very nice talk, one trick on RX to get chirp compression with low time sidelobes, is to time window the reference chirp signal before being placed in the rx matched filter. cheers
Dear professor,thank you for the perfect explanation.But I have a question,when we use FFT to ensure the position of one object,for chirp from Rx?or for multipling chirp from TX and RX？
I'm getting a little bit confused here, is the signal that you send similar to a barker code or is it a complete different thing ? I heard that barker code were designed to have a thin peak after autocorelation
At first, thanks a lot for the nice tutorial. all make sense. worth to mention, that in practical chirped radar both s(t) and matched filer is realised in analog domain. which makes it really elegant solution.
The reason to use chirp signal instead of others that was given in video can also be used for other form of signals. The main advantage of chirp signal is its capability to provide Doppler frequency shift of moving targets, directly and with controllable resolution.
Could you please answer my very basic question? Why digital radar (non magnetron radar) only need very small power compared to analog (magnetron) radar? Regards.
I think, the question is rather whether you got a primary RADAR, where you are receiving only reflections from the target or secondary RADAR, where the target, i.e. an air plane is sending back a signal containing information like identification, height etc. The primary needs high power, the secondary needs low power only.
Doesn't the broader frequency spectrum of the chirp help locate objects of different sizes? I always thought sonar pings were like that because the wave would pass right 'through' objects smaller than the wavelength.
I think you're mixing a few things up. Wider bandwidth provides more accurate timing resolution. So features of an object (target) can be resolved with higher accuracy. In terms of your comment about sonar, it's important to remember that sonar pulses are compression waves (not transversal like EM waves). Water particles are actually being compressed and decompressed as the wave propagates, so there is no chance of the wave "passing through" objects.
thanks you, does that means the higher frequencies pass the system faster than the lower frequencies? Because you get one sharp peak, but one send a signal over a longer time, or does that filter at the receiver (s^*) produces the narrow peak?
Dear Professor, thank you very much for the tutorial, this video has been a tremendous help for me when I was starting to learn about radar, this really has made it easier for me to learn all the follow-up material regarding radar. I have one question if you don't mind me asking. I am currently operating a radar instrument, inside the module, a hardware for signal mixing to retrieve the delta function signal (dechirping) is already installed. In the manual, I read that signal mixing is done, in mathematical sense by multiplying the complex conjugate of the received signal and the transmitted one. I am curious about how and why the complex conjugate operation is performed to the received signal. I am aware that complex conjugate means the inversion of sign to the phase or the reversing of oscillation in real world sense. But I don't understand how it's actually happening to the received signal. I would really appreciate if you can help me with an answer because I cannot seem to find one by searching myself and I don't really have any background on electrical or communication science. Thank you very much again for the wonderful tutorial.
I'm glad you like the video. The complex conjugate is a phase rotation, that inverts the original phase from the transmitter. This video might help you to understand this process more: "What is a Matched Filter?" ua-cam.com/video/Ci-EjiMJo3I/v-deo.html and also this one might help too "What Does "Linear Phase" Mean?" ua-cam.com/video/aQ__XatMxJo/v-deo.html
Hi sir can radar read iron deep buried using freaquency?i dont have knowledge about electonics but i see some module generator they used in making locators
Hi professor, once again a great tutorial! I have a small question though. If we have a scenario where the Doppler frequency changes with time, for example in a LEO satellite or in underwater communications, then can that signal be also called as a chirp signal? Thanks!
Yes, that would cause the frequency to change with time. But it would be over a longer time period. LEO satellites move fast, but not as fast as the chirp sweep used in usual radar signals.
@@iain_explains this will be a weird request... I'm a computer engineer in the US with 17 years of experience who wants to move somewhere else and I'm seriously considering Australia. I've looked into work visas but it seems it's best to find a sponsor employer to start the process. My specialty is PCB, firmware, and software design. I was thinking about Sydney, Perth, or Darwin (what a list right haha). If there is anything you could do to help direct me toward a business that could use my skills, or some way to find such a business, that would be amazing. I have no idea where to start. Thanks for your consideration. I've been a sub to your channel for a while. Helps me remember some of the skills I learned and haven't used in a while.
