Military Embedded Systems

Sensor networks and warfare


September 25, 2015

Ray Alderman

VITA Standards Organization

EVOLUTION OF WARFARE BLOG. In this segment, we will take a look at the sensor networks used by our intelligence community and military and their importance to maintaining secrecy in warfare. As Sun Tzu stated in "The Art of War," "secret operations are essential in war; upon them the army relies to make its every move."

The earliest of these electronic fences appears after World War II and then another during the VietNam war. Even though the technology was unrefined back then, the concepts proved invaluable in future implementations of ground-based, airborne, shipborne, and satellite intelligence networks.

In 1961, the Sound Surveillance System (SOSUS) network was deployed by the U.S. Navy. Thousands of sensitive hydrophones were dropped onto the seafloor in the Greenland-Iceland-United Kingdom (GIUK) gap, the Mid-Atlantic Ridge off the east coast of the U.S., in the Barents Sea near Russian submarine bases, and in some Pacific ocean locations. The primary mission of this network was to detect and track Russian submarines, but all naval traffic could be monitored.

The sensor arrays would detect enemy subs and send their signals to mainframe computers (upgraded to Cray supercomputers later in the program), in the basement of the Naval Ocean Systems Center in Norfolk, Virginia. The computers would discern which specific Russian boat it was (from screw sound analysis) and continuously map it during its mission. That positional information was shared with the U.S. submarine fleet, who could then play cat-and-mouse games with the Russian subs.

This network played a part in the Cuban Missile Crisis in October 1962. Eleven U.S. Navy ASW (anti-submarine warfare) equipped destroyers knew that twelve Russian submarines had come across the SOSUS fence heading into the Atlantic, but the system had just been deployed and did not have the database of screw sounds to identify each ship. Using land based HFDF (high frequency direction finding) and P2 Orion sub hunter data, U.S. destroyers intercepted, hounded, and cornered one of them: the Russian submarine B-59. The captain of that sub wanted to fire a nuclear-tipped torpedo into the destroyers, believing that a nuclear war had already started over the Cuba blockade, but Vasili Arkhipov, the fleet commander aboard the B-59, refused to let the captain fire the nuclear weapon and ordered him to surface instead. Thanks, Vasili.

In October of 1986, Russian nuclear submarine K-219 suffered an explosion in one of its missile tubes while submerged about 600 miles northeast of Bermuda. SOSUS sensors detected several Russian freighters changing their courses and heading toward Bermuda, an unusual occurrence. The SOSUS network also knew the K-219 was in that area. Nearby, sailors on the nuclear sub USS Augusta heard the explosion on their hydrophones and changed course to investigate. The explosion onboard K-219 so damaged the sub that it eventually sank, with its remaining 15 nuclear-tipped missiles, to the bottom off the coast of North Carolina. In 1988, a Russian hydrographic ship found the wreck three miles down, sitting upright, with several of its missile tubes forced open. The missiles, along with their nuclear warheads, were missing. Decide for yourself who took the missiles from the tubes and where they went. You can read about both incidents in Peter Huchthausen’s books, “October Fury” and “Hostile Waters” respectively.

Operation Igloo White

In 1967, US Air Force OP-2 Neptune aircraft, helicopters, and F-4 Phantom jet fighters dropped 20,000 battery-powered sensors into the jungle in Laos, along the Ho Chi Min Trail. Some sensors were attached to camouflage parachutes, to hang-up in the jungle canopy invisibly, and some looked like darts with long blades of grass for their tails. The long tails were actually antennas made to look like weeds on the ground, once their points embedded in the soft dirt.

The objective of this “sensor network" was to detect troop and supply movements by the NVA (North Vietnamese Army) at points along the Trail and report them. The seismic sensors would detect vibrations from marching soldiers and heavy vehicles. The RF sensors would detect vehicle ignition systems operating on trucks. Chemical sensors would detect ammonia as large groups of soldiers create a lot of urine in the hot jungle.

The sensor signals were transmitted to Navy EC-121R or QU-22B reconnaissance planes circling nearby, and relayed to dual IBM 360/40 mainframe computers (later upgraded to 360/65 machines) sitting at an air base near Nakhon Phanom, Thailand. The sensor positions would be mapped by the computers and handed over to the pilots for their bombing runs in near real time. The program was very successful through 1969, until the NVA figured out why they were being hit regularly and accurately. So, they drove cattle near the seismic sensors to simulate troop movements, and hung buckets of urine near the ammonia sensors while their troops and supply convoys took another route. You can read more about this network in the books “Kill Chain” by Andrew Cockburn and “The History of Big Safari” by Col. Bill Grimes.

Ground-based intercept networks

In 1912, Herbert Yardley got a job as a clerk and telegraph operator at the U.S. State Department in Washington. He then secretly began to try his hand at deciphering the coded messages between the state department and the White House, since his boring job was simply to pass them through. This is one of the first modern examples of encrypted communications being intercepted and exposed to cryptanalysis.

