United States. [Brig. Gen. Walter Agee, USAF, Acting Deputy Director of Intelligence, "Memorandum for the Coordinator of Joint Operations: Proposed U.S.-Canadian Agreement," 7 June 1948.]

Sources disagree on when the CANUSA Agreement was finalized. Martin Rudner, one of the best informed commentators on CSE matters, says the agreement was finalized in May 1948. [Canada's Communications Security Establishment: From Cold War to globalisation, NPSIA Occasional Paper 22, 2000.] But this date seems incompatible with the June 1948 Agee memorandum. David Bercuson and Jack Granatstein put the date at some time in 1949. [D.J. Bercuson and J.L. Granatstein, The Dictionary of Canadian Military History, Oxford University Press, 1992, p. 43.] Others have suggested that the "Security Agreement between Canada and the United States of America" signed 15 September 1950 may have been the CANUSA agreement.

Letter from USCIB to CRC, 27 June 1949My own sense is that 1949 is about right. In fact, this declassified letter from the Chairman of the U.S. Communications Intelligence Board to the Chairman of the Communications Research Committee (the Canadian government interdepartmental committee then responsible for SIGINT policy), dated 29 June 1949, may be the actual document that finalized the agreement.

In any case, the Public Accounts for FY 1949-1950 show that Robert S. McLaren, Canada's first SIGINT liaison officer to the U.S., received his moving allowance and the initial installment of his representation allowance during that fiscal year (i.e., between 1 April 1949 and 31 March 1950). This also suggests that CANUSA was finalized in 1949 or very early in 1950.

Canada and the U.S. continue to post liaison officers at each other's agencies. Information about one former Senior U.S. Liaison Officer/Ottawa (SUSLO/O), Velva Klaessy, can be found on the NSA website.

[Update 30 May 2009: According to Secret Sentry, the letter above is indeed the document that finalized the agreement. The agreement was formally signed in November 1949.]

Monday, June 13, 2005

R291 on the recruiting site

The official blurb on the Reserve Communicator Research Operator (R291) trade is now up on the Canadian Forces Recruiting website. "Training takes approx 128 days spread over two summers."
Communication Reserve Communicator Research Operators work approximately one to two evenings a week and up to two weekends a month at their local unit. Once trained there are full time opportunities for employment with the Canadian Forces, either in Canada or deployed overseas, such as with a United Nations operation.

Sunday, June 12, 2005

In the news: CSE's expansion

Jim Bronskill of the Canadian Press reports on CSE's ongoing expansion: "Canada's secret eavesdropping agency is undergoing its biggest expansion in decades as it takes on a greater role in the fight against terrorism" (CP story, Globe and Mail pick up, Winnipeg Sun pick up...).

Its biggest expansion ever, if you want to get pedantic about it, and its fastest. CSE has had three major growth episodes during its 59-year history: an increase of about 600 over the 15 years between the agency's inception in 1946 and 1961; an increase of about 300 over the 10 years between 1980 and 1990; and the current increase, which is projected to total about 750 personnel over the six years between 2001 and 2007. When it reaches its final target level of 1,650 "full-time equivalents", CSE will have nearly twice as many employees as it had the year the Berlin Wall came down (1989 average: 835) and more than three times the average number it had over the Cold War as a whole (525).

Anyway, a good article by one of the only reporters (or the only reporter?) who regularly covers the Canadian intelligence beat. Read it while you can: newspapers don't tend to leave their stuff available online very long anymore.

Thursday, June 09, 2005

This date in history: XU began operations

On this date in history, 9 June 1941, Canada's first cryptanalytic agency, the Examination Unit, began operations. Housed in rooms 202 and 203 of the National Research Council Annex on Montreal Road in Ottawa, the XU had an initial staff of nine: Herbert O. Yardley, his assistant Edna Ramsaier, mathematician Dr. Gilbert de Beauregard Robinson (no relation), RCMP Constable Robert McLaren, Dr. Douglas Cameron, Richard Rudey (or Ruddy?) from the NRC, Vern Gavel from the Royal Canadian Corps of Signals, and two typists.

Gilbert deB Robinson Yardley and Ramsaier were gone within six months' time. Robinson (shown at right) later ended up running the XU but returned to his position in the mathematics department at the University of Toronto at the end of the war. Some of the other members of the original staff ended up making a career in signals intelligence. Robert McLaren stayed on to become one of CBNRC's initial employees and later served as CBNRC's first liaison officer at AFSA, forerunner to NSA. Vern Gavel also stayed on with CBNRC after the war, eventually retiring in 1972.

Previous posts about the XU here and here.

Tuesday, June 07, 2005

Coulter speaks again

More tidbits from the soon-to-be-departed Chief of CSE, Keith Coulter. On 4 May, Coulter testified to the Subcommittee on Public Safety and National Security of the House of Commons' Standing Committee on Justice, Human Rights, Public Safety and Emergency Preparedness, which is examining the operations of the Anti-Terrorism Act of 2001. He previously testified to the Special Senate Committee on the Anti-Terrorism Act on the same subject (blogged about here).

