Code Warriors
Page 30
The first electronic surveillance satellite, GRAB, was launched in June 1960 in an over-the-pole orbit to collect signals from key Soviet air defense radars.
Flying in a polar orbit just five hundred miles high, GRAB swept a continually shifting swath of territory as the earth turned below during the hour and a half it took to make each pass; at that low altitude it flew through the cone of microwave signals coming from a targeted Soviet radar installation for only a minute or less at a time. A tape recorder collected the signals and then played them back as the satellite passed over Hawaii. The first ground stations were literally no more than mobile two-man huts; a hand wheel affixed to a metal shaft running down from the ceiling between the two operator stations permitted the antenna on the roof to be steered in the right direction.31
The first GRAB satellite had a life of only three months, but from this modest beginning a new age of SIGINT collection had begun that would eventually make most of NSA’s traditional ground intercept stations, as well as the extremely dangerous ferret missions, obsolete. Satellites afforded unimpeded line of sight, at least when they happened to be in the right spot overhead, and were especially well suited to pick up radar and other microwave signals, which (unlike HF signals, which are deflected by the ionosphere) travel relatively unimpeded through the atmosphere and remain readable even thousands of miles into space.
Getting to that promising new age, though, required traversing not only a host of technological challenges but the usual rough bureaucratic road crowded with other agencies all working at cross-purposes. The Air Force was racing ahead with its own program of ELINT space probes designed to piggyback, on a space-available basis, with its Samos photoreconnaissance satellites intended to pick up where the now-canceled U-2 overflights had left off mapping Soviet ICBM sites. The F-series ELINT probes would each be able to cover the entire territory of the Soviet bloc about every five days. Three F-1 satellites were planned to be launched along with Eseries photo satellites in 1960 and 1961, but only one made it into orbit.32
Overall responsibility for ELINT meanwhile remained a bureaucratic free-for-all. Halfhearted attempts to give NSA the job had been largely shrugged off by the Navy and Air Force, which since 1952 had been operating their own joint ELINT collection and analysis program. With LeMay’s single-minded focus on SAC’s ability to “execute the SIOP”—to deliver his “Sunday punch” of nuclear bombs in an all-out attack on the Soviet Union, as prescribed in the Single Integrated Operations Plan—the Air Force commander had insisted that as far as he was concerned the only purpose of ELINT collection was to locate the targets for his bombers and missiles and the Soviet air defenses that stood in the way of their carrying out the mission. In laying out plans for the Samos program, the Air Force gave top priority to collecting radar information on the antiballistic-missile system the Soviets were developing to intercept American ICBMs. Next on the list were conventional air defense radars in remote locations: Gage and Token, which the GRAB satellite was sent to map and probe, were search and surveillance radars associated with the SA-1 surface-to-air missile batteries, and their deployment in the north of the Soviet Union along the polar route that U.S. bombers would fly to carry out a nuclear strike placed them out of reach of the border-hugging ferret planes and ground-based intercept stations. At the very bottom of the Air Force’s intelligence requirements for Samos, at “Priority IV,” was “COMINT—to the extent development proves feasible.” The military men were concerned about radars, not chitchat.33
Meanwhile CIA was staking out for itself the lead role in running the entire satellite program. CIA already had its own ELINT office, and by virtue of having been in charge of the U-2 program had built up a large photointerpretation staff, housed in the former Steuart Motor Company Building at 6th and K Streets in downtown Washington, which became the obvious choice to handle photographs taken by the new satellites as well. Just before leaving office in January 1961, Eisenhower gave CIA overall supervision of overhead reconnaissance; under a special access program code-named Talent-Keyhole, CIA controlled who was permitted to have any information about the program or its products, and the agency denied NSA all but a handful of the limited number of T-K clearances. Although NSA worked out a separate deal with the Air Force for sharing ELINT data, it still was shut out from decisions on the design or tasking of any of the signals-collecting satellites.34
