For all of my flying career I have been given a standard explanation about how an ILS works. In the military it was described in accordance with RAF Publication AP3456 and in civilian life it was described in the ATPL exams and associated training material. With some minor differences, it is described as a pair of lobes at 90 Hz and 150 Hz. On the centreline, the two lobes overlap and you get a centred signal. To the left or right, you will get only the 90 Hz or the 150 Hz signal. The diagram looks something like this:

After reading and watching material about a particular incident in the USA (United Express UA-4933), I realized my understanding of ILS did not explain how this accident occurred. It turns out the explanation above is grossly over-simplified and creates several dangerous misconceptions that have caused some other accidents and near-misses (it is vaguely close to how the original blind landing aids were constructed but it is far away from the ILS).

It is actually quite simple to pull apart the widely used explanation when you put your mind to it. A transmitter at 90 Hz or 150 Hz would be enormous so it must of course be a modulation on a carrier signal. The 2 signals would have to be perfectly matched for the centreline to be exactly down the middle – even a slightly higher power on one side would skew the beam. Finally, the multiple orange antennas as shown in the title picture are producing lots of lobes not just 2 so there must be something else going on.

Whilst not going into the depth that an ILS engineer needs, I think a little more clarity is useful. I have tried to minimize the maths and analogue radio theory to produce a practical description. There is a link at the end to an excellent description of the radio theory involved. First you do need a little theory; I have kept the maths to plus and minus so do not be scared off. I have probably offended many radio engineers for I which I apologise.

Carrier Waves and Sidebands

Without going into too much detail, you can superimpose one signal on top of original carrier signal and modulate the original signal. Once received, this extra signal can be extracted again from the original carrier signal and used in systems or intercoms. For example a voice signal can be added to a carrier signal and transmitted; this is how ATC radio works.

What you actually get is a composite signal which consists of the original carrier wave with a pair of sidebands that are the original carrier frequency plus or minus the extra signal. If you superimpose 2 signals on the original carrier frequency, you would get one carrier and 4 sidebands (each extra signal plus or minus). When you transmit the original carrier and the sidebands, this is Carrier plus Side Bands (CSB). If you supress the carrier and just transmit the sidebands, you get Side Band Only (SBO). Since both CSB and SBO come from the same transmitter, the receiver gets both at the same time and adds them together. But phasing is critical…

Adding signals – In phase or not

When you take two signals, if they are in-phase with each other, they create a stronger signal. If completely out of phase, they cancel each other out. Any other relationship and you get a new waveform.

Adding CSB and SBO

We can use a combination of the CSB and SBO to create a new waveform. It can be thought of a simple addition just to help initial comprehension. Think of an in-phase signal as +1 and an out-of-phase signal as -1. This works for the simple theory we have seen above; two in-phase signals get stronger (1+1=2). An in-phase and an out-of-phase cancel each other out (1-1=0).

We can use this principle and phase shifting to do something useful with the CSB and SBO signals. For the CSB signal we will put both the 90 Hz and 150 Hz signal in-phase:

• CSB (+90 Hz and +150 Hz)

For the SBO, we can feed different modulations on each side of the transmitter area. On the left of the centreline, we can feed the 90 Hz signal in-phase and the 150 Hz out-of-phase

• Left SBO(+90 Hz and -150 Hz)

On the right of the centreline, we put the whole signal out-of-phase (there are other reasons for doing this. See later with respect to multiple antennas)

• Right SBO(-90 Hz and +150 Hz)

When the CSB and SBO are received together in the aircraft they are coherent (coming from the same original transmitter) so are combined in the receiver, giving the following result.

The relationship is actually more refined than this view and is a linear one. The further from the centreline that you get, the stronger the respective signal gets than the other (90 or 150 Hz). This drives the course deviation bar on the HSI – when the 90 Hz signal is stronger than 150 Hz signal the course bar goes right (steer right to get back on the centreline) and vice versa. The relationship looks like this (notice the signals cancel each other out on the centreline).

This is great but it does not provide the cone-shaped guidance we use on the ILS. There is something else going on.

Signals From Multiple Antennas

If you have multiple directional antennas next to each other transmitting on the same frequency, you can create a new signal pattern based on how these signals interact.

In Phase

If the signals are in phase, you get a new signal of roughly the same shape but narrower and with double the amplitude.

Out of Phase

Now instead put the signals from each antenna completely out of phase with EACH OTHER (180 degrees of phase difference) something amazing happens. In the centre you get a strong null as the signals cancel each other out, but either side you get a signal. Each side of the signal is from its respective antenna.

The ILS Localizer Antenna

The ILS localizer antennas uses both types of outputs from multiple antenna (in-phase and completely out of phase). The arrays are from 8 to 20 antenna.

All of the antenna transmit a CSB signal (carrier wave plus side bands) that is in-phase across all of the antennas. This produces a single lobe aligned with the centreline. The 90 Hz and 150 Hz signals are both in-phase in this signal as discussed above.

We can put the SBO signal out on the split out-of-phase lobes. As discussed earlier, the 90 Hz and 150 Hz signals are either in-phase or out-of-phase on each side of the transmitter array.

This gives us the following antenna pattern and signals.

We can see some useful outcomes here:

• The transmitters are at the far end of the runway
• The central yellow cone is your centreline.

