Part Two: The Seige
Editor's Note
Essentially the same editorial conventions followed in Part One have been applied to Part Two. Slightly more liberty has been taken in condensing the original text. Where the meaning of the sentence is unaltered, such deletions are not identified. Portions of the narrative that are summarized rather than directly transcribed are again in italics. USM-808 was ASA's 8th Radio Research Field Station at Phu Bai. USN414J was sub-unit one of the USMC's 1st Radio Battalion at Da Nang; J4 was the detachment at Khe Sanh.
* * * * *
SIGINT had helped U.S. military commanders in tracing the movements of NVA units into the Khe Sanh sector and in recovering their order of battle. Colonel Lownds credited SIGINT with providing him valuable early warning as various enemy forces entered the Khe Sanh area. NVA units moving southward could continue along the infiltration routes to the A Shau Valley or to the triborder area of Laos, Cambodia, and Kontum Province of South Vietnam. The question was whether or not they would make Khe Sanh itself their objective. When SIGINT revealed that the units had remained in Laos opposite Khe Sanh for some time, Colonel Lownds predicted that the objective of the units was Khe Sanh. His prediction was confirmed by other intelligence sources, and a clear pattern of NVA deployments against Khe Sanh emerged. Some days later, [the officer in charge of the Marine SIGINT detachment] came to him and reported: “Colonel, they're here!”—ARDF had indicated that 304th Division troops were unmistakably inside South Vietnam. SIGINT helped Colonel Lownds estimate the size of the force he faced; ARDF now told him where those forces were located.
SIGINT Support in the Early Part of the Battle
SIGINT during the first few days of the battle provided the supported commanders with information on the enemy's organization and command structure, his dispositions, and his tactical intentions. Although ARDF fixes furnished the most immediately usable combat intelligence, analysis of the numerous enemy tactical communications collected during these early days also provided valuable information on enemy activities. USM-808 [Phu Bai] reported that the NVA 29th Regiment of the 325C Division, which was involved in the initial actions had resumed tactical communications activity for the first time since the previous September. In the first few days of the campaign, in-country SIGINT analysts followed closely the communications passed by tactical intelligence elements of the 325C Division and outlined the activities of these intelligence units in their reports. The categories of information being collected by the enemy intelligence units—Allied dispositions, movements, and helicopter and other air activity, for example—as revealed in enemy messages intercepted, constituted in themselves an indication of North Vietnamese tactical interests and intentions. SIGINT released on 26 January stated that tactical intelligence elements of the 101D Regiment were collecting intelligence at least as early as 23 January, when one message mentioned Hill 595. Several days later, two tactical intelligence elements associated with the 101D Regiment were reporting Allied helicopter activity west and southwest of Khe Sanh.1
Meanwhile, SIGINT was providing the Khe Sanh force and others with information on enemy artillery operations. For example, SIGINT derived from messages passed by tactical intelligence elements pointed to the presence of 12.7-mm guns in the area, the first SIGINT confirmation of enemy antiaircraft guns near Khe Sanh. In another instance, an NVA tactical voice net came up on 20 January, apparently to serve artillery elements, since information passed on the net concerned NVA reconnaissance and fire direction. SIGINT also yielded material on the evolving enemy command structure in the Khe Sanh area, revealing that the Khe Sanh Area Front had added at least seven new subscribers, including headquarters of the 304th Division and two of its subordinate elements. By this time, SIGINT was reporting that KSAF was communicating with 19 subscribers, including Hanoi, two other division HQ, and the entity responsible for communications along the Ho Chi Minh Trail. On 22 January, an ARDF fix on the 320th Division headquarters located the headquarters well south of the DMZ in the vicinity of Cam Lo, considerably southeast of its 15 January location. The relocation of the headquarters to the Cam Lo area signified to SIGINT analysts that the 320th Division was scheduled for duty east of Khe Sanh, in the central DMZ area. Before that time, in view of the division's sporadic communications with KSAF and the presence of elements of the division near Khe Sanh, SIGINT analysts had suspected that the 320th would join the concentration near Khe Sanh.
