How do sensors handle interference due to the reflected ray?

[JC2000]

TL;DR Summary

If the emitting and receiving instruments are adjacent to each other then wouldn't the emitted wave interfere with the reflected wave and affect the sensors function?

I am unable to see how a sensor where the receiving and emitting instruments are next to one another deal with wave interference between the emitted and reflected wave.

 

[Baluncore]

Can you be more specific about the type of sensor and the wavelength you are considering.

 

[JC2000]

Can you be more specific about the type of sensor and the wavelength you are considering.

I was considering ultrasonic and infrared sensors in general, didn't have any specific models in mind...

 

[Baluncore]

When a receiver is adjacent to a transmitter, the two will probably alternate in time. If they both operate at the same time then the channels must be separated in frequency (ultrasonic), wavelength (infrared), or by being spread spectrum with different spreading codes.

I do not understand what you mean by “reflected wave”. A receiving transducer absorbs energy from the signal, but some energy incident on the transducer is reflected, and so is lost.

A transmitter can overload and desensitise a nearby receiver if there is insufficient separation in space or time.

 

[JC2000]

By reflected wave I meant the wave that is reflected off the reflecting surface (or the object which is to be detected). My question was based on the diagram above. I was wondering if the incident wave would interfere with the reflected wave.

Your point about spacing the pings from the emitter makes sense, however is it possible to emit a single wavefront? Given that even two concentric wavefronts in a single emission would probably lead to interference (?)

 

[berkeman]

View attachment 268641

By reflected wave I meant the wave that is reflected off the reflecting surface (or the object which is to be detected). My question was based on the diagram above. I was wondering if the incident wave would interfere with the reflected wave.

Not if the media is linear at the amplitudes of the incident and reflected waves (combined). It takes non-linear media for two waves to affect each other. So the answer for air and water is "no".

 

[Baluncore]

My question was based on the diagram above. I was wondering if the incident wave would interfere with the reflected wave.

There are many techniques available to separate the transmitted and reflected signals. Before anyone technique can be identified you need to specify a particular application and situation, rather than the most general case.

For a continuous wave radar with a fixed frequency, the detected signal will have a doppler frequency determined by the radial velocity of the target.

Alternatively, by transmitting a linear frequency sweep, (chirp), the transmitted and reflected signals can be separated. A synchronous detector multiplies the received signal by the transmitted signal which generates a difference frequency. The detected transmitted signal will then appear as a drifting DC component in the detector. The detected reflected signal will have a frequency offset determined by the change in transmitter frequency while the reflected signal was in transit.

 

[JC2000]

Not if the media is linear at the amplitudes of the incident and reflected waves (combined). It takes non-linear media for two waves to affect each other. So the answer for air and water is "no".

By linear media do you mean non-dispersive media?
I looked up the definition and it says that a non-dispersive medium is a medium where the speed of a wave is independent of its frequency.

Does this mean that for ER in a non dispersive medium, $c = f \lambda$ does not hold?
Based on my possibly erroneous understanding of optical interference, how are the interference fringes created if air is linear and thus the amplitudes do not interact? Or is it that the surface on which the fringes are created is non linear and hence the fringes are visible as superposition occurs only on that surface?

This is the first time I have come across these terms(!), where can I start reading about it in a systematic manner?

 

[JC2000]

There are many techniques available to separate the transmitted and reflected signals. Before anyone technique can be identified you need to specify a particular application and situation, rather than the most general case.

I see.

Where can I learn more about the chirp signal? From your answer I have got some sense of how this process is done but I think it would be useful to thoroughly understand the entire concept.

 

[berkeman]

By linear media do you mean non-dispersive media?

Not necessarily. Dispersive means what you found (propagation velocity depends on frequency), and non-linear means that the properties of the media depend on the amplitude of the wave.

Based on my possibly erroneous understanding of optical interference, how are the interference fringes created if air is linear and thus the amplitudes do not interact? Or is it that the surface on which the fringes are created is non linear and hence the fringes are visible as superposition occurs only on that surface?

Simple optical interference is just constructive and destructive interference. The two waves add and subtract depending on the relative phases.

Non-linear effects would be like when the media's index of refraction can change depending on the amplitude of the wave or set of waves. Think about a mass on a spring -- as long as the vibrations are small, you are in the linear part of Hooke's Law $F = -k\delta{x}$. But if the amplitude gets big enough, you start getting into plastic deformation of the spring, and the linear version of Hooke's Law doesn't work anymore...

 

[Baluncore]

Where can I learn more about the chirp signal?

https://en.wikipedia.org/wiki/Continuous-wave_radar#Sawtooth_frequency_modulation

 

[.Scott]

Where can I learn more about the chirp signal? From your answer I have got some sense of how this process is done but I think it would be useful to thoroughly understand the entire concept.

I just hunted around for a better description of this - to no avail. So let me describe it.
By the way, I work on a team that develops automotive radar.

There are actually several transmit and receive antennas on the face of the radar unit - and these extra antennas along with appropriate software allow the system to determine the direction of the target as well as it's range.

But let's stick with just two antennas - 1 transmit, 1 receive. The radar stuff is all driven by a single MMIC chip. The linear frequency modulated signal is generated and sent to two places: The transmit antenna and a mixer. The signal from the receive antenna is also directed to the mixer. The mixer performs a simple analogue multiply operation - and it doesn't need to be a good multiply. When you multiply two sine waves, one of the frequency components is the frequency difference. This signal is directed to a low-pass filter that blocks everything except that frequency difference.

That signal is then digitized and transmitted off the MMIC chip to more conventional digital electronic technology. Where it is immediately processed through an FFT - converting it from the time domain to the frequency domain. Since the frequency is proportional to range, this provides a range map. As an example, let's say that we collect 2048 samples and do a FFT that converts this into a 1024 complex value - we lose half the sampling bins but keep the phase information.

Then you collect this range map for several more chirps. How many depends of the requirements and design of the radar - but let's say you collect 256. But a warning: I am simplifying this.

Now you have 256 rows of range data. And as you scan down from the top row to the bottom row, you will see some targets holding their range, some moving away from the sensor, and others moving towards it. We want that Doppler information. To understand the next step, imagine what is happening to the signal from a single target moving away from the sensor at a constant speed. The round trip distance to it can be measured in wavelengths. At the mixer, this would result in oscillating constructive and destructive interference - proportional to the range component of the target's speed. And, for any given target, that oscillation will appear in those 256 rows.

So the next step: Apply a 256-element FFT to each column of range bins. What we will get is a "peak" where the Doppler matches the corresponding 256-element FFT frequency. The entire array is now called a range-Doppler map.
You actually get the speed as a modulo value - so you need to pick you chirp width and other parameters to get the Doppler resolution and range that you need. For example, let's say that you design things to give you a modulo value of 120 mph - and let's say that a particular target shows up in the 50 mph bin. It could be approaching at 50 mph, 170 mph, or 290 mph or it could be moving away at 70 mph, 190 mph or 310 mph. This ambiguity can be resolved by capturing another range Doppler map and attempting to match up targets on one R/D map with the next by actually looking for the targets moving from one range bin to the next.

So... How "thoroughly" do you want to understand this concept?

 

[Baluncore]

https://www.radartutorial.eu/02.basics/Frequency Modulated Continuous Wave Radar.en.html