With his zeal for organizational efficiency and cost control honed at Harvard Business School and the Ford Motor Company, the new defense secretary, Robert McNamara, took office determined to impose some order on the sprawling defense establishment, and he moved quickly to centralize the satellite programs in a newly created Defense Department organization, the National Reconnaissance Office, that brought together the separate CIA, Air Force, and Navy projects. NSA’s new director, Vice Admiral Laurence Hugh Frost, protested that NSA had the authority under the National Security Council’s directives to develop its own signals intelligence satellites; McNamara’s staff brushed off Frost’s objections, pointing out that the secretary of defense was the “executive agent” for carrying out NSC decisions and he had all the authority he needed to lay down the law and stop the squabbling. McNamara told the admiral his decision was final: NSA would receive the product from the satellite intercepts, but NRO would run the program.35
Frost, in office only a year and a half, was not a brisk executive in the mold of McNamara’s “whiz kids.” In meetings with senior Pentagon and White House officials he habitually spoke in so low a voice, literally mumbling his words, that no one could understand what he was saying. Although an experienced naval intelligence officer, he had little experience with civilians. Bewildered by the large staff at NSA that did not always follow orders as he was accustomed to having them followed in his thirty-five years in the Navy, he surrounded himself with fellow Navy men, which further isolated him from the agency’s rank and file, who considered him ineffectual, uncommunicative, and given to occasional unpredictable outbursts in which he vented his frustrations on subordinates, usually picking the nearest available target, deserved or not. He once humiliatingly dressed down Frank Raven, a top NSA civilian cryptanalyst, doing the whole “finger-on-the-chest bit,” as one embarrassed witness to the scene described it.
When Frost threatened to go to the White House to have McNamara’s decision on the signals satellites overturned, the impatient defense secretary seized the opportunity to fire him; he was gone by July 1962. Frost ended his naval career with an ignominious tour commanding the Potomac River Naval Command.36
Yet another spoil fought over in the unending interagency wars for control of new types of signals intelligence and new methods of interception was the telemetry data radioed back to earth from Soviet missiles and space missions. These mostly VHF-band signals posed a special collection challenge since they required line of sight to a missile’s path as it rose from Soviet test ranges and launching sites in Kazakhstan, particularly the Baikonur Cosmodrome and the antiballistic-missile test facility at Sary Shagan. The data was encoded using what was called pulse position modulation, which represented information such as acceleration, fuel pressure, and temperature by the time interval between a steadily repeated reference pulse and another pulse that stood for each variable. NSA intercept stations recorded the data on special one-inch-wide magnetic tape on twelve-inch reels. “We’d get them from field military posts in strange places,” recalled John O’Hara, who worked on Soviet telemetry at NSA in the early 1960s, “and a lot of them came in in very bad shape.” Although the Soviets sometimes tried to hide the signals by using frequencies very close to TV broadcast channels, the data itself was unencrypted. U.S. posts in northern Turkey, Peshawar in Pakistan, the northern island of Hokkaido in Japan, and central Norway were the prime ground-intercept locations. Back at Fort Meade, the tapes of the intercepted radio pulses would be transcribed by chart recorders as traces on four-hundred-foot-long rolls of emulsion paper for analysis.37
Just as
with ELINT, CIA and the Air Force also each laid claim to telemetry intelligence. Satellites were tailor-made for this mission, giving direct coverage of even the remotest locations; a satellite in geostationary orbit, about twenty-two thousand miles above the earth, could even remain permanently “parked” over Baikonur and Sary Shagan, orbiting at the same rate the earth turns below to keep a constant watch on the missile sites. CIA proceeded to develop a series of signals intelligence satellites code-named Spook Bird and later Rhyolite to do just that—again without clearing anyone at NSA to know about the project.
