I have a rough list of topics for future articles, a scratchpad of two-word
ideas that I sometimes struggle to interpret. Some items have been on that list
for years now. Sometimes, ideas languish because I'm not really interested in
them enough to devote the time. Others have the opposite problem: chapters of
communications history with which I'm so fascinated that I can't decide where
to start and end. They seem almost too big to take on. One of these stories
starts in another vast frontier: northeastern Canada.
It was a time, rather unlike our own, of relative unity between Canada and the
United States. Both countries had spent the later part of World War II planning
around the possibility of an Axis attack on North America, and a ragtag set of
radar stations had been built to detect inbound bombers. The US had built a
series of stations along the border, and the Canadians had built a few north of
Ontario and Quebec to extend coverage north of those population centers. Then
the war ended and, as with so many WWII projects, construction stopped. Just a
few years later, the USSR demonstrated a nuclear weapon and the Cold War was
on. As with so many WWII projects, freshly anxious planners declared the
post-war over and blew the dust off of North American air defense plans. In
1950, US and Canadian defense leaders developed a new plan to consolidate and
improve the scattershot radar early warning plan.
This agreement would become the Pinetree Line, the first of three
trans-Canadian radar fences jointly constructed and operated by the two
nations. For the duration of the Cold War, and even to the present day, these
radar installations formed the backbone of North American early warning and the
locus of extensive military cooperation. The joint defense agreement between
the US and Canada, solidified by the Manhattan Project's dependence on Canadian
nuclear industry, grew into the 1968 establishment of the North American Air
Defense Command (NORAD) as a binational joint military organization.
This joint effort had to rise to many challenges. Radar had earned its place as
a revolutionary military technology during the Second World War, but despite
the many radar systems that had been fielded, engineer's theoretical
understanding of radar and RF propagation were pretty weak. I have
written
here
before
about over-the-horizon radar, the pursuit of which significantly improved our
scientific understanding of radio propagation in the atmosphere... often by
experiment, rather than model. A similar progression in RF physics would also
benefit radar early warning in another way: communications.
One of the bigger problems with the Pinetree Line plan was the remote location
of the stations. You might find that surprising; the later Mid-Canada and DEW
lines were much further north and more remote. The Pinetree Line already
involved stations in the far reaches of the maritime provinces, though, and to
provide suitable warning to Quebec and the Great Lakes region stations were
built well north of the population centers. Construction and operations would
rely on aviation, but an important part of an early warning system is the
ability to deliver the warning. Besides, ground-controlled interception had
become the main doctrine in air defense, and it required not just an alert but
real-time updates from radar stations for the most effective response. Each
site on the Pinetree Line would require a reliable real-time communications
capability, and as the sites were built in the 1950s, some were a very long
distance from telephone lines.
Canada had only gained a transcontinental telephone line in 1932, seventeen
years behind the United States (which by then had three different
transcontinental routes and a fourth in progress), a delay owing mostly to the
formidable obstacle of the Canadian Rockies. The leaders in Canadian
long-distance communications were Bell Canada and the two railways (Canadian
Pacific and Canadian National), and in many cases contracts had been let to
these companies to extend telephone service to radar stations. The service was
very expensive, though, and the construction of telephone cables in the
maritimes was effectively ruled out due to the huge distances involved and
uncertainty around the technical feasibility of underwater cables to
Newfoundland due to the difficult conditions and extreme tides in the Gulf of
St. Lawrence.
The RCAF had faced a similar problem when constructing its piecemeal radar
stations in Ontario and Quebec in the 1940s, and had addressed them by applying
the nascent technology of point-to-point microwave relays. This system, called
ADCOM, was built and owned by RCAF to stretch 1,400 miles between a series of
radar stations and other military installations. It worked, but the
construction project had run far over budget (and major upgrades performed soon
after blew the budget even further), and the Canadian telecom industry had
vocally opposed it on the principle that purpose-built military communications
systems took government investment away from public telephone infrastructure
that could also serve non-military needs.
These pros and cons of ADCOM must have weighed on Pinetree Line planners when
they chose to build a system directly based on ADCOM, but to contract its
construction and operation to Bell Canada [1]. This was, it turned out, the
sort of compromise that made no one happy: the Canadian military's
communications research establishment was reluctant to cede its technology to
Bell Canada, while Bell Canada objected to deploying the military's system
rather than one of the commercial technologies then in use across the Bell
System.