Sorry, I'm not across the current job market for PCB, firmware, and software design. Certainly there are lots of companies doing this here in Australia, so I'm sure you'll find opportunities. Good luck in your search.
Thank you, professor. In the case of MIMO radar, can the radar send chirp signals for its operation or different types of signals?. as a request, we need more videos about signal processing of the radar systems.
Dear professor, I have some difficulty in reading paper on wireless communciations about derivation process and models (including how to summarize a paper and how to get the "point" of the paper and so on). If you could give some advice about this, I would appreciate it!!! Thanks a lot!
Yes, papers are often written with a focus on the technical details, rather than on the "explanation". I'll give some thought to making a video as you've suggested.
Thank you, professor. After matching the filter different targets will produce different peaks. But I am considering that if the two targets are close together this means that it is difficult to produce the scene as you have depicted, and instead the second peak appears halfway down the first peak. Is it possible that the falling part of the first peak will affect the value of the second peak? If this holds true what method should be used to separate the two peaks? Thanks again for your clear explanation.
Yes, that's right. That's always the challenge in radar - how to seperate close reflectors/targets. In general, the wider the bandwidth of the chirp, the narrower the returning "impulse" that comes out of the receiver filter. The returns from different targets add linearly, so there are many "deconvolution" techniques that have been developed to try to seperate the returns. Here's a reference to a paper I wrote on this topic back in 1996 (so long ago it makes me feel old!): I.B. Collings and D.A. Gray, "Deconvolution Techniques for Non-coherent Radar Images'', in Proc. of the Int. Symp. on Signal Processing and its Applications (ISSPA), Gold Coast, Australia, pp. 113-116, August 1996.
Hi Prof. Ian, can you perhaps make a video where it is about non-bijective mappings of random variables, e.g. as in a PDF that is uniformly distributed with line of symmetry at the origin. So one half is in the negative and the other half is in the positive and where the task is also about quantization levels.
@@iain_explains as far as suggestions go what would be nice is to make a series on the practical skills needed in SP. Not in depth but like what each do and where they come into play and why. Say Matlab and then C to run the algos and why we choose C and practical reasons why you choose X algo for an application etc. Also a "raodmap" of SP would be cool. Comms will generally use XYZ radar ABC audio DEF. That would be useful. good vids though learning a lot
I’m sorry, dumb question. Early in the video you say something that sounds like “rect function”. Is this what you are saying? If not, can you clarify what your are saying?
You can find the details on Wikipedia, as follows (written in LaTeX): {\displaystyle x(t)=\sin \left[\phi _{0}+2\pi \left({\frac {c}{2}}t^{2}+f_{0}t ight) ight]} where c is the chirp rate, given by {\displaystyle c={\frac {f_{1}-f_{0}}{T}}}
Dear Prof. Ian, could you please explain why even a mean-free white noise can have a constant spectrum? Doesn't "zero mean value" and a constant spectrum, including at 0 Hz, contradict each other? I always thought 0 Hz = DC = mean value. I hope my question is reasonably understandable. Thanks for your awesome curses!
Great question. The "spectrum" that you're referring to is the Power Spectrum. This can (sort of) be viewed as an "average spectrum of the squared value". The term "white noise" refers to a Random Process - not an exact specific waveform. Each exact specific waveform has a Fourier Transform. But when you are talking about "random waveforms" (or more specifically, Random Processes) you need to talk about its "average" properties. For more details, see: "Autocorrelation and Power Spectral Density (PSD) Examples in Digital Communications" ua-cam.com/video/XWytSLZZP1A/v-deo.html
this was the 4th video I have seen on this. Also was the only one that made some sense. THanks for posting this.
I'm glad you found it helpful.
very nice talk, one trick on RX to get chirp compression with low time sidelobes, is to time window the reference chirp signal before being placed in the rx matched filter. cheers
Yes, good point. I should add "time windowing" to my topic list for a future video.