NSA had meanwhile been developing its own plans for a communications intercept satellite in geostationary orbit over the Soviet Union and China that could pick up telephone signals from microwave repeater towers that escaped into space due to the curvature of the earth as they were beamed from one link to another along the relay chain. Even at VHF and UHF frequencies huge high-gain antennas were required to capture a detectable signal at such extreme distances, and the satellites presented an unprecedented feat of space engineering involving unfurling a seventy-five-foot-diameter umbrella-like array. The Defense Department tried to kill the program, but at the end of 1965 the rival spy agencies arranged an uneasy truce; CIA agreed to include an NSA COMINT package within their Rhyolite mission and to establish a joint processing center, with NSA supplying half of the telemetry staff and all of the COMINT staff.38 The cost of maintaining this overhead presence would grow from tens of millions of dollars per satellite in the 1960s to a billion dollars apiece by the 1990s.
As the race to the moon increasingly began to dominate the U.S.-Soviet competition, NSA spent tens of millions of additional dollars setting up deep-space-tracking stations with huge parabolic dish antennas, spaced out along the equator at three locations (Asmara, Ethiopia, was one), specifically to grab data from Soviet lunar and Mars probes and rush to NASA the findings, which were expected to be “of great value” to the U.S. Apollo manned lunar landing program.39 If space was where the money was, then the U.S. spy agencies were going to be there, too.
—
Spy scandals, unprecedented congressional scrutiny, new ethical questions, revolutionary technologies, and undiminished interagency rivalries were more than enough to keep NSA’s managers busy dealing with internal crises at the start of the new decade of the 1960s, but world events were about to intrude with their own more urgent reality.
Nikita Khrushchev’s warmth and enthusiasm for the Communist regime of Fidel Castro’s Cuba were animated in part by the simple nostalgic feelings of an old Marxist-Leninist. “I felt as though I had returned to my childhood!” exclaimed Khrushchev’s deputy Anastas Mikoyan, a veteran of the Bolshevik revolution, upon meeting the Cuban leader for the first time. Here was a youthful, energetic, and committed apostle of Communism who had secured a beachhead for the revolutionary movement in Latin America; to Khrushchev, always agonizingly sensitive to the perceived stature of the Soviet Union on the world stage, the fate of Castro’s regime was a test of Soviet power and importance. The botched attempt by fourteen hundred Cuban exiles to overthrow the Cuban Communists in April 1961, the fiasco known to history as the Bay of Pigs Invasion, had been a galling embarrassment for the new Kennedy administration (whose denial of U.S. involvement had quickly proved anything but “plausible”), but to Khrushchev it added another reason to deploy the full military might of the USSR to defend Cuba—and to maintain “Soviet prestige in that part of the world,” as he later explained.40
Kennedy had run for the presidency decrying the “missile gap” between the United States and the Soviet Union. It had not taken McNamara more than a few weeks as secretary of defense to discover that all of Khrushchev’s boasts of Soviet missile strength—he had once declared that Russia was turning out missiles “like sausages”—was a bluff. Photographs taken by the new U.S. reconnaissance satellites over the previous six months showed that the Soviets had ten operational ICBMs, compared to fifty-seven in the U.S. force, and the disparity was rising rapidly. The Soviets’ SS-6 and SS-7 missiles moreover took hours to fuel and had to have their unstable liquid propellant drained every thirty days to prevent them from blowing up on the launch pad; the new U.S. Minute-man missile, entering final testing, was powered by solid propellant and could be launched in minutes.
McNamara had been at his job three weeks when the Defense Department’s public affairs officer came to him and said that as he had not yet met the Pentagon press corps, he ought to hold a press conference. McNamara tried to demur, noting that he was still learning the ropes, but finally agreed.
The first question was, “Mr. Secretary, what have you learned about the missile gap?”