The View in Washington
At the national level, [the U.S. Intelligence Board] reviewed SIGINT and collateral information on the early phase of hostilities at Khe Sanh and drew some inferences: “North Vietnamese forces in the area of Khe Sanh and the DMZ have begun the offensive against U.S. positions for which the buildup has been under way for several weeks. Reports suggest that the enemy attacks are to be intensified just prior to and after the Tet holiday. Prisoners have confirmed that the mission of the NVA 325C Division is to launch attacks in the Khe Sanh area, while SIGINT has reflected a further concentration of major elements of the NVA 304th Division in the area, including its suspect artillery regiment.” In a hint at the coming Tet Offensive, the report also noted that "there is mounting evidence that the NVA operations in the DMZ/Khe Sanh area are part of a general plan to intensify Communist operations throughout much of South Vietnam; there have been numerous indications of preparations for coordinated attacks in the near future elsewhere in the two northernmost provinces and in the central highlands." (See "The 1968 Tet Offensive, Part One: Uncertain Signals" for a broader picture of the situation in January, 1968.)
USIB analysts were concerned about the possibility that the Khe Sanh campaign was merely a prelude to a major enemy effort to seal off the two northern provinces. At the theater level, MACV was trying to evaluate the extent of the threat to Khe Sanh and to ascertain whether the enemy planned major activity elsewhere in the I Corps Tactical Zone just before the main attacks on Khe Sanh got under way. It was clear that MACV's primary concern was the immediate threat to Khe Sanh, while national level consumers feared that the units near Khe Sanh might stage damaging attacks in other areas. SIGINT could, and did, answer many questions—but not this one.
Exploiting the Enemy's Tactical Voice Communications
One valuable source of direct and timely SIGINT support to the Khe Sanh base during the campaign came from enemy tactical voice communications that USN-414J4 was able to intercept and exploit. USM-808 supplemented this effort by processing voice traffic intercepted by the airborne collection platforms operating under USM-808's guidance and tasking, and both USM-808 and NSA provided technical backup. The enemy began using tactical voice communications in the Khe Sanh area during the third week of January. At first the Khe Sanh crew taped the traffic and sent it to Da Nang for translation, but this process was much too slow for tactical exploitation. A voice intercept team, including an NSA civilian, was dispatched to Khe Sanh. The American linguists developed the voice networks serving artillery elements and other NVA combat units.
The heaviest volume of voice intercept came from artillery communications and from the 304th Division headquarters net, although the team also heard a considerable volume of voice traffic on regimental nets. Towards the end of January, Col. Lownds became aware of a South Vietnamese SIGINT team supporting the ARVN 37th Ranger Battalion in a corner of the base about a mile away from the Marines’ listening post. The ARVN troops were of course natural born linguists, and the two groups were merged. Directed by the Marines, the Vietnamese were able to dramatically reduce translation times. Intelligence of immediate value was immediately relayed to Col. Lownds by telephone. On one occasion, the actual order to fire was heard over an artillery voice net. The command post was instantly notified. The warning siren sounded, and the Marines headed for cover before the first shells landed—a fact evidently noted by the enemy. Passing of fire orders over the voice net did not occur again; orders thereafter apparently went over more secure communications facilities. Artillery firing plans, however, continued to appear periodically on the voice net and constituted valuable warnings for the men at Khe Sanh.
SIGINT derived from voice intercept also gave forewarning of enemy ground probes. Often in late afternoon, the enemy sent out messages ordering a force of undetermined echelon to launch a night probe. Although the location of the probe would not be given, the Marines knew it would probably be against the ARVN Ranger sector since they almost always occurred in that area. On receiving the warnings the ARVN Rangers and U.S. Marines would reinforce the sector and ask for extra artillery support. The probes were invariably blunted and the enemy usually incurred heavy casualties. About 90 percent of these particular enemy ground actions were predicted in advance. usually with at least a 4-hour warning.
