Identification friend or foe

Radar coverage of the Chain Home system, 1939–40

Already before the deployment of their Chain Home radar system (CH), the RAF had considered the problem of IFF. Robert Watson-Watt had filed patents on such systems in 1935 and 1936. By 1938, researchers at Bawdsey Manor began experiments with "reflectors" consisting of dipole antennas tuned to resonate to the primary frequency of the CH radars. When a pulse from the CH transmitter hit the aircraft, the antennas would resonate for a short time, increasing the amount of energy returned to the CH receiver. The antenna was connected to a motorized switch that periodically shorted it out, preventing it from producing a signal. This caused the return on the CH set to periodically lengthen and shorten as the antenna was turned on and off. In practice, the system was found to be too unreliable to use; the return was highly dependent on the direction the aircraft was moving relative to the CH station, and often returned little or no additional signal.[12]

It had been suspected this system would be of little use in practice. When that turned out to be the case, the RAF turned to an entirely different system that was also being planned. This consisted of a set of tracking stations using HF/DF radio direction finders. Their aircraft radios were modified to send out a 1 kHz tone for 14 seconds every minute, allowing the stations ample time to measure the aircraft's bearing. Several such stations were assigned to each "sector" of the air defence system, and sent their measurements to a plotting station at sector headquarters, who used triangulation to determine the aircraft's location. Known as "pip-squeak", the system worked, but was labour-intensive and did not display its information directly to the radar operators. A system that worked directly with the radar was clearly desirable.[13]

The first active IFF transponder (transmitter/responder) was the IFF Mark I which was used experimentally in 1939. This used a regenerative receiver, which fed a small amount of the amplified output back into the input, strongly amplifying even small signals as long as they were of a single frequency (like Morse code, but unlike voice transmissions). They were tuned to the signal from the CH radar (20–30 MHz), amplifying it so strongly that it was broadcast back out the aircraft's antenna. Since the signal was received at the same time as the original reflection of the CH signal, the result was a lengthened "blip" on the CH display which was easily identifiable. In testing, it was found that the unit would often overpower the radar or produce too little signal to be seen, and at the same time, new radars were being introduced using new frequencies.

Instead of putting Mark I into production, a new IFF Mark II was introduced in early 1940. Mark II had a series of separate tuners inside tuned to different radar bands that it stepped through using a motorized switch, while an automatic gain control solved the problem of it sending out too much signal. Mark II was technically complete as the war began, but a lack of sets meant it was not available in quantity and only a small number of RAF aircraft carried it by the time of the Battle of Britain. Pip-squeak was kept in operation during this period, but as the Battle ended, IFF Mark II was quickly put into full operation. Pip-squeak was still used for areas over land where CH did not cover, as well as an emergency guidance system.[14]

Even by 1940 the complex system of Mark II was reaching its limits while new radars were being constantly introduced. By 1941, a number of sub-models were introduced that covered different combinations of radars, common naval ones for instance, or those used by the RAF. But the introduction of radars based on the microwave-frequency cavity magnetron rendered this obsolete; there was simply no way to make a responder operating in this band using contemporary electronics.

In 1940, English engineer Freddie Williams had suggested using a single separate frequency for all IFF signals, but at the time there seemed no pressing need to change the existing system. With the introduction of the magnetron, work on this concept began at the Telecommunications Research Establishment as the IFF Mark III. This was to become the standard for the Western Allies for most of the war.

Mark III transponders were designed to respond to specific 'interrogators', rather than replying directly to received radar signals. These interrogators worked on a limited selection of frequencies, no matter what radar they were paired with. The system also allowed limited communication to be made, including the ability to transmit a coded 'Mayday' response. The IFF sets were designed and built by Ferranti in Manchester to Williams' specifications. Equivalent sets were manufactured in the US, initially as copies of British sets, so that allied aircraft would be identified upon interrogation by each other's radar.[14]

IFF sets were obviously highly classified. Thus, many of them were wired with explosives in the event the aircrew bailed out or crash landed. Jerry Proc reports:

Alongside the switch to turn on the unit was the IFF destruct switch to prevent its capture by the enemy. Many a pilot chose the wrong switch and blew up his IFF unit. The thud of a contained explosion and the acrid smell of burning insulation in the cockpit did not deter many pilots from destroying IFF units time and time again. Eventually, the self destruct switch was secured by a thin wire to prevent its accidental use."[15]

The United States Naval Research Laboratory had been working on their own IFF system since before the war. It used a single interrogation frequency, like the Mark III, but differed in that it used a separate responder frequency. Responding on a different frequency has several practical advantages, most notably that the response from one IFF cannot trigger another IFF on another aircraft. But it requires a complete transmitter for the responder side of the circuitry, in contrast to the greatly simplified regenerative system used in the British designs. This technique is now known as a cross-band transponder.