In 1917, he became an officer in the Army and established MI-8, the code and cypher section of Army Military Intelligence, known as "The Black Chamber." Over the next decade, he and his group would pilfer copies of encrypted cables and telegrams coming through the commercial telegraph offices, mostly those to and from foreign embassies on American soil, and decode them. Soon, the radio would offer safer and greater opportunities to surreptitiously grab communications traffic.

William Friedman took over the code and cypher initiatives in 1929, and outlined the structure and mission of modern COMINT (COMmunications INTelligence) by creating the processes for signal intercept, goniometry (direction finding), traffic analysis, decryption of encoded messages, and eventually, RFP (Radio FingerPrinting). In 1930, the Signal Intelligence Service, a division of the War Department, was born. It would later become the NSA. We will focus here on the interception networks used in COMINT. The history of cryptology is a broad and sticky topic, and will be the subject of another blog.

First, let’s get some acronyms out of the way, to avoid confusion. IMINT is IMagery INTelligence, starting with film cameras on aircraft and satellite platforms, and evolving into the high0resolution electronic imaging sensors we have today. This includes infrared (IR) sensors. SIGINT is SIGnal INTelligence, the interception of electronic signals, and this group is divided into two categories: COMINT is COMmunications INTelligence, interception of communications between two people, either voice or digital; and ELINT is Electronic Intelligence, the interception of signals not used specifically for communications, like RADAR. Under ELINT, there is the subdivision, FISINT (Foreign Instrumentation Signals INTelligence), which are signals coming from foreign airborne, surface, and submerged systems.

Then, under that heading, we have another subdivision: TELINT (TELemetry INTelligence), the interception of signals from missiles and other unmanned platforms that send their position, speed, engine status, and other operational data to its controllers (machine to human). Next, we have MASINT: Measurement And Signature INTelligence, the interception of signals that identify targets by their unique characteristics (like the screw sounds of a specific ship, or the RADAR signature of a specific type of missile). Then, of course, there is HUMINT, intelligence gained from human sources like captured terrorists or information from informers. As you can see, most of these sources are based on advances involving technology. There could be some other xxxINT designations based on new collection techniques, but they may be classified at this point.

From Friedman’s work, we have the primary elements of COMINT. First, we must find the radio signals to intercept with our receiver. That is done by “spinning the dial” until you find them. Once found, that frequency is logged for continuous use. Second, we must establish the location of that transmitter: direction finding (DF). That is done by multiple receivers, in geographically different places using directional antennas, tuning-in the specific frequency. That yields an azimuth from each receiver. Where those different azimuths cross is where the transmitter is sitting on a map. Later, as technology improved, the carrier waveform of that specific transmitter could be photographed, analyzed, and its unique fingerprint established. Next, the intercept operators copy the message. The date-time stamp and message externals are given to the traffic analyst who logs them in for that transmitter. He also maps the location since the transmitter moves with its unit, and he can track those movements with the next DF intercept. The message then goes to the cryptanalysts for decoding. As time went on, hundreds of locations and thousands of people were eventually involved in this process, both civilian and military.

Fast forward to today where all of this is done by computers and sophisticated frequency-hopping multi-channel radio receivers. Waveforms are analyzed and fingerprinted by digital signal processors (DSPs) using complex algorithms. All the “externals”, encrypted traffic, frequencies, and DF location information is logged into huge databases on large computer networks. There are still many ground-based intercept stations in the world today, run by the intelligence agencies and military intelligence groups.

Ship-borne sensor networks

The same basic systems were installed on U.S. Navy ships in the 1960s to accomplish COMINT missions on enemy naval communications on the seas and along enemy coastlines. At first, all the equipment was packed into shipping containers and attached to the decks of U.S. Navy destroyers. Destroyers attract a lot of attention being warships, and that’s not good for secret SIGINT missions. So, small cargo ships were outfitted with the equipment and antennas so as not to attract attention.

The plan was for a fleet of these spy ships, but only about seven were ever built and put into service. One of those was the USS Pueblo. In January 1968, the North Koreans attacked, commandeered, and took the Pueblo to one of their ports with much of the classified equipment and sensitive information on board intact. They eventually released the crew, but the Pueblo was never returned. It is anchored in the Botong River near Pyongyang to this day. Twenty-four months after the Pueblo was lost, all of the SIGINT ships in the U.S. Navy were decommissioned and sold for scrap. Ships are just too slow and vulnerable for SIGINT missions. However, the Russians still have their SIGINT boats, disguised as fishing trawlers. More than 40 of them are still out there according to books on the topic.

There’s a lot more to this subject, but not enough space for it here. If you want more details and a lot more history, read James Bamford’s book, “The Puzzle Palace”. We’ll look at airborne sensor networks and satellite networks in our next installment.


Featured Companies

VITA Standards Organization

9100 Paseo del Vita
Oklahoma City, OK 73131