If the power level of the CSB is varied, it makes no difference the shape of the central cone. Note the central yellow cone disappears at the touch down point on the runway.

But if you change the power output of the SBO signals, you can tailor the central section of the approach signal to tailor the width of the “on centreline” signal and where the touch down point is (eg longer runway or inset threshold). Note the on-centreline area is now wider and goes to zero width half way down the runway. This is part of the calibration process that happens on ILS; the beam width is adjusted by varying the SBO power.

Idents and System Failures

The ident code you hear on the ILS is only sent on the CSB signal. So if you cannot hear the ident signal you know the CSB is inoperative and you will not fly the approach. But what about a failure of the SBO signal?

If the SBO signal fails, this can be extremely dangerous. You will still hear the ident and you will get just the CSB signal. This has equal portions of 90 Hz and 150 Hz and will drive the course bar to the centre no matter where you are with respect to the beam. Whilst the beam is shown as a nice defined area, you are likely to pick this up outside the centreline due to sidelobes (see below). This is very dangerous as you would have a very compelling erroneous indication. This is also what happens during testing of the ILS. Watch out for NOTAMs and actually listen to the ident (which should change if it is on test).

Fortunately, the ILS should have a backup transmitter for this type of failure and should be monitored to check it is actually transmitting (interestingly this is always done in Europe but not in the USA). There have been instances where this monitoring has been interrupted and this should be NOTAM’d. It can still go wrong though (see NZ60 incident here: NZ60 – A free lesson).

Sidelobes and false localizers

When a transmitter radiates a directional beam, there is inevitably some transmissions that leak out in a direction you did not intend and these are called sidelobes. This produces weaker beams of energy outside of the main beam which could be inadvertently picked up by an aircraft. These would give the pilot a compelling but wrong guidance in the wrong direction. This is particular dangerous if a CSB sidelobe is inadvertently picked up (“on-centreline” indication when you definitely are not)

These erroneous signals can be overpowered with an extra “clearance” signal. This produces the appropriate full left or full right signals to ensure that the weak incorrect signal is not received.

The clearance signal can either be transmitted on the same transmitter array as the rest of the ILS signal or on a separate transmitter array. The signal can be transmitted on a slightly different carrier signal (8 kHz difference from the ILS frequency). This can be tolerated by the receiver as they are built to fairly wide tolerances (ILS has been around for a long time and old analogue equipment was not particularly accurate). These are called “Two frequency” ILS.

The sidelobes have been minimized with modern antenna which also means that the radiation in the wrong direction (ie back course) has also been minimized and this explains the reduction in back-course ILS.

Of note, the protected area for an ILS now makes sense. The shorter range, but wider area is the clearance signal, whereas the main beam is the longer central beam.

Errors in ILS – Snow and planes

The quality of the ILS signal can be affected by obstacles or surface contamination in front of the transmitter. These objects and surface conditions can change the position of the centreline or provide false guidance. Watch this video for an example of how snow can cause ILS errors (the ILS in this case was NOT monitored – ILS signal disrupted by snow. Also look at this article where I got some of the diagrams below –Snow effects on ILS signal propagation . I can highly recommend Mentour other incident analysis videos and Engineering Pilots articles).

Glideslope

In many textbooks the glideslope is described as being “the same principle as the localizer”. This is again an over-simplification. One crucial part about the glideslope signal is that the antenna area is normally vertical and the reflections from the ground must be considered In order to get a nice tight beam you would normally need a quite long array. This is a problem in an airfield environment as you don’t want the Eiffel Tower in the middle of the airfield. Thus some more magic occurs.

The glidelsope aims to provide guidance down to the touch down point. The antenna cannot be placed at the end of the runway as for the localizer so you will find it next to the touch down area, offset slightly from the runway. The antenna array is typically only 2 or 3 transmitters working on a UHF frequency.

The main difference with the glideslope antenna is the ground reflects some signal and forms the CSB and two SBO beams.

This ground reflection also does the phase change that is required in the lower beam. The surface in front of the glideslope antenna needs to be clear of obstacles or contamination out to significant distances which restricts sitting of glideslopes. Refer back to the NZ60 incident for how this can go wrong – the depth of snow can shallow out the apparent glideslope.

The picture above Is taken from the very good article by the Engineering Pilot. Read It here: https://www.engineeringpilot.com/post/snow-effects-on-ils-signal-propagation

Of note there are now other types of glideslope antenna that use a lateral array of antennas and do not use the reflection of signals (which needs a large flat area which may not be available). These new type of antennas are called “End Fire” antenna.

Summary

There is always a little more depth to be explored in theoretical knowledge in the aviation world. Hopefully this more detailed look at how an ILS works might help resolve an issue one day or get a pilot to stop and think about an odd indication that does not quite match what they expect. Check NOTAMs, watch for snow and always scratch the theoretical knowledge itch.

For a fuller description, I highly recommend a YouTube video here:

Use of CSB/SBO technique in the Instrument Landing System (ILS) – leepd60

Take a look at some of my other articles:

• Lights, helipad, action! The problem with new helicopter pad lights
• Helicopter on Fire – Could accident investigators have learned more?
• The Ultimate Medical Helicopter – Selecting the right machine for HEMS
• Aerial teamwork – Practical tips for working together by staying apart in emergency aviation
• HEMS Landing Sites – Reliable places to drop your medics