@@iain_explains often a hamming window is used
Good stuff! Thanks for such well explained intuitive content. Subscribed!!
Awesome. I'm glad you like the channel!
Dear professor,thank you for the perfect explanation.But I have a question,when we use FFT to ensure the position of one object,for chirp from Rx?or for multipling chirp from TX and RX？
I'm getting a little bit confused here, is the signal that you send similar to a barker code or is it a complete different thing ? I heard that barker code were designed to have a thin peak after autocorelation
At first, thanks a lot for the nice tutorial. all make sense.
worth to mention, that in practical chirped radar both s(t) and matched filer is realised in analog domain. which makes it really elegant solution.
Yes. Good to point that out.
The reason to use chirp signal instead of others that was given in video can also be used for other form of signals. The main advantage of chirp signal is its capability to provide Doppler frequency shift of moving targets, directly and with controllable resolution.
Der professor,
Thank you for this wonderful explanation. You can make them practical to understand. Everything becomes clear when I watch your courses
I'm so glad to hear you're finding the videos helpful.
Good show. The idea that chirp is an initialism never occurred to me, but it makes sense now.
thanks, sir for the video I have done my project using chirp as a basis function it has made my concepts clearer now
Glad it helped
Could you please answer my very basic question? Why digital radar (non magnetron radar) only need very small power compared to analog (magnetron) radar?
Regards.
Sorry, I'm not a hardware expert.
I think, the question is rather whether you got a primary RADAR, where you are receiving only reflections from the target or secondary RADAR, where the target, i.e. an air plane is sending back a signal containing information like identification, height etc. The primary needs high power, the secondary needs low power only.
Thank you, Please can make a video explaining the signal form in each LTE stage transmitter. Thanks
Do you have anything specific in mind? Have you seen that I've got lots of videos on the channel about OFDM?
Doesn't the broader frequency spectrum of the chirp help locate objects of different sizes? I always thought sonar pings were like that because the wave would pass right 'through' objects smaller than the wavelength.
I think you're mixing a few things up. Wider bandwidth provides more accurate timing resolution. So features of an object (target) can be resolved with higher accuracy. In terms of your comment about sonar, it's important to remember that sonar pulses are compression waves (not transversal like EM waves). Water particles are actually being compressed and decompressed as the wave propagates, so there is no chance of the wave "passing through" objects.
Thanks, on the upper left image, isn't the period time a bit smaller?
It seems as if T marks the double period time.
I have used T to represent the length of the pulse (not the period of the sinusoid).
thanks you, does that means the higher frequencies pass the system faster than the lower frequencies?
Because you get one sharp peak, but one send a signal over a longer time, or does that filter at the receiver (s^*) produces the narrow peak?
This might help: "ua-cam.com/video/2kMSLqAbLj4/v-deo.html" ua-cam.com/video/Ci-EjiMJo3I/v-deo.html
really fascinating concept enjoyed the video thanks for it !
Glad you enjoyed it!
Thanks a lot, Professor, Could you kindly explain LoRa as well, Is there any reference for Matlab Coding ?
Thanks for the topic suggestion. I've put it on my "to do" list.
sir It was one of the hard topics for me before I saw this video of yours you are a life saviour thanks sir 😊
I'm so glad to hear that the video has helped you.
Dear Professor, thank you very much for the tutorial, this video has been a tremendous help for me when I was starting to learn about radar, this really has made it easier for me to learn all the follow-up material regarding radar.
I have one question if you don't mind me asking. I am currently operating a radar instrument, inside the module, a hardware for signal mixing to retrieve the delta function signal (dechirping) is already installed. In the manual, I read that signal mixing is done, in mathematical sense by multiplying the complex conjugate of the received signal and the transmitted one. I am curious about how and why the complex conjugate operation is performed to the received signal. I am aware that complex conjugate means the inversion of sign to the phase or the reversing of oscillation in real world sense. But I don't understand how it's actually happening to the received signal.