The distinct lack of enthusiasm on the part of both parties involved was a bad
omen for the future of this Pinetree Line communications system, but as it
would happen, the whole plan was overcome by events. One of the great struggles
of large communications projects in that era, and even today, is the rapid rate
of technological progress. One of ADCOM's faults was that the immense progress
Bell Labs and Western Electric made in microwave equipment during the late '40s
meant that it was obsolete as soon as it went into service. This mistake would
not be repeated, as ADCOM's maritimes successor was obsoleted before it even
broke ground. A promising new radio technology offered a much lower cost
solution to these long, remote spans.
At the onset of the Second World War, the accepted theory of radio propagation
held that HF signals could pass the horizon via ground wave propagation,
curving to follow the surface of the Earth, while VHF and UHF signals could
not. This meant that the higher-frequency bands, where wideband signals were
feasible, were limited to line-of-sight or at least near-line-of-sight links...
not more than 50 miles with ideal terrain, often less. We can forgive the
misconception, because this still holds true today, as a rule of thumb. The
catch is in the exceptions, the nuances, that during the war were already
becoming a headache to RF engineers.
First, military radar operators observed mysterious contacts well beyond the
theoretical line-of-sight range of their VHF radar sets. These might have been
dismissed as faults in the equipment (or the operator), but reports stacked up
as more long-range radar systems were fielded. After the war, relaxed
restrictions and a booming economy allowed radio to proliferate. UHF television
stations, separated by hundreds of miles, unexpectedly interfered with each
other. AT&T, well into deployment of a transcontinental microwave network, had
to adjust its frequency planning after it was found that microwave stations
sometimes received interfering signals from other stations in the chain...
stations well over the horizon.
This was the accidental discovery of tropospheric scattering.
The Earth's atmosphere is divided into five layers. We live in the troposhere,
the lowest and thinnest of the layers, above which lies the stratosphere.
Roughly speaking, the difference between these layers is that the troposphere
becomes colder with height (due to increasing distance from the warm surface),
while the stratosphere becomes warmer with height (due to decreasing shielding
from the sun) [2]. In between is a local minimum of temperature, called the
tropopause.
The density gradients around the tropopause create a mirror effect, like the
reflections you see when looking at an air-water boundary. The extensive
turbulence and, well, weather present in the troposhere also refract signals on
their way up and down, making the true course of radio signals reflecting off
of the tropopause difficult to predict or analyze. Because of this turbulence,
the effect has come to be known as scattering: radio signals sent upwards,
towards the troposphere, will be scattered back downwards across a wide area.
This effect is noticeable only at high frequencies, so it remained unknown
until the widespread use of UHF and microwave, and was still only partially
understood in the early 1950s.
The locii of radar technology at the time were Bell Laboratories and the MIT
Lincoln Laboratory, and they both studied this effect for possible
applications. Presaging one of the repeated problems of early warning radar
systems, by the time Pinetree Line construction began in 1951 the Lincoln
Laboratory was already writing proposals for systems that would obsolete it.
In fact, construction would begin on both of the Pinetree Line's northern
replacements before the Pinetree Line itself was completed. Between rapid
technological development and military planners in a sort of panic mode, the
early 1950s were a very chaotic time. Underscoring the ever-changing nature of
early warning was the timeline of Pinetree Line communications: as the Pinetree
Line microwave network was in planning, the Lincoln Laboratory was
experimenting with troposcatter communications. By the time the first stations
in Newfoundland completed construction, Bell Laboratories had developed an
experimental troposcatter communications system.
This new means of long-range communications would not be ready in time for the
first Pinetree Line stations, so parts of the original ADCOM-based microwave
network would have to be built. Still, troposcatter promised to complete the
rest of the network at significantly reduced cost. The US Air Force, wary of
ADCOM's high costs and more detached from Canadian military politics,
aggressively lobbied for the adoption of troposcatter communications for the
longest and most challenging Pinetree Line links.
Bell Laboratories, long a close collaborator with the Air Force, was well aware
of troposcatter's potential for early warning radar. Bell Canada and Bell
Laboratories agreed to evaluate the system under field conditions, and in 1952
experimental sites were installed in Newfoundland. These tests found reliable
performance over 150 miles, far longer than achievable by microwave
and---rather conveniently---about the distance between Pinetree Line radar
stations. These results suggested that the Pinetree Line could go without an
expensive communications network in the traditional sense, instead using
troposcatter to link the radar stations directly to each other.