Coulter, accompanied by Deputy Chief Corporate Services Barb Gibbons, Director General Policy and Communications John Ossowski, and Director and General Counsel Legal Services David Akman, provided extensive new details on CSE's budget and staffing.
  • CSE's budget prior to September 11th, 2001 was $140 million per year.
  • The December 2001 federal budget approved a 25% increase in CSE's budget, to take effect over several years.
  • The March 2004 budget approved an additional 25% increase in CSE's budget, to take effect by fiscal year 2007-08, for a cumulative increase of 57% , at which time CSE will have a budget of $220 million per year.
  • The 2001 budget also approved a 35% increase in CSE's staffing, from about 950-1,000 full-time equivalents (FTEs) to about 1,300.
  • The 2004 budget approved a further 25% increase in CSE's staffing, i.e., another 350 FTEs, for a total staff of 1,650 in FY 2007-08.
The 1,650 figure is more than 100 FTEs higher than the previous publicly acknowledged figure of 1,546. (CSE's current staff size is 1,381.)

The testimony also provides some general information on CSE's intelligence-gathering priorities. Coulter's written submission (PDF file) notes that
CSE has greatly increased its focus on security issues. CSE now devotes the majority of its foreign intelligence efforts to gathering and reporting intelligence on issues such as terrorism, proliferation and cyber threats. CSE also supports deployed Canadian Forces operations abroad.
His spoken testimony, however, provides the more precise claim that
Right now, if you look at the reporting as one metric on this, over 75% of our business is in the security domain, and that's a little broader than terrorism. That's proliferation as well. It is counter-intelligence as well. It's cyber-threats as well. And these days it is hugely a support to military operations ... because we have troops deployed abroad and we're very involved in helping to intercept communications so they can paint the picture of what the local threats are to them.
The testimony contains lots of other interesting comments as well.

Thursday, June 02, 2005

The FRD-10: An endangered species

FRD-10 array

In the early 1960s the U.S. Naval Security Group began deploying a network of large high-frequency direction-finding (HF-DF) circularly disposed antenna arrays, the AN/FRD-10s, to detect, monitor, and plot the location of Soviet submarines and other radio emitters in the Atlantic and Pacific oceans.

Sometimes dubbed Elephant Cages or Dinosaur Cages, the FRD-10 arrays were enormous structures. In the centre of each array was a two-storey operations building, about 40 metres square, where the station personnel worked. Surrounding this building were two concentric rings of HF antennas, one for shorter HF wavelengths, containing 120 sleeve monopoles, and one for longer HF wavelengths, containing 40 folded dipoles. The shorter wavelength ring was about 260 metres in diameter and the longer wavelength ring was about 230 metres in diameter. Inside each ring there was also a large wire screen, supported by 80 towers, which was designed to prevent HF signals from crossing the array and interfering with its operations. The inner screen, corresponding to the longer HF wavelengths, was roughly 36 metres high. A horizontal ground screen about 390 metres in diameter surrounded the entire site. (Aerial views of an FRD-10 array here.)

Fourteen of the huge arrays were eventually deployed by the NSG (not counting two built at Sugar Grove, WV, for communications rather than intelligence-gathering):
  • Adak, Alaska
  • Edzell, Scotland
  • Galeta Island, Panama
  • Guam
  • Hanza, Okinawa
  • Homestead, Florida
  • Imperial Beach, California
  • Marietta, Washington
  • Northwest, Virginia
  • Rota, Spain
  • Sebana Seca, Puerto Rico
  • Skaggs Island, California
  • Wahiawa, Hawaii
  • Winter Harbor, Maine
Another two were built by the Canadian Forces Supplementary Radio System: one at Gander, Newfoundland and one at Masset, British Columbia (both built in 1970-71).

The FRD-10 arrays became the backbone of the BULLSEYE net, the Atlantic and Pacific HF-DF nets. They were supplemented by a number of smaller, simpler CDAAs known as Pushers, including a Canadian Pusher in Bermuda. (Canada also has Pushers deployed at Leitrim and Alert.)

The FRD-10s offered four major improvements over their predecessors, the GRD-6 in NSG service and the GRD-501 in Canadian service:
  1. transmissions could be recorded for immediate or subsequent DF-ing;
  2. bearings were four times as accurate;
  3. antenna gain was about four times higher than previous systems; and
  4. the system had the ability to select wanted signals and reject interfering signals or noise.
As noted in the Supplementary Radio Activities Consolidation Plan (30 May 1966), the improvement expected as a result of deploying the FRD-10s was "a combination of more accurate and reliable fixes, producing reduced search areas in ocean areas of prime responsibility so fresh in time as to enable maritime commanders to deploy their forces more economically and with much greater prospect of making contact with the target than is now the case."

The following maps show the locations of FRD-10 (black circles with pink dots) and Pusher arrays (empty black circles) during the system's heyday.