I would really appreciate if you can help me with an answer because I cannot seem to find one by searching myself and I don't really have any background on electrical or communication science. Thank you very much again for the wonderful tutorial.
I'm glad you like the video. The complex conjugate is a phase rotation, that inverts the original phase from the transmitter. This video might help you to understand this process more: "What is a Matched Filter?" ua-cam.com/video/Ci-EjiMJo3I/v-deo.html and also this one might help too "What Does "Linear Phase" Mean?" ua-cam.com/video/aQ__XatMxJo/v-deo.html
How are the filters implemented?
Hi sir can radar read iron deep buried using freaquency?i dont have knowledge about electonics but i see some module generator they used in making locators
Yes, there are companies that make ground-penetrating radars for a range of applications.
Thanks for this. Helps to understand that it's complicated.
Glad it was helpful!
Hi professor, once again a great tutorial! I have a small question though. If we have a scenario where the Doppler frequency changes with time, for example in a LEO satellite or in underwater communications, then can that signal be also called as a chirp signal? Thanks!
Yes, that would cause the frequency to change with time. But it would be over a longer time period. LEO satellites move fast, but not as fast as the chirp sweep used in usual radar signals.
Fantastic Explanation.
Thanks. Glad you liked it.
whta is a narrowband waveform?
amazing professor with an amazing concepts thanks lain!
Thanks for your nice comment. I'm glad you like the videos.
Sir can you do a video on range Doppler compensation algorithms?
Thanks for the topic suggestion. I've put it on my "to do" list.
Dear professor, can you make a video that explain what is "Rate-Splitting Multiple Access" ?
Thanks for the suggestion Tường. I've put it on my "to do" list.
Excellent video. Very clear presentation. Is your accent from Australia?
Yeah mate! 😁
@@iain_explains this will be a weird request... I'm a computer engineer in the US with 17 years of experience who wants to move somewhere else and I'm seriously considering Australia. I've looked into work visas but it seems it's best to find a sponsor employer to start the process. My specialty is PCB, firmware, and software design. I was thinking about Sydney, Perth, or Darwin (what a list right haha).
If there is anything you could do to help direct me toward a business that could use my skills, or some way to find such a business, that would be amazing. I have no idea where to start.
Thanks for your consideration. I've been a sub to your channel for a while. Helps me remember some of the skills I learned and haven't used in a while.
Sorry, I'm not across the current job market for PCB, firmware, and software design. Certainly there are lots of companies doing this here in Australia, so I'm sure you'll find opportunities. Good luck in your search.
@@iain_explains no worries, you miss all of the shots you don't take, was worth a try. Thanks!
Thank you, professor. In the case of MIMO radar, can the radar send chirp signals for its operation or different types of signals?. as a request, we need more videos about signal processing of the radar systems.
Yes, MIMO Radar can use chirp signals. Thanks for your suggestion of more videos on radar. I will add this to my "to do" list.
Dear professor, I have some difficulty in reading paper on wireless communciations about derivation process and models (including how to summarize a paper and how to get the "point" of the paper and so on). If you could give some advice about this, I would appreciate it!!! Thanks a lot!
Yes, papers are often written with a focus on the technical details, rather than on the "explanation". I'll give some thought to making a video as you've suggested.
@@iain_explains Your reply really inspires me and give me a hope! I am looking forward to seeing it!
Thank you, professor. After matching the filter different targets will produce different peaks. But I am considering that if the two targets are close together this means that it is difficult to produce the scene as you have depicted, and instead the second peak appears halfway down the first peak. Is it possible that the falling part of the first peak will affect the value of the second peak? If this holds true what method should be used to separate the two peaks? Thanks again for your clear explanation.
Yes, that's right. That's always the challenge in radar - how to seperate close reflectors/targets. In general, the wider the bandwidth of the chirp, the narrower the returning "impulse" that comes out of the receiver filter. The returns from different targets add linearly, so there are many "deconvolution" techniques that have been developed to try to seperate the returns. Here's a reference to a paper I wrote on this topic back in 1996 (so long ago it makes me feel old!): I.B. Collings and D.A. Gray, "Deconvolution Techniques for Non-coherent Radar Images'', in Proc. of the Int. Symp. on Signal Processing and its Applications (ISSPA), Gold Coast, Australia, pp. 113-116, August 1996.
