>>> 2021-03-13 CONELRAD (PDF)

Here in the year 2021, an extended, concerted effort across multiple levels of the government and the telecommunications industry has made it possible for the government to send short text messages to cell phones. Most of the time, it even works. This sophisticated, expensive capability is widely used to send out mistyped descriptions of vehicles potentially containing abducted children, and nearly nothing else.

Before we lived in the modern era of complicated technology that barely works, though, the Civil Defense administration developed an emergency notification solution that was simple and barely worked: CONELRAD. CONELRAD is, of course, short for Control of Electromagnetic Radiation, which in a way is what all radio transmitters do. In the case of CONELRAD, though, the control aspect has special significance.

CONELRAD, introduced in the early stages of the cold war in the '50s, was intended to provide timely information to the public about an incoming Soviet bombing mission. Because bombers, presumably delivering nuclear weapons, were relatively slow and could be detected relatively quickly, early public warning of attack could save many lives. This was especially true due to the generally lower-yield nuclear weapons in consideration at the time. The problem, though, was finding a way to get warning and instructions out in a matter of minutes.

This is actually a bit of a misrepresentation of the history of CONELRAD, but it's a very common one since the emergency alerting feature of CONELRAD was the most widely advertised and the most successfully implemented. In actuality, though, CONELRAD was designed as an active defense system in addition to an emergency alerting system. This part of CONELRAD is not so well known.

Understanding this requires a trip back to World War II, and specifically the air campaigns occurring over Britain (by the Germans) and Germany (by the Allies). During WWII, air navigation was in its infancy. Navigation for fast-evolving situations like bombing runs was based on sighting landmarks and dead reckoning ("pilotage"), which is already difficult at night and especially difficult when the targets are using active denial techniques (blackouts) to make landmarks difficult to see. Bombing runs, though, were far more effective at night when air defense personnel suffered the same challenges---of it being difficult to see things when it's dark.

The result was a huge drive for radio-navigation technology, which would work just as well at night as during the day. Although radio-navigation would later involve all kinds of interesting encoding techniques [1], the simplest and earliest radio technology for air navigation was a simple directional receiver. A small loop antenna outside the fuselage could be rotated around to identify the angle at which a signal is nulled, giving the direction to the signal. This allowed airplanes to fly towards radio transmitters whether or not they could see anything, which as you can imagine was tremendously useful to bombers.

The commercial radio stations in major cities quickly went from a valuable communication asset during a blackout to an unintended navigation aid for the enemy. In Britain and Germany, where this technique was seeing active use in targeting cities, countermeasures had to be developed.

One option is obvious: when incoming bombers are detected, just shut off radio stations. This deprives the bombers of guidance but also deprives the community of information, which would be especially critical following a nuclear attack.

CONELRAD offered a smarter solution: keep (some) radio stations on the air to deliver information, but have them operate in such a way that they would be confusing and useless to aircraft.

The history of air defense in the US is a somewhat strange one in large part because the US has never actually had a need for them. That is, there has never been a significant bombing campaign on the mainland US. As a result, much of the US air defense infrastructure has always been hypothetical in many ways. The CONELRAD proposal fell into a perfect time window when a Soviet nuclear attack had become a public concern but was still expected to be delivered by aircraft. Air defense, briefly, was a focus of the Cold War.

So, form this perspective of CONELRAD functioning primarily as an active denial system for air defense, let's take a look at how it worked.

When an Air Defense Control Center (ADCC) detected an incoming bombing mission, a set of leased telephone lines between the ADCC and major radio stations would be used to activate CONELRAD. The activated radio stations would first alternate their transmitters off and on, five seconds each, twice. Then, a 1kHz tone would be transmitted for 15 seconds.

AM radio stations designed as emergency stations would then switch their transmitters (or more likely switch to an appropriately configured standby transmitter) to either 640 or 1240 kHz. These radio stations would then broadcast emergency information, but with a twist: following a pre-arranged schedule, the stations on 640 and 1240 would alternately shut down and start up their transmitters, every few minutes, on a cycle of several stations This would frustrate any aircraft trying to navigate by these stations as their "target" would keep changing positions.

In the mean time, all other radio stations would simply shut down, making the cycling AM stations the only option.

There are a few things to unpack about this. First, the five-second off/on cycling of activated radio stations and 1kHz tone are both measures to allow automated receivers to take action when an alert is issued. This feature of emergency radio broadcasts persists today in the form of a dual-frequency attention tone and SAME preamble repeated thrice. Various automated receivers were offered for CONELRAD and some radio broadcasters used automatic receivers that disabled their transmitter in response to either the monitored station's carrier dropping or the 1kHz tone.

Second, some radio stations would be expected to change their transmit frequency and power. It is not clear that there was a specific need for CONELRAD transmitting stations to reduce power, I suspect it may be an artifact of the frequency change process. Some stations might achieve the frequency change by switching to an already-prepared standby transmitter, which was often of lower power due to infrequent use, and other stations actually changed the carrier frequency of their primary transmitter... but did not have time to adjust the other stages of the transmitter (and antenna), resulting in poor tuning and low efficiency.

As a result, following the CONELRAD activation sequence there would often be a long and uncomfortable silence as CONELRAD transmitting stations reconfigured. In practice, they didn't always come back at all, because the frequency change-out process was complex and presented a substantial risk of things going wrong. Just the five-second off-on cycle came with great risk; large transmitters used large tubes that operated at high temperature and often did not respond well to rapid changes in power.

As a result, CONELRAD implementation was costly and complex for participating radio stations. This introduced a lot of friction to CONELRAD adoption, and while it's extremely difficult to find detailed information, it seems that CONELRAD's deployment was always severely limited. I have read before that very, very few CONELRAD transmitting stations every fully implemented the ability to cycle their transmitters on and off in an ongoing cycle, it was viewed as too difficult and risky. There were relatively few tests of full CONELRAD capabilities, which left a lot of questions around the actual performance of the system.

Perhaps because of the technical complexity of the active denial component, later public discussion of CONELRAD generally identifies it only as an emergency communications system. The air defense mission was largely forgotten.

CONELRAD faced challenges beyond its own complexity. The development of ICBMs made the air defense component obsolete, as ICBMs used inertial (dead-reckoning) navigation that could not be confused by radio station trickery. In 1963, CONELRAD was replaced by the Emergency Broadcasting System. The EBS entirely eliminated the air defense component, allowing stations to continue to operate on their normal frequencies for the purpose of delivering messages.

EBS would shortly after be replaced by the Emergency Alert System, EAS, which for the most part is still what we use today---but it has been augmented by a baffling number of accessory systems which handle various types of alerts over various media. This includes Wireless Emergency Alert (WEA), the system which has a modest success rate in delivering text to smartphones.

Despite CONELRAD's relatively short lifespan and limited success, its design was highly influential on emergency alerting systems since. The basic pattern of a set of key radio stations broadcasting an attention sequence which causes other radio stations to switch to an emergency mode remains in used today. Modern radio stations use a device called an ENDEC which, depending on the radio station, typically monitors some other radio station further up the tree for the transmission of an attention sequence. In this way emergency alerts propagate downwards from key (entry point) radio stations to all other radio stations.

The modern structure of EAS is just so much more complicated than you would ever reasonably expect, which means it's my kind of thing. Maybe I'll write about it some time.

[1] In fact, some more advanced radio-navigation technologies were in use prior to 1950 including for WWII bombing operations. This is an interesting topic that I hope to write about in the future.