Manual Morse Intercept and Processing
Although manual Morse intercept and processing by the Khe Sanh SIGINT detachment was less productive than voice intercept, traffic analysts who worked on the Morse intercept operations made significant contributions to on-base intelligence. The traffic analysts provided valuable guidance for ARDF through their knowledge of enemy manual Morse nets in the Khe Sanh area and kept the Marine force informed on SIGINT order of battle information that Phu Bai and NSA produced from traffic analysis of the Morse communications. Through its own coverage and that of the airborne collection and ARDF resources it controlled, USM-808 passed hundreds of ARDF location reports, tactical reports (TACREP’s), artillery warnings, and so forth to the intelligence staff of the 3d Marine Division at Dong Ha. Khe Sanh exploited low-level tactical manual Morse communications serving the NVA 325C Division's regiments and reconnaissance elements, which used little tactical voice. Thanks to augmented two-way secure communication, Khe Sanh received timely SIGINT from outside its immediate area, while Khe Sanh’s raw intercepts were more quickly forwarded to Phu Bai or direct to NSA for analysis.
SIGINT and Firepower
General Westmoreland called on all three Services to render close-in and general fire support to the Khe Sanh defenders. Much of the intelligence on which the Army targeted its 175-mm guns, the Marine, Air Force, and Navy targeted their tactical fighters, and the Air Force targeted its B-52 bombers came from the NIAGARA all-source intelligence evaluation team at MACV in Saigon. (See "Operation NIAGARA in a previous article.) Much of the team's data, in turn, came from various ARDF sources. ARDF was the primary source for targeting the B-52 program, which normally allocated 48 sorties per day in support of Khe Sanh, with additional missions flown against special targets. MACV applauded the role of ARDF in ARC LIGHT targeting. A message to the commander of the 8th Radio Research Field Station at Phu Bai noted that “Gen. Westmoreland has ordered TAC air, artillery, and B-52 strikes based on fix[es] reported by your unit. BG Davidson [head of MACV intelligence] is extremely pleased with the good and timely reporting and has asked that you be especially watchful during the next few hours for any reflections of the above operation. [NSA Representative, Vietnam] extends its congratulations to the 8th RRFS for its continuing fine reporting efforts.” No mention was made of the EC-47 and ASA crews who made the fixes. Information from the NIAGRA operation amplified by SIGINT from Phu Bai was used to target a devastating 30-plane B-52 strike against the Khe Sanh Area Front headquarters on 30 January.
And More of the Same . . .
The North Vietnamese made a number of ground attacks on the various hill positions that the Marines held beyond the base perimeter but attempted only one major ground assault on the base itself. That assault, in about regimental strength, came on the night of 29 February-1 March. On 29 February, an unidentified artillery element reported that it was going to conserve its ammunition in order to have enough to “support at night.” Later that day, the same source stated. "Tonight, the infantry will attack. Our artillery will assist." About that time, the North Vietnamese probed the western side of the perimeter, causing Colonel Lownds to speculate that the probe was a feint and that the North Vietnamese would attempt to break through on the eastern side of the perimeter in the ARVN Ranger area.
Basing his actions on SIGINT and his own knowledge of the surrounding area, Colonel Lownds made his plans to repulse the attack. He picked what he believed to be the most logical route of approach for the enemy force and selected a killing zone along the route. Next he estimated how long it would take the force co traverse the zone, and then arranged for a B-52 strike on the zone, to be followed by an artillery barrage. He also reinforced the ARVN perimeter. During the night of 29 February-1 March NVA troops three times tried to break through the ARVN sector of the perimeter but each time were driven back with heavy losses. The main NVA force, however, never reached the perimeter—it was caught in Colonel Lownds' killing zone by the B-52's and the artillery. During the barrage, the Khe Sanh SIGINT detachment reported that the two NVA stations in the area were still operating. Suddenly one operator announced that shells had just barely missed his position. The near miss was immediately reported to Colonel Lownds, who adjusted the artillery fire accordingly. When the next rounds landed, the two stations went off the air and, Colonel Lownds noted with satisfaction, "were never heard again."