When the Mark II was revealed in 1941 during the Tizard Mission, it was decided to use it and take the time to further improve their experimental system. The result was what became IFF Mark IV. The main difference between this and earlier models is that it worked on higher frequencies, around 600 MHz, which allowed much smaller antennas. However, this also turned out to be close to the frequencies used by the German Würzburg radar and there were concerns that it would be triggered by that radar and the transponder responses would be picked on its radar display. This would immediately reveal the IFF's operational frequencies.

This led to a US–British effort to make a further improved model, the Mark V, also known as the United Nations Beacon or UNB. This moved to still higher frequencies around 1 GHz but operational testing was not complete when the war ended. By the time testing was finished in 1948, the much improved Mark X was beginning its testing and Mark V was abandoned.

Mark X started as a purely experimental device operating at frequencies above 1 GHz, the name refers to "experimental", not "number 10". As development continued it was decided to introduce an encoding system known as the "Selective Identification Feature", or SIF. SIF allowed the return signal to contain up to 12 pulses, representing four octal digits of 3 bits each. Depending on the timing of the interrogation signal, SIF would respond in several ways. Mode 1 indicated the type of aircraft or its mission (cargo or bomber, for instance) while Mode 2 returned a tail code.

Mark X began to be introduced in the early 1950s. This was during a period of great expansion of the civilian air transport system, and it was decided to use slightly modified Mark X sets for these aircraft as well. These sets included a new military Mode 3 which was essentially identical to Mode 2, returning a four-digit code, but used a different interrogation pulse, allowing the aircraft to identify if the query was from a military or civilian radar. For civilian aircraft, this same system was known as Mode A, and because they were identical, they are generally known as Mode 3/A.

Several new modes were also introduced during this process. Civilian modes B and D were defined, but never used. Mode C responded with a 12-bit number encoded using Gillham code, which represented the altitude as (that number) x 100 feet - 1200. Radar systems can easily locate an aircraft in two dimensions, but measuring altitude is a more complex problem and, especially in the 1950s, added significantly to the cost of the radar system. By placing this function on the IFF, the same information could be returned for little additional cost, essentially that of adding a digitizer to the aircraft's altimeter.

Modern interrogators generally send out a series of challenges on Mode 3/A and then Mode C, allowing the system to combine the identity of the aircraft with its altitude and location from the radar.

The current IFF system is the Mark XII. This works on the same frequencies as Mark X, and supports all of its military and civilian modes.

It had long been considered a problem that the IFF responses could be triggered by any properly formed interrogation, and those signals were simply two short pulses of a single frequency. This allowed enemy transmitters to trigger the response, and using triangulation, an enemy could determine the location of the transponder. The British had already used this technique against the Germans during WWII, and it was used by the USAF against VPAF aircraft during the Vietnam War.

Mark XII differs from Mark X through the addition of the new military Mode 4. This works in a fashion similar to Mode 3/A, with the interrogator sending out a signal that the IFF responds to. There are two key differences, however.

One is that the Interrogation pulse is followed by a 12-bit code similar to the ones sent back by the Mark 3 transponders. The encoded number changes day-to-day. When the number is received and decoded in the aircraft transponder, a further cryptographic encoding is applied. If the result of that operation matches the value dialled into the IFF in the aircraft, the transponder replies with a Mode 3 response as before. If the values do not match, it does not respond.

This solves the problem of the aircraft transponder replying to false interrogations, but does not completely solve the problem of locating the aircraft through triangulation. To solve this problem, a delay is added to the response signal that varies based on the code sent from the interrogator. When received by an enemy that does not see the interrogation pulse, which is generally the case as they are often below the radar horizon, this causes a random displacement of the return signal with every pulse. Locating the aircraft within the set of returns is a difficult process.

During the 1980s, a new civilian mode, Mode S, was added that allowed greatly increased amounts of data to be encoded in the returned signal. This was used to encode the location of the aircraft from the navigation system. This is a basic part of the traffic collision avoidance system (TCAS), which allows commercial aircraft to know the location of other aircraft in the area and avoid them without the need for ground operators.

The basic concepts from Mode S were then militarized as Mode 5, which is simply a cryptographically encoded version of the Mode S data.

The IFF of World War II and Soviet military systems (1946 to 1991) used coded radar signals (called Cross-Band Interrogation, or CBI) to automatically trigger the aircraft's transponder in an aircraft illuminated by the radar. Radar-based aircraft identification is also called secondary surveillance radar in both military and civil usage, with primary radar bouncing an RF pulse off of the aircraft to determine position. George Charrier, working for RCA, filed for a patent for such an IFF device in 1941. It required the operator to perform several adjustments to the radar receiver to suppress the image of the natural echo on the radar receiver, so that visual examination of the IFF signal would be possible.[16]

By 1943, Donald Barchok filed a patent for a radar system using the abbreviation IFF in his text with only parenthetic explanation, indicating that this acronym had become an accepted term.[17] In 1945, Emile Labin and Edwin Turner filed patents for radar IFF systems where the outgoing radar signal and the transponder's reply signal could each be independently programmed with a binary codes by setting arrays of toggle switches; this allowed the IFF code to be varied from day to day or even hour to hour.[18][19]