Consider a comparison laid out by the Air Force: one of the most complex
communications requirements for the Pinetree Line was a string of stations
running not east-west like the "main" line, but north-south from St. John's,
Newfoundland to Frobisher Bay, Nunavut. These stations were critical for
detection of Soviet bombers approaching over the pole from the northeast,
otherwise a difficult gap in radar coverage until the introduction of radar
sites in Greenland. But the stations covered a span of over 1,000 miles, most
of it in formidably rugged and remote arctic coastal terrain. The proposed
microwave system would require 50 relay stations, almost all of which would be
completely new construction. Each relay's construction would have to be
preceded by the construction of a harbor or airfield for access, and then
establishment of a power plant, to say nothing of the ongoing logistics of
transporting fuel and personnel for maintenance. The proposed troposcatter
system, on the other hand, required only ten relays. All ten would be colocated
with radar stations, and could share infrastructure and logistical
considerations.
Despite the clear advantages of troposcatter and its selection by the USAF, the
Canadian establishment remained skeptical. One cannot entirely blame them,
considering that troposcatter communications had only just been demonstrated in
the last year. Still, the USAF was footing most of the bill for the overall
system (and paying entirely for the communications aspect, depending on how you
break down the accounting) and had considerable sway. In 1954, well into
construction of the radar stations (several had already been commissioned), the
Bell Canada contract for communications was amended to add troposcatter relay
in addition to the original microwave scheme. Despite the weaselly contracting,
the writing was on the wall and progress on microwave relay stations almost
stopped. By the latter part of 1954, the microwave network was abandoned
entirely. Bell Canada moved at incredible speed to complete the world's first
troposcatter long-distance route, code named Pole Vault.
One of the major downsides of troposcatter communications is its inefficiency.
Only a very small portion of the RF energy reaching the tropopause is
reflected, and of that, only a small portion is reflected in the right
direction. Path loss from transmitter to receiver for long links is over -200
dB, compared to say -130 dB for a microwave link. That difference looks smaller
than it is; dB is a logarithmic comparison and the decrease from -130 dB to
-200 dB is a factor of ten million.
The solution is to go big. Pole Vault's antennas were manufactured as a rush
order by D. S. Kennedy Co. of Massachusetts. 36 were required, generally four
per site for transmit and receive in each direction. Each antenna was a 60'
aluminum parabolic dish held up on edge by truss legs. Because of the extreme
weather at the coastal radar sites, the antennas were specified to operate in a
120 knot wind---or a 100 knot wind with an inch of ice buildup. These were
operating requirements, so the antenna had not only to survive these winds, but
to keep flexing and movements small enough to not adversely impact performance.
The design of the antennas was not trivial; even after analysis by both Kennedy
Co. and Bell Canada, after installation some of the rear struts supporting the
antennas buckled. All high-wind locations received redesigned struts.
To drive the antennas, Radio Engineering Laboratories of Long Island furnished
radio sets with 10 kW of transmit power. Both D. S. Kennedy and Radio
Engineering Laboratories were established companies, especially for military
systems, but were still small compared to Bell System juggernauts like Western
Electric and Northern Electric. They had built the equipment for the
experimental sites, though, and the timeline for construction of Pole Vault was
so short that planners did not feel there was time to contract larger
manufacturers. This turn of events made Kennedy Co. and REL the leading experts
in troposcatter equipment, which became their key business in the following
decade.
The target of the contract, signed in January of 1954, was to have Pole Vault
operational by the end of that same year. Winter conditions, and indeed spring
and fall conditions, are not conducive to construction on the arctic coast.
All of the equipment for Pole Vault had to be manufactured in the first half of
the year, and as weather improved and ice cleared in the mid-summer, everything
was shipped north and installation work began. Both militaries had turned down
involvement in the complex and time-consuming logistics of the project, so
Bell Canada chartered ships and aircraft and managed an incredibly complex
schedule. To deliver equipment to sites as early as possible, icebreaker CCGS
D'Iberville was chartered. C-119 and DC3 aircraft served alongside numerous
small boats and airplanes.