Click to see full-sized image

Click to see full-sized image

In the mid-1990s, however, the NSG began to close down its FRD-10 arrays. The demise of their Soviet targets, a desire to refocus collection efforts and cut costs, and, presumably, a decision to rely on alternative ocean surveillance technologies has led to the near-extinction of the FRD-10. Canada's two arrays are the only ones left in service. The others have since been dismantled. (Al Grobmeier has written more on the fate of the NSG FRD-10s.)

The Canadian FRD-10s themselves were converted in 1997 so they could be remotely operated from CFS Leitrim. Presumably they are primarily used for HF intercept operations now, although DF remains as a secondary capability. The Masset and Gander arrays no longer have other FRD-10s to work with, but they can still work with the Canadian Pusher arrays at Leitrim and Alert, and presumably with other HF-DF sites still operated by UKUSA allies and other partners. (This leaked NSA document (see p. 3) confirms that the agency continues to operate the "world-wide CROSSHAIR HFDF geolocation service", and that 2nd Parties and other countries participate.)

The first step in this direction may have been Project Polo (G1777), which was established in the late 1980s to "modernise the CFSRS High Frequency/Direction Finding (HF/DF) system at CFS Masset and Gander, and to equip CFS Alert for netted DF Operations."

Did Canada see the demise of the FRD-10 network coming? It seems unlikely that the authors of Challenge and Commitment, the 1987 Defence White Paper (précis: "The Soviet Threat will go on forever"), expected the UKUSA allies' main ocean surveillance networks to be shut down within a decade, so assuming the two sites remain useful and we're not just stuck with a couple of White Elephant Cages, maybe we got lucky.

Update 25 August 2005: Corrected to include the FRD-10 built at Marietta, Washington, and update info on the fate of the NSG FRD-10s. The FRD-10 at Marietta was dismantled in 1972, possibly as a result of the Masset array's entry into service. [A comment dated 12 September 2015 (see below) states that the Marietta array had continuing problems with inaccurate bearings due to the presence of nearby aluminum smelters, so it was probably pointless to keep the site in service once Masset was complete.]

Update 14 December 2007: As noted in the comments, the NSG detachments at US Army/Air Force FLR-9 CDAA sites also participated in the BULLSEYE net, as did some older NSG sites that continued operating the GRD-6 system (and older Canadian sites operating the GRD-501 system) for a number of years.

Update 24 June 2009: Information on the roughly equivalent Soviet Krug HF-DF network here.

Update 9 March 2015: Updated Pacific HF-DF map to include the Panama FRD-10, as suggested in comments.

Update 13 September 2015: Added some information on problems at Marietta, as suggested in comments, and updated a few other dated sections of the text.

Wednesday, June 01, 2005

More on the R291 trade

An article in the 25 May 2005 edition of Maple Leaf provides more information about the new Reserve Communicator Research occupation.
The new occupation will both complement and augment (where necessary) Regular Force operations in three broad areas: tactical electronic warfare, strategic signals intelligence and linguistic support. ...
Currently there are over 100 members of 772 EW Squadron in Kingston, Ont. whose primary mission is to provide augmentation on operations to their Regular Force counterpart, 2 EW Squadron. The Res Comm Rsch occupation has been developed with the Army in mind to produce a better communicator in-theatre by broadening their knowledge of SIGINT. Personnel who understand strategic SIGINT capabilities and capacities while working the EW mission at the tactical and operational level will provide commanders with better information faster. ...
“A Reserve Communications occupation will pay significant dividends by helping sustain current and future CF operations, thus reducing the toll of high operational tempo,” said Col Dave Neasmith, commander of the Canadian Forces Information Operation Group (CFIOG). The operational demands on Communicator Research personnel have steadily grown. With only 600 Regular Force personnel in the occupation, deploying personnel to all theatres and to sea has placed a severe strain on its members who must also maintain 24/7 support to all theatres via CFIOG units here in Canada.
Only 600 Reg Force 291ers? The PML in 1998 was 665, so this new number probably shouldn't come as a surprise, but CFIOG's establishment was recently reported to be 900. Granted, that number would also include other trades, and it was acknowledged that the Group is not at full strength. But maybe the Reg Force number is also soon to go up?

Retraining of 772 EW Squadron personnel will begin in 2006. Other Communication Reserve units will also be involved:
In addition to the unit in Kingston, a troop will be established in Ottawa at 763 Comm Regiment to provide operational support to CFIOG. This troop will supply Res Comm Rsch operators on a full- and part-time basis to augment their Regular Force counterparts at CFS Leitrim, as well as deploy into theatres of operations. To address the need for personnel with specialized language skills, five test sites have been initially chosen to recruit native speakers—Victoria, Winnipeg, Toronto, Montréal and Halifax. By reallocating existing Comm Res positions from across Canada, the goal is to have the complete Res Comm Rsch occupation fully staffed by 2009-2010.
Still no word on the CFIOG Support Dets proposal, however, unless the plan to recruit specialized language speakers is part of or has now replaced that idea.