@@iain_explains thank you so much, professor. I will read this paper.
Thank you, clear and simple!!
Glad it helped!
Prof. Iain, you rocks!!
Thanks a lot!
Nice one! I'm glad you like the videos.
Hi, all your videos are really interesting
Glad you like them!
Hi Prof. Ian, can you perhaps make a video where it is about non-bijective mappings of random variables, e.g. as in a PDF that is uniformly distributed with line of symmetry at the origin. So one half is in the negative and the other half is in the positive and where the task is also about quantization levels.
Sounds like an interesting question, but I'm not exactly sure what you mean. Can you be more specific?
@@iain_explains as far as suggestions go what would be nice is to make a series on the practical skills needed in SP. Not in depth but like what each do and where they come into play and why. Say Matlab and then C to run the algos and why we choose C and practical reasons why you choose X algo for an application etc. Also a "raodmap" of SP would be cool. Comms will generally use XYZ radar ABC audio DEF. That would be useful. good vids though learning a lot
I’m sorry, dumb question. Early in the video you say something that sounds like “rect function”. Is this what you are saying? If not, can you clarify what your are saying?
You'll have to give me a time-stamp. I don't have time to watch my videos through, trying to find specific words I may have said, sorry.
Thank you prof. Ian, Can I have what you talk about in the video in writing?
Yes. Summary sheets for all of my videos can be found at iaincollings.com
@@iain_explains Prof. Ian there are only images, no text
Can you do a video on LoRa technology?
Thanks for the suggestion. It's on my "to do" list.
Interpulse modulation? After match filter??
No, I said Intra-Pulse Modulation. ie. modulation within the pulse. In other words, changing (modulating) the frequency within the pulse.
Hi professor, what is the mathmatic description of the chirp signal? It looks like a A*cos(w*exp(t)*t), where t = 0:T to me but I cannot be sure.
You can find the details on Wikipedia, as follows (written in LaTeX): {\displaystyle x(t)=\sin \left[\phi _{0}+2\pi \left({\frac {c}{2}}t^{2}+f_{0}t
ight)
ight]} where c is the chirp rate, given by {\displaystyle c={\frac {f_{1}-f_{0}}{T}}}
Is this the same as LFMOP?
Yes.
Dear Prof. Ian, could you please explain why even a mean-free white noise can have a constant spectrum? Doesn't "zero mean value" and a constant spectrum, including at 0 Hz, contradict each other? I always thought 0 Hz = DC = mean value. I hope my question is reasonably understandable. Thanks for your awesome curses!
Great question. The "spectrum" that you're referring to is the Power Spectrum. This can (sort of) be viewed as an "average spectrum of the squared value". The term "white noise" refers to a Random Process - not an exact specific waveform. Each exact specific waveform has a Fourier Transform. But when you are talking about "random waveforms" (or more specifically, Random Processes) you need to talk about its "average" properties. For more details, see: "Autocorrelation and Power Spectral Density (PSD) Examples in Digital Communications" ua-cam.com/video/XWytSLZZP1A/v-deo.html
@@iain_explains Thank you very much for the good explaination! :)
This is pretty cool. Thank you.
I'm glad you liked it.
Dear sir, we found one unknown signal from satellite spectrum , i am not sure if from radar. C-band frequency around 4GHz.
There are some satellites that use C band for their synthetic aperture radar payloads.
Excellent!
Glad you liked it!
I think LoRa uses Cihrps too.
Yes, that's right. It can be viewed as a version of "spread spectrum".
By reading the title, my first reaction is because it sending a pulse signal (Delta function)
A "pulse signal" is not the same thing as a "Delta function". A "pulse" has a time duration.
I don't think chirp is an acronym. It's just chirp after its very characteristic sound.
Compressed High Resolution Pulse Radar
Word salad .