All told, it took about seven months to manufacture and deliver equipment to
the Pole Vault sites, and six months to complete construction. Construction
workers, representing four or five different contractors at each site and
reaching about 120 workers to a site during peak activity, had to live in
construction camps that could still be located miles from the station. Grounded
ships, fires, frostbite, and of course poor morale lead to complications and
delays. At one site, Saglek, project engineers recorded a full 24-hour day with
winds continuously above 75 miles per hour, and then weeks later, a gust of 135
mph was observed. Repairs had to be made to the antennas and buildings before
they were even completed.
In a remarkable feat of engineering and construction, the Pole Vault system was
completed and commissioned more or less on schedule: amended into the contract
in January of 1954, commissioning tests of the six initial stations were
successfully completed February of 1955. Four additional stations were built to
complete the chain, and Pole Vault was declared fully operational December of
1956 at a cost of $24.6 million (about $290 million today).
Pole Vault operated at various frequencies between 650 and 800 MHz, the wide
range allowing for minimal frequency reuse---interference was fairly severe,
since each station's signal scattered and could be received by stations further
down the line in ideal (or as the case may be, less than ideal) conditions.
Frequency division multiplexing equipment, produced by Northern Electric
(Nortel) based on microwave carrier systems, offered up to 36 analog voice
circuits. The carrier systems were modular, and some links initially supported
only 12 circuits, while later operational requirements lead to an upgrade to 70
circuits.
Over the following decades, the North Atlantic remained a critical challenge
for North American air defense. It also became the primary communications
barrier between the US and Canada and European NATO allies. Because Pole Vault
provided connections across such a difficult span, several later military
communications systems relied on Pole Vault as a backhaul connection.
An inventory of the Saglek site, typical of the system, gives an idea of the
scope of each of the nine primary stations. This is taken from "Special
Contract," a history by former Bell Canada engineer A. G. Lester:
(1) Four parabolic antennas, 60 feet in diameter, each mounted on seven mass
concrete footings.
(2) An equipment building 62 by 32 feet to house electronic equipment, plus a
small (10 by 10 feet) diversity building.
(3) A diesel building 54 by 36 feet, to house three 125 KVA (kilovolt amperes)
diesel driven generators.
(4) Two 2500 gallon fuel storage tanks.
(5) Raceways to carry waveguide and cables.
(6) Enclosed corridors interconnecting buildings, total length in this case
520 feet.
Since the Pole Vault stations were colocated with radar facilities, barracks
and other support facilities for the crews were already provided for. Of
course, you can imagine that the overall construction effort at each site was
much larger, including the radar systems as well as cantonment for personnel.
Pole Vault would become a key communications system in the maritime provinces,
remaining in service until 1975. Its reliable performance in such a challenging
environment was a powerful proof of concept for troposcatter, a communications
technique first imagined only a handful of years earlier. Even as Pole Vault
reached its full operating capability in late 1956, other troposcatter systems
were under construction. Much the same, and not unrelated, other radar early
warning systems were under construction as well.
The Pinetree Line, for all of its historical interest and its many firsts,
ended as a footnote in the history of North American air defense. More
sophisticated radar fences were already under design by the time Pinetree Line
construction started, leaving some Pinetree stations to operate for just four
years. It is a testament to Pole Vault that it outlived much of the radar
system it was designed to support, becoming an integral part of not one, or
even two, but at least three later radar early warning programs. Moreover,
Pole Vault became a template for troposcatter systems elsewhere in Canada, in
Europe, and in the United States. But we'll have to talk about those later.
[1] Alexander Graham Bell was Scottish-Canadian-American, and lived for some
time in rural Ontario and later Montreal. As a result, Bell Canada is barely
younger than its counterpart in the United States and the early history of the
two is more one of parallel development than the establishment of a foreign
subsidiary. Bell's personal habit of traveling back and forth between Montreal
and Boston makes the early interplay of the two companies a bit confusing. In
1955, the TAT-1 telephone cable would conquer the Atlantic ocean to link the US
to Scotland via Canada, incidentally making a charming gesture to Bell's
personal journey.
[2] If you have studied weather a bit, you might recognize these as positive
and negative lapse rates. The positive lapse rate in the troposphere is a
major driver in the various phenomenon we call "weather," and the tropopause
forms a natural boundary that keeps most weather within the troposphere.
Commercial airliners fly in the lower part of the stratosphere, putting them
above most (but not all) weather.
Log in