The 29 February attack was the last major ground assault in the Khe Sanh area. During late February and early March North Vietnamese units began to show the results of th e severe poundings by Allied air and artillery, often targeted by ARDF. Although it would not become apparent for some weeks, the issue at Khe Sanh had already been decided. However, as of early March, national-level consumers still feared that the Khe Sanh campaign was only the prelude to a major enemy offensive designed to wrest the northernmost provinces in I CTZ from Allied control. At Khe Sanh, Ground probes all but ceased, but enemy shelling did not. The volume of fire in March actually exceeded that of February.
On 6 March an enemy message revealed that the communications troops of the NVA 325C Division's 101D Regiment were to turn over their station to unidentified replacements. SIGINT on 13 March reported that tactical communications serving the 101D Regiment had not been heard for a week. USM-808 [8th RRFS, Phu Bai] postulated that this communications change indicated a scheduled repositioning of the division. On 20 March USM-808 reported that ARDF fixes indicated the 325C Division headquarters and its 95C and 101D Regiments had withdrawn to Laos. Thus, the SIGINT evidence for withdrawal of the 325C was firm.
Epilogue
At the end of March enemy tactical pressure near Khe Sanh was stepped up for what turned out to be the last time. On 30 March, just three kilometers from Khe Sanh, a company-sized Marine patrol engaged an enemy unit estimated to be of battalion size. During the action the North Vietnamese lost 115 killed. SIGINT provided tactical commanders with an enemy assessment of this encounter when it intercepted a message from an unidentified station reporting on a "battalion-sized" American attack on its positions. The message revealed that the NVA units involved had taken heavy casualties and that an entire battalion would be used to recover the dead.
Meanwhile, COMUSMACV created an all-service Provisional Corps, Vietnam, headquarters that assumed temporary command of all U.S. forces in I CTZ and undertook planning for Operation PEGASUS; the relief of Khe Sanh. The operation would be spearheaded by the 1st Cavalry Division, with ASA’s 371st Radio Research Company and its three Huey UH-1 LEFT BANK ARDF helicopters in support. The range of responses by the action units to the [LEFT BANK and fixed wing] ARDF fixes included air and artillery strikes, reconnaissance, and ground probes. The entire process was usually accomplished within 10 minutes of an ARDF fix, representing a major improvement in the timeliness of ARDF support in that the [sanitized] information was passed directly to the action units themselves as well as to the division G-2.2 Many tactical commanders acknowledged that SIGINT provided the major part of the immediately usable tactical intelligence available to their commands during the defense of Khe Sanh and in actions in the northern provinces of South Vietnam. In discussing the value of SIGINT support to his command, Col. Charles A. Morris, who played a major role in planning Operation PEGASUS in his capacity as G-2 of the Provisional Corps, Vietnam, stated: “From my Corps G-2 point of view, SI was the only reliable source of early warning on enemy infiltration and movements … At the Corps and higher level in SVN, 85 percent of the early warning intelligence derived from SIGINT … In terms of early recognition and early warning of the approach of the enemy forces and tracking them toward position locations, it is the best source.”
_____
Footnotes.
1. Enemy electronic warfare capabilities were much greater than most U.S. commanders knew or cared to admit. The NVA was frequently able to exploit poor U.S. communications security practices, mostly through intercepting and interpreting tactical voice traffic.
2. The SIGINT support for PEGASUS was somewhat unusual in that the LEFT BANK choppers were organic to the 101st rather than being controlled by ASA. For details, see the full text of Focus on Khe Sanh, pages 86-89.
Article transcribed and uploaded by Joe Martin
20 July 2020