McNamara innocently replied, “Oh, I’ve learned there isn’t any, or if there is, it’s in our favor.” He was astonished when a few seconds later the room emptied in a mad rush as the reporters raced to the door to break the sensational story the new defense secretary had just handed them.41
The Soviets did, however, have an abundance of medium-range missiles capable of delivering nuclear weapons; similar U.S. weapons based in Britain, Italy, and Turkey were already targeted at Russia. In August 1961, Khrushchev had given another yank on the West’s anatomy in Berlin with the sudden erection of a wall dividing the East and West halves of the city. By the next year he had decided on a more provocative demonstration of Soviet force with a move that would bring the two superpowers to the brink of nuclear war. Sending nuclear-armed medium-range ballistic missiles to Cuba would not only defend Castro’s revolution and “throw a hedgehog down Uncle Sam’s pants,” as Khrushchev put it, but would restore the credibility of the Soviet nuclear threat that McNamara’s announcement had so dramatically deflated. It would be no more than giving the Americans “a little of their own medicine,” he insisted in his memoirs, to have dozens of missiles pointing at them.42
Throughout the Soviet military buildup in Cuba the Soviets insisted on absolute secrecy. Several times Castro and his aides questioned the wisdom of that policy, arguing that if they indeed had the legal and moral right to match U.S. nuclear forces with their own, why engage in duplicity and not just announce the Cuban-Soviet military alliance openly? It was never clear how much Khrushchev thought through the entire gambit: years later his speechwriter, at an extraordinary conference of American and Soviet officials who met to reexamine the crisis, said there had always been “irrational reasons” behind Khrushchev’s decision. In any case, the Soviet leader waved away Castro’s concerns, confidently asserting that if the Americans were presented with a sudden fait accompli they would have no choice but to accept the presence of Soviet missiles ninety miles from their shores, just as the Soviets had swallowed U.S. missiles based right over their border in Turkey.43
At NSA, Cuba was the responsibility of the B1 office, since July 1961 under the direction of Juanita Morris Moody, who had gotten her start at Arlington Hall during World War II and since risen through the ranks to become one of the most senior women at the agency. Juanita Morris, like Cecil Phillips, was the quintessential accidental cryptanalyst: she was also from western North Carolina, had also left college after a year, having attended Western Carolina College from 1942 to 1943, and also happened to be living where Army recruiters from Arlington Hall were scouring for GS-2 clerks to hire during the war. While waiting for her background check to be completed she was assigned to an unclassified course in elementary cryptanalysis and was crestfallen when her clearance came through and she had to drop what she already found a fascinating study and go to work sorting messages and punch cards.
One problem that Arlington Hall’s cryptanalytic branch had put aside during the war was the German diplomatic code called GEE, a one-time-pad system used for the highest-level traffic. Morris started working on it on her own time after her day’s shift was done and soon was one of the key members of a small after-hours group that tried to find a way into the system. They succeeded in January 1945 when, with the extensive help of IBM sorts, they discovered an un
derlying pattern in the one-time-pad sheets, the product of a mechanical printing system the Germans had used to generate the sheets automatically. It was one of Arlington Hall’s most brilliant feats of pure cryptanalysis, and helped to inspire and encourage the effort to attack the Soviet one-time-pad messages.44
In 1961, disappointed at being moved from A Group, where she had been assigned for much of the 1950s, Moody was reassured by Louis Tordella, who had become NSA’s deputy director after Engstrom retired in 1958, that she was going to have the agency’s “most interesting job for the next five years.” The B1 office had begun to pick up a few wispy trails of the growing Soviet military relationship with Cuba the previous year, when monitoring of the Czechoslovak air force’s ground control voice circuits picked up Spanish-speaking pilots flying piston-engined trainers at Czech airfields.45 NSA had begun conducting “hearability” tests to see what Cuban communications could be intercepted from the closest U.S. listening posts, a Navy installation at Puerto Rico and the Army’s large World War II–era station at Vint Hill, Virginia. The answer was little. But a Navy destroyer, the USS Massey, was able to pick up signals from Cuba’s microwave-based telephone network when it circled the island in July 1960.46 A conversation between operators at two Cuban airfields intercepted the following May offered another clue of the deepening Soviet military relationship